.
GLEN CANYON
ENV!RONMENiAL STUOIES
P. O. BOX 1811
.
FLAGSTAFF, AZ 86002

GLEN CANYON DAM RELEASES AND DOWN STREAM
RECREATION: AN ANALYSIS OF USER
PREFERENCES AND ECONOMIC VALUES

l

(U.S.) Glen Canyon Environmental Studies
Flagstaff, AZ

Jan 87

71'0.oo

LND--z.oo
G 55~

•
1
1

rqo 4- t
v.3

e.3

PB88-183546

P B 8 B- 1 8 3 5 ..+ 6

..
•
GLEN CANYON DAM RELEASES AND
DOWNSTREAM RECREATION: AN
ANALYSIS OF USER PREFERENCES
AND ECONOMIC VALUES

Recreation
of the
Glen Canyon Environmental Studies

By
Richard C. Bishop
Kevin J. Boyle
Michael P. Welsh
Robert M. Baumgartner
Pamela R. Rathbun
HBRS

•

January, 1987

REPRODUCED BY

U.S. DEPARTMENT OF COMMERCE
NATIONAL TECHNICAL
INFORMATION SERVICE
SPRINGFIELD, VA. 22161

50272-10

REPORT DOCUMENTATION , l.

AAGE

4.

~

•

.

3.

REPORT NO.
GCES/27/87

Title and Subtitle
Glen Canyon Darn Releases and Downstream Recreation:
An Analysis of User Preferences and Economic Values.

Author<s> Richard C. Bisho??, Kevin J. Boy le, Mi chae 1 ~.
Welsh. Robert M. BaumoarrnAr. ~nn P~mAla R. R~rhh11n
9. Performinc Orcanizatlon Name and Address
Huberlein - Baumgartner Research Service
SBRS Executive Building
4513 Vernon Boulevard
Madison, ~VI 53705

1.

Recipient's Accession No.

Report Date
January 1987

5.

~

8. Performinc Orcanization Rept. No.
10.

Project/Task/Work Unit No.

Contract(C) or Grant(G) No.
cc>4-CS-40-0l 780

11.

(GJ

13. Type of Report & Period Covered
Sponsorinc Orcanization Name and Address
Final
Glen Canyon Environmental Studies
Bureau of Reclamation, Upper Colorado Region
14.
P.O. Box 11568
Salt Lake Citv TT~ R4147-1~nR
15 Supplementary, NQtes
.
,
,
. ,
Freparea in cooperation with the Glen Canyon Environmental Studies.
Report in two parts: Final Report (201 oages) and Ao9endices (195 pages~
for a total of 396 pages.
u;. Abstract (Limit: 200 words)
This study assesses the impact of Glen Canyon Dam releases on rafting
(white-water boating and day-use rafters) an angling recreationists in
Glen Canyon and Grand Canyon National Park using attribute and contingent
valuation surveys. Several sources of information were utilized in this
study: knowledgeable people (fishing quides, rafting guides, resource
managers, and GCES researchers), seven formal survys (includinq attribute
surveys), and contiDgent valuation survey to quantify, in dollars, the
effects of dam releases on the recreational exoerience. The goal of the
study was to assess the imoact of alternative annual flow release oatterns
f0r Glen Canyon Darn on recreationists in the aqgregate. Flow regimes
combining high constant flows in the summer months with moderate or low
flows during the remainder of the year would be likely to oroduce the
largest recreational benefits. Extreme high or low flows will adversely
affect all river recreation, with flows below apt?roximately 5,000 cubic
feet per second and above 35,000 cubic feet oer second to both boaters
and anglers.

12.

17.

•

Document Analysis a. Descriptors

b. Identifiers/Open.Ended Terms

c. COSATI Field/Group
lL Av.ilabllity Statement
i~o restriction on distribution
Available from :~ational Technical Inf or-

Security Class (This Report)
I 21. No. of Paees
UNCLASSIFIED
3 9~
20. Security Class <This Pase>
- i 22. Prire ·
.__~m~~~......,~;~~~-......,.c~~~.....,-··~·;~=~-......-·_s~p-r_i~·n~.~q.~f~i~e~l-d~.--=V~A~..;;.2~2~1~6~1;_......i.___u~~-C~L-A~S~S~I-F:o-=I~E~D:...-~-~~--~·--~~·~<see ANSl-Z39.18)
See Instructions on Reverse
OPTIONAL FORM 272 (4-77)
-.
(formerly NTIS-35)
Department of Commerce
1
_...;.-19.

j

ABSTRACf

•

The research reported here is part of the Glen Canyon Environmental
Studies (GCES), a series of studies examining the downstream effects of
alternative Glen Canyon Dam releases on the environments of Glen Canyon
National Recreation Area and Grand Canyon National Park. This
particular study focused on recreational effects. Three types of
recreation were- identified as being potentially affected by river
flows: white-water boating, trout angling, and one-day raft trips.
Seven formal surveys were conducted. One was a mail survey of
white-water commercial trip guides and private trip leaders to
ascertain their views on how dam releases affect white-water boating.
Samples of white-water boaters, anglers and day-use rafters each
participated in "attribute surveys" to identify the important
attributes of recreational quality and define which are sensitive to
flows. Separate samples from the three recreation groups also
participated in contingent-valuation surveys to quantify, in dollars,
the effects of flows. In addition, this study used the expertise of
fishing guides, rafting guides, resource managers, and other GCES
researchers.
Glen Canyon Dam releases have substantial impacts on both trout fishing
and white-water boating. Anglers placed the highest value, $126 per
trip, on constant flows of about 10,000 cubic feet per second (cfs).
Both lower and higher flows adversely affect fishing values. For
example, a fishing trip is worth only $60 at a constant flow of 3,000
cfs and $94 at a constant flow of 25,000 cfs. White-water boaters
prefer constant high flows in the 29,000 to 33,000 cfs range, with
maximum values of $898 per trip for commercial passengers and $688 per
trip for private white-water boaters. At a constant flow of 20,000
cfs, the commercial trip value declines to about $550, while the
comparable figure at 45,000 cfs is $732. Daily fluctuations in flows
are detrimental to both groups.
Day-use rafters' reported values were not affected by the range of
alternative flows evaluated in this study.

•

Although flow preferences of white-water boaters-and anglers appear to
conflict, use data indicates that with the exceptions of f.!ay and
September, the greatest use levels for each activity occur at different
times of the year. Flow regimes combining high constant flows in the
summer months of May-September with low to moderate constant flows for
the remainder of the year would produce the largest recreational
benefits. Under ideal conditions, annual benefits could be as high as
$12.4 million.

ACKB<JII.EDGMEBTS

•

Each stage of this research project required a careful assessment of
the results of previous stages and a well thought out plan for the
accomplishment of future stages in the research process. Members of
the Recreation Subteam made significant contributions to this process
and the overall success of the study. Members of the Recreation
Subteam are:
David Wegner, GCES Manager, U.S. Bureau of Reclamation
Michael O'Donnell, GCES Recreation Subteam Leader,
U.S. Bureau of Reclamation
Martha Hahn-O'Neill, Grand Canyon National Park
Curtis Brown, U.S. Bureau of Reclamation
Larry Belli, Glen Canyon National Recreation Area
John Loomis, U.S. Fish and Wildlife Service
Tom Heberlein and Bo Shelby of HBRS also provided considerable
experti~e in the develoµnent of the survey instruments and
interpretation of the data.
In addition, a number of others made important contributions to this
study. In alphabetical order, they are:
Nancy Brian, U.S. Bureau of Reclamation
Nancy Brocato, U.S. Bureau of Reclamation
Sue Cherry, Grand Canyon National Park
Kim Crumbo, Grand Canyon National Park
Deborah Gibbs, U.S. Bureau of Reclamation
Steve Hodapp, Grand Canyon National Park
Jay Hokenstrom, U.S. Bureau of Reclamation
Robert K. Lanky, U.S. Bureau of Reclamation
Lauren Lucas, U.S. Bureau of Reclamation
Henry ~addux, Arizona Game and Fish
Mike Rolute, U.S. Bureau of Reclamation
Larry Schluntz, U.S. Bureau of Reclamation
Alice Thomas, Western Area Power Administration
John Thomas, National Park Service
Beth Wedge, Grand Canyon National Park

•
•

Finally, we would like to thank ~argaret Syse of HBRS who typed and
revised numerous drafts of this final report, as well as Richard
Prestine and Brian Klockziem of HBRS who performed much of the data
management and computer work for this project .

TABLE OF OORTEHTS

'
•

Page
LIST OF TABLES •

v

LIST OF FIGURES.

ix

..

.
CHAPTER 1• INTRODUCTION . .
Background • . • • •
. • • • • • • •
Recreation and Stream Flows.
• • • • •
Goals and Approaches of the Study ••
Study Plan
. • • .
Summary • • • • • • •

1
1

2

3
6
6

CHAPTER 2. RECREATION VALUES AND PUBLIC DECISIONS ••
Introduction . . . •
• • • .
• • • • •
. • • .
Why Dollar Values? • • . • • .
. • • • . •
• ••••
Value Definitions. • • . • • •
• • • • • • •
• •••
Benefits ~qual Surplus Values, Not Expenditures.
• ••••
Adaptation of Value Concepts to Colorado River Recreation.
Why Not the Travel-Cost Method?. . . • .
• • • .
On the Validity of Contingent Valuation • • • • • •
Summary . • • . •

9
9
9

CHAPTER 3. RESE!RCH PROCEDURES.
Introduction • .
Survey Procedures. .
Background . .
Sampling • . • •
Survey Techniques.
Survey Pretests. • . . • • • . • • . . • . .
Survey Implementation •.
Designing the Attribute Surveys.
Designing the Contingent-Valuation Surveys . • • • . .
Values to be Estimated .
Definition of Item • • . . . •
Payment Vehicle. . . • . . • .
Choice of a Contingent-Valuation Technique •
Data Analysis. • • • • • •
Supplemental Data Required
Estimating Surplus Values-from Dichotomous-Choice Responses.

26
26

11

15
16
18
19
25

27
27

28
28
29
30
31

33
35
35
37
37
37
38
42

45

S 1.lIIllila ry. . . . . . . . . . . .

CHAPTER 4.

..

SURVEY OF WHITE-WATER RAFT TRIP COMMERCIAL GUIDES
AND PRIVATE TRIP LEADERS .
Introduction . . • •
Background . • . . .
Survey Procedures. •
Sampling • • • . • • • •
Response Rate. • • • • • • • • • • • •
Guide Survey Results
Evaluation of Constant Flow Levels
i

49
49
49
50
50
50
51
52

TABLE OF CONTENTS (Continued)

Effect of Constant Flows on Attributes of a White-Water
Trip . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
61

Evaluation of Daily Fluctuations in Flow Level . . • . .
Effect of Daily Fluctuations on Attributes of White-Water

Trips. . . . . . . . . . . . . . . . . . . . . . . . . . . 63

.

Preference Ratings of Four Flow Release Scenarios.
• • . 65
Evaluation of Scenarios.
• • • • • • • •
• 68
Summary. • • • • • • • . •
. . . •
72
CHAPTER 5. WHITE-WATER BOATER SURVEYS AND RESULTS •
Introduction • • • • • • • • • • • .
• • • •
Background • . . • • • • .
Attribute Survey Procedures. •
Sampling . . . . . .

. . . . . . . .

• .
.
•
. • .

CHAPTER 6. GLEN CANYON ANGLER SURVEYS AND
Introduction
Background· .
. . • . .
Attribute Survey Procedures.
Sampling • . • . . .
Response Rate. . . • . .
Attribute Survey Results . . .
Glen Canyon Angling Experience • .
Important Attributes • • • . •
Contingent-Valuation Survey Procedures
Sam.pling

. . .

RESULTS

76
. 76
• • • • • 76

• • • •
ii

• .

. . .

78
78
81

83
83
87
89
89
91
91
91
94
96
96
98
99
100
101
102
10 3

. • 108
108
108
109
. . 1C9
110

. • • •
. • • • •
• • • • •
•

. . . . . . . . . . . . . .

Response Rate. . • • • • •

74

. . . . .

_ Response Rate. • • •
• • • • • • • •
Attribute Survey Results
. • • • •
• •••..
Trip Attributes~ • • .
Attributes of Rapids •
• •••••••
Attributes of Campsites. . • • • . . • • • • • • • • .
• .
Sensitivity of Selected Attributes to Flows.
. •.
Respondent Perceptions of Flow Levels.
. ..
Contingent-Valuation Survey Procedures
. • • •
Sampling . . • • . • • . • . • .
Response Rate. . . . . . • . • .
. . . • .
Contingent-Valuation Survey Results.
.
Actual Trip. . • • • • • • • • . • . . •
•
Scenarios. . . • • . • • • • . • .
.
Case 1 - Constant Flow of 5, 000 cfs. . • . •
. • • • •
Case 2 - Average Flow of 5,000 cfs with Daily Fluctuations •
Case 3 - Constant Flow of 13,000 cfs . . • • • . . . .
Case 4 - Constant Flow of 22,000 cfs . . . . . . • . .
• .
Case = - Average Flow of 22,000 cfs With Fluctuations.
Case 6 - Constant Flow of 40,000 cfs .
.
Change 1 - Beaches Reduced
• • • •
•
SllID.IIla ry . . . . . . . . . . . . . . .

74
74

• 110
. • 110
. . • . • 112
. . . . • • . 117

. . . • • • 117
• 118

•

TABLE OF CONTENTS (Continued)

'

Contingent-Valuation Survey Results. •
Actual Trip. . •
. . • . . .
Scenarios. . . . . . . . . . . .

•

Case 1
Case 2
Case 3
Case 4
Case 5
Case 6
Case 7
Change
Change

1
2

. • • • •

• 1f9
• 119

. . . . . . . .

Constant Flow of 3,000 cfs.
• •••.•
Average Flow of 3, 000 cfs with Fluctuations •
•
Constant Flow of 10,000 cfs . • • • • • • •
. •
Average FJ.ow of 10,000 cfs with FJ.uctuations.
•
Constant Flow of 25,000 cfs . • • • . . • . .
.
Average FJ.ow of 25,000 cfs with FJ.uctuations • • • •
Constant Flow of 40,000 cfs •
• .•••.•••
- Bigger Fish • • • • • • • • •
• .••••
- Chances of ·Getting Skunked.
• • • • • .

Summary. . . . . . .

122

122
123
124
126
127
128
129
130
130

. . . • . . . . . . . . . . 131

Final Observations • . • • • • . • . . • • • .

• • • • • • 132

CHAPI'ER 7. DAY-USE RAFTERS SURVEYS AND RESULTS. .
Introduction . • . • . • . •
Background • • • • • • • • •
Attribute Survey Procedures.
Sampling . . • . . .
Response Rate. , . . . .
Attribute Survey Results . .
Contingent-Valuation Survey Procedures .

. • • • • • . 135
• 135
• 135
136
. • 136
138
. 139
• 143
143
• • 144

Sampling . . . . . . . . . . . . .

Response Rate. . • . • . •
Contingent-Valuation Survey . .
Summary. . . . . . . . . . •

• • • • • • • • • 14 5

146

CHAPI'ER 8.

EVALUATION OF ANNUAL FLOW RELEASE PATTERNS
FROM GLEN CANYON DAM .
. • . • .
Introduction • . . . .
Flow Value Functions . . . •
White-Water Boater and Angler Use Rates. .
. .•••••.•
Annual Flow Regimes Involving Releases of 8.25 Million

147
147
147
153

Acre-Feet. . . . . . . . . . . . . . . .
154
Evaluation of Annual Flow Regimes for 8.25 Million Acre-Feet • . 157
. 161
Three High-Water Regimes .
. . . .
• 168
Summary . . . . . . ! •

.

CHAPI'ER 9. CONCLUSIONS.
White-Water Boating Conclusions • . . • . .
Conclusion 1
...•••••
Conclusion 2 • .
Conclusion 3 . • . • . . . • .
• •
Conclusion 4
. . . . •
Conclusion 5 • • . • • • • • • • •
• • • •
Glen Canyon Fishing Conclusions.
. .
Conclusion 6 . . . . . . .
iii

• • • .
.....
• • • •
•
• •

170
170
17 0
171
172

. . . • • . 172
• • • • • 17 2

. . . . . . 173

173

TABLE OF CONTENTS (Continued)
Page
• • ·- • • 174

Conclusion 7 •
. • • •
. . . • . . . . 175
Conclusion 8
.•••••
• • 175
Conclusion 9
. • • • .
175
Glen Canyon Day-Use Rafting Conclusions ••
. . . . . . . 175
Conclusion 10 . . • • • . . • • . .
Conclusions From Combining White-Water Boating and
176
Fishing Values •
• 176
Cone! usion 11.
• 176
Conclusion 12. • .
. 177
Conclusion 13.
Cone! usion 14. •
177
178
Some Final Thoughts.
REFERENCES .

182

GLOSSARY •

186

iv

•

LIST OF TABLES

'
•

4-1.

Commercial Guide and Private Trip Leader Survey Response
. • • •
Rate • • . . . . .

51

4-2.

Preferred Constant Flow Levels •

52

4-3.

Reported Minimum Constant Flow Levels for Running River
Trips Safely with Passengers •
• • • .

54

Reported Maximum Constant Flow Level for Running River
Trip Safely with Passengers. • •

55

Constant Flow Level Requirements for Use and Access to
Attraction Sites • • • • • • . •

60

Constant Flow Level Requirements for Access and
Availability of Campsites. . . •
. • • •

61

Reported Mean Tolerable Daily Changes in Flow Levels for
Those Who Have Experienced Daily Fluctuations of at Least
15,000 cfs in the Grand Canyon . . •
• • . . • . . •

62

Impact of Daily Fluctuations in Flow Level on Procedures
at Major Rapids. . . . • . . . . . •
. • • • .

63

Impact of Daily Fluctuations in Flows on Campsite
Selection and Use. . . . . . . . . • . . . . • . .

64

4-4.
4-5.
4-6.
4-7.

4-8.
4-9.

.

4-10. Impact of Daily Fluctuations in Flows on Trip Itineraries.

65

4-11. Hean Ranking of the Four Flow Scenarios.

67

ll-12. Percent Ranking Four Scenarios as Unacceptable .

67

4-13. Potential Problems with Scenario D (Moderate Daily Changes).

68

4-14. Potential Problems with Scenario A (Constant Daily Flows)
. . . . .
During May through August. . . • . . • • • .

69

4-15. Potential Problems with Scenario B (Combination Flow
Pattern) During June through August. . . . . • . . . .

70

4-16. Potential Problems with Scenario B (Less Severe Daily
Changes) During September through May . • .

71

4-17. Potential Problems with Scenario C (Severe Daily Changes)
During June through August • • • • • . . •
• • • • • •

72

v

LIST OF TABLES (Continued)

5-1.

White-Water Boaters' Attribute Survey Response Rate •

• 77

5-2.

White-Water Boater Attribute Survey Response Rates by
Trip Type and Year. • • . • • •
• • • • •

• • • 77

5-3.

Attributes that Contribute Most to An Excellent or Perfect
Grand Canyon Raft Trip. • • • •
• • • . • • • • • . •

••

78

5-4.

Attributes that Contribute Most to a Poor Grand Canyon Raft Trip. 80

5-5.

Importance Ratings· of Grand Canyon Raft Trip Attributes • • • • • 81

5-6.

Important Attributes of Rapids Increasing or Decreasing
Enjoyment by User Group • • • • • •
. • • • • •

• • • -82

5-7.

Rating of Attributes of Campsites .

5-8.

Respondents Walking Around Rapids by Flow Level Experienced . • • 84

5-9.

Amount of Time for Hiking and Attraction Sites •.

. . 83

85
• • • • • 86

5-10. Reported Crowding by Flow Level Experienced . •

5-11. Reported Crowding at Campsites by Flow Level Experienced . . . . • 87
5-12. Flow Level Preferences by Flow Level Experienced. .

~

. •

• 88

5-13. -Respondents Reporting Awareness of Changes in Flow Level.

89

5-14. Evaluation of Effects of Daily Fluctuations on Perceptions
of a Natural Setting.

. . . 90

5-15. Daily Flow Levels Experienced by Respondents to the Grand
. • . .
Canyon Boater Contingent-Valuation Survey • . •

90

5-16. White-Water Boater Contingent-Valuation Survey Response Rate . . . 91
5-17. Estimated Scenario Surplus Values for White-Water Boaters
( $ PER TRIP).
.....

.....

.

5-18. Rating of the 5,000 cfs Constant Flow Scenario Relative to
.. ...
Actual Trip .

.

-5-19. Respondents'

.

.

...

95

96

Preferences for Fluctuations at Low Flow Levels . . 97

5-20. Rating of the 13,000 cfs Constant Flow Scenario Relative to
Actual Trip • • • . • • . • • • • . • • • . . • • • • . . • • . . 98
vi

;.

LIST OF TABLES (Continued)

'

5-21. Rating of the 22,000 cfs Constant Flow Scenario Relative to
Actual Trip . . . . . . . . . . . . . . .

•

. . . .

• • 99

5-22. Respondents' Preferences for Fluctuations at Moderately High
Flow Levels . . . . . . . . . . .

. . . . . . . .. ·.

. 101

5-23. Rating of the 40,000 cfs Constant Flows Scenario Relative
to Actual Trip. . . • • . • • • • • • • • . . • • •
. •• 102
6-1.

Respondent Reported Fishing Experience At Specified Flow
• • 111

Levels. . . . . . . . . . . . . . . . . .

• 112

6-2.

Respondents Having Experience With Various Flow Levels.

6-3.

Reasons For Fishing in Glen Canyon. • • • • • •

6-4.

Attributes Contributing Most to an Excellent or Perfect
Glen Canyon Fishing Trip. • • • • • . • • • • • • • • •

• 114

6-5.

Attributes Contributing to a Poor q1en Canyon Fishing Trip.

• 115

6-6.

Perceived Effects of Low Flow Levels on a Glen Canyon
Fishing Experience •.

• 117

• • 113

• • 118

6-7.

Glen Canyon Angler CV Survey Response Rate.

6-8.

Average Expenditures Reported by Anglers on the Trip When
On-Site Interview Completed • . . • . . • •
• . . . •

• 119

Estimated Glen Canyon Angler Surplus Values ($ Per Trip).

• 121

6-9.

6-10. Rating of the 3,000 cfs Constant Flow Scenario Relative
to Actual Trip. . •
. • . • . . .
• • • • . • .

• . 123

6-11. Rating of the 3,000 cfs with Fluctuations Scenario Relative
to the 3,000 cfs Constant Flow Scenario •
. • . • .
. . 124
6-12. Rating of the 10,000 cfs Constant Flow Scenario Relative
to Actual Trip. • • . • • • .
. . . . . . •

. . . 125

6-13. Rating of the 10,000 cfs with Fluctuations Scenario
Relative to the 10,000 cfs Constant Flow Scenario .

. . . 126

6-14. Rating of the 25,000 cfs Constant Flow Scenario Relative
to Actual Trip. • • • • • • • • • • • • • • • • • • . . . • • . 127

vii

I

LIST OF TABLES (Continued)

6-15. Rating of the 25,000 cfs with Fluctuations Scenario
Relative to the 25,000 cfs Constant Flow Scenario . .

. . 128

6-16. Rating of the 40,000 cfs Constant Flow Scenario Relative
to Actual Trip .
. ..

.........

7-1.

...

129

.

137

Daily Flow Level Experienced by Respondents to the
Glen Canyon Rafter Attribute Survey.
..

..

.

7-2.

Glen Canyon Rafters' Attribute Survey Response Rate.

7-3.

Attributes that Contribute Y~st to an Excellent or Perfect
One-Day Raft Trip. • . • • • • • • . • • . • • • •
• 139

7-4.

Attributes that Contribute Most to a Poor One-Day Raft
Trip . . • .

. 141

7-5.

Respondents' Knowledge of the Expected Water Level

. 142

7-6.

Respondents' Evaluation of the Water Level During Their
Raft Trip. . . . • . • • •
. • . • •

7-7.

Daily Flow Levels Experienced by Respondents to the
Glen Canyon Rafter Contingent-Valuation Survey

7-8.

Glen Canyon Rafters CV Survey Response Rate. •

8-1.

Grand Canyon White-Water Boating and Glen Canyon Angling
1985 Use Rates (Trips) . . • . . .

138

142
• 143
. • 144
. 154

8-2.

Alternative Annual Flow Regimes.

156

8-3.

Evaluation of Flow Regimes (Annual Flow = 8.25 ~~llion
Acre-Feet) . . . .
. . . • • . . . . . . . . .

159

8-4.

Data for Simplified 1984 Flow Regime

163

8-5.

Data for Simplified 1985 Flow Regime

164

8-6.

Evaluation of 1984 and 1985 Flow Regimes

165

.-,
8 -1.

The Unconstrained Optimal Recreation Flew Regime
(All Flows Constant) . • . . . . • . . . . . . . . . . . . . 167

viii

•

LIST OF FIGURF.S

•

-

1-1

Diagram of the Relationships Among Study Components.

7

2-1

Theoretical Flow Value Curve for an Individual • •

17

3-1

Question Asking White-Water Boaters About the Most
Important Reason for Taking Trip • . • • . • •

• 32

3-2

White-Water Boater Positive Attribute Question

3-3

Day-Use Rafter Negative Attribute Question . • .

• 32

3-4

Glen Canyon Angler Positive Attribute Question .

. 34

3-5

Glen Canyon Angler Expenditure and Actual Trip CV Question • 39

3-6

White-Water Boater Constant Flow Scenario (5,000 cfs) and
Associated Valuation Question . .

3-7

Glen Canyon Angler Environmental Impact Scenario
(Larger Fish) and Associated Valuation Question . . • . • . . 41

3-8

Relationship Between Tuber Sur pl us Values and
Answering No to a Dichotomous-Choice CV Question . . . . •• 46

3-9

Cumulative Distribution Function for a
Continuous Distribution of Surplus Values.

4-1

Guides' Constant Flow Level Preference Ratings

4-2

-

32

. 40

• • 47

• 53

Proportion of Respondents Who Reported Stopping to Scout
Rapids at Constant Flows . .
. . . . . • • . . • . 57

t~jor

4-3
4-4

5-1

..

Proportion of Respondents Who Have Passengers Walk Around
Rapids at Constant Flows • . . . . • • • •

58

Proportion of Respondents Who Run Motor or Row Nore or Less
Than Usual to Compensate for the River Current at Constant
Flows. . • •

59

Relationship Between Surplus Values and Flew Levels for
Respondents' Actual Trip ($ Per Trip). . • .

• • 93

5-2

Commercial Boater Surplus Values for Constant Flow
Scenarios and Actual Trip ($ Per Trip) • • • • . •

5-3

Private Boater Surplus Values for Constant Flow Scenarios
and Actual Trip ($ Per Trip) . • . • • • • • • • • • • • . • 106
ix

• 105

LIST OF FIGURES (Continued)
5-4

Commercial Passenger and Private Boater Surplus Values for
Fluctuating Flow Scenarios ($ Per Trip). . . . • • .
. 107

6-1

Glen Canyon Angler Flow Value Functions for Constant and
Fluctuating Flow Levels ($ Per Trip) . . . .

. 133
. • 149

8-1

White-Water Boater Flow Value Functions.

8-2

Angler Flow Value Functions. .

8-3

Fl ow Value Functions (Constant Flows).

8-4

Flow Value Functions (Fluctuating Flows)

• 152

8-5

Calculation of Annual Recreation Benefits for Alternative
Fl ow Regimes .

. 158

• • • 150

x

151

•

..

CHAPTER 1
IHTRODUCl'IOR
Background
The research reported here is one part of the Glen Canyon
Environmental Studies (GCES). The GCES are a joint research effort
sponsored by the Bureau of Reclamation and the National Park Service
focusing on the Colorado River below Glen Canyon Dam and above Lake
Mead. When Glen Canyon Dam was completed and Lake Powell began
filling in 1963, basic environmental parameters such as water
temperature, silt loading, and annual stream flow patterns in the
Colorado River below the dam were changed substantially. The result
is a much modified aquatic and riparian environment along the nearly
300 miles of the Colorado River flowing through Glen Canyon National
Recreation-Area and Grand Canyon National Park. The GCES comprise a
broad interdisciplinary project designed to determine the effects of
flow release patterns from Glen Canyon Dam on that environment.
A wide range of flows can be released at the dam. When all eight
turbines of the Glen Canyon Power Plant are in full operation, 33, 500
cubic feet per second (cfs) of water is being released~ In addition,
there are four 96-inch diameter pipes (bypass tubes) that allow water
to -bypass the power plant. They have a combined capacity of 17,000
cfs. Under extreme conditions, when the reservoir (Lake Powell) is
full, spillways capable of releasing an additional 278,000 cfs are
al so available.
At the other extreme, the dam is capable of releasing as little as
1,000 cfs, although minimal flows of 3,000 cfs in the months of April
through September are generally maintained for environmental
considerations and the benefit qf downstream recreation. Additional
constraints are placed on dam releases by.legal and institutional
requirements, and by releases from upstream reservoirs. Still, while
operating within these constraints, there is broad latitude in
choosing how much water will be released at any point in time and in
choosing the pattern for water releases on an annual, monthly, or
even daily basis. Different flow release patterns will have
different effects on the downstream environment and the GCES are
seeking to understand and document these effects.

..

The research reported here focused specifically on the relationships
between stream flows and river based recreation in Grand Canyon
National Park and Glen Canyon National Recreation· Area, below Glen
Canyon Dam and above Lake Mead. Although the casual visitor viewing
the Grand Canyon from the South Rim may enjoy seeing the river, the
quality of her/his experience is not likely to be heavily dependent
on how much water is released from the dam. There are three groups

2

of recreationists, however, that may be directly and substantially
affected by stream flows. The first group will be referred to as
"white-water boaters." These are people using a variety of rafts and
boats to take a white-water trip on the Colorado River between Lee's
Ferry and the Lake Mead. Such trips involve negotiation of the
famous white-water rapids of the Grand Canyon. The second group is
composed of "Glen Canyon anglers" who typically fish the Colorado
River between the.dam and Lee's Ferry. The third group of
recreationists are "day-use rafters" in Glen Canyon. These people
take one-day trips through Glen Canyon (above Lee's Ferry) on large,
motor-driven rafts. The volume of water in the river is a
potentially important parameter in the recreational environments of
these three groups and the effect of varying flow release patterns
from Glen Canyon Dam on these three groups is the focus for the
research reported here.

•

Recreation and Stream Flows
White-water boaters typically depart from Lee's Ferry. They use a
variety of craft with two types of distinctions being relevant for
our work. First, boats are distinguished by whether or not a motor
is used. "Motorized trips" utilize relatively large rafts, often
measuring 37 feet in length and carrying as many as 25 passengers.
Nonmotorized boats are generally smaller, and include oar rafts,
paddle rafts, dories, kayaks, and occasionally other craft as well.
The second major distinction is between commercial and private
trips. Commercial trips are organized by 21 commercial rafting
companies which, for a fee, supply guides, boats, food, and much of
the other equipment needed by passengers. Private trips are
organized by groups of individual:=: who provide their own equipment
and supplies, usually sharing expenses. In 1985, 11,374 commercial
passengers and 2,368 private boaters traveled down the Colorado River
through Grand Canyon Nati_onal Park.
Lee's Ferry plays a major role in Glen Canyon fishing in that it is
the only major access point on the river for anglers. The National
Park Service provides a boat launching ramp, dock, parking area, rest
rooms, and a campground at this river location. Most of the fishing
takes place in boats on the 15 miles of river upstream from Lee's
Ferry. A smaller number of anglers fish from the shore near Lee's
Ferry. Some fishing also occurs as an incidental activity on
white-water boating trips, but there are not currently enough anglers
fishing on these trips to warrant specific attention in- this study.
For purposes of this study, the effect of flow levels on incidental
fishing on white-water trips will not be evaluated.
The current Glen Canyon fishery is a product of the Glen Canyon Dam.
The cold water released from Lake Powell has combined with other
ideal environmental conditions to create a very productive trout

"

3
t

fishery with rainbow trout as the dominant species. Within a few
years after completion of Glen Canyon Dam, anglers began enjoying an
outstanding fishing opportunity with many fish of trophy quality
being taken during the late l960's and the 1970's. More recently,
the number of trophy-sized fish taken by anglers has declined
substantially, most likely as a result of heavy fishing pressure. As
many as 52,000 angler days were recorded in the Glen Canyon fishery
during 1983, but participation has been -highly variable from one year
to the next. In response to this trend in catch rates, the Arizona
Game and Fish Department, as the agency that manages the fishery, has
enacted increasingly stringent regulations to decrease fishing
pressure and the number of fish taken by anglers.
Day-use rafters are people taking commercial raft trips through Glen
Canyon above.Lee's Ferry. This recreational experience is also a
by-product of Glen Canyon Dam. The trips typically originate at the
dam and participants spend the day floating/motoring down to Lee's
Ferry. However, at relatively high flow levels when bypass tubes are
in operation, trips depart from Lee's Ferry to avoid the turbulence
below the dam, and motor upstream before floating down. Thus, the
most obvious effect of flow release patterns on day-use rafters
occurs when the bypass tubes are in operation. Participation in Glen
Canyon raft trips has been growing in recent yea.rs with about 8,500
individuals taking a trip in 1985.
Goals and Approaches of the Study
The overall goal for the portion of the GCES reported here is to
evaluate the impacts of alternative flow release patterns from Glen
Canyon Dam on white-water boating, fishing, and day-use rafting on
the Colorado River below the dam. To achieve this goal, four
interrelated research procedures were implemented.

•

The first stage in the research involved "attribute surveys" of
recreationists. The effects of varying flow release patterns are
transmitted to recreationists largely through changes in the quality
of the experiences. It was necessary to know more about the
characteristics or attributes of the recreational experience that
influence recreational quality and which of these attributes are
influenced by stream flow. Consider white-water boating for
example. Even casual observation of white-water boating would
indicate that running rapids is an important attribute of a trip down
the Colorado River. However, the relationships between recreational
quality, rapids, and flow rates had not previously been documented .
As flow rates increase, the hydraulics and waves at some rapids are
intensified, while other rapids are washed out. How do boaters feel
about this trade-off? At very high flows, safety concerns may become
important, and passengers may have to walk around one or more of the
major rapids. Also, how important are rapids compared to other

4

attributes of the white-water boating experience? An understanding
of the implications of alternative flow release patterns on the
recreational experience requires a thorough understanding of the
recreational activities, including the role of flow-sensitive
attributes. This understanding was gained by conducting attribute
surveys of individuals from all three recreation groups.

•

Second, our research drew on the.knowledge of river guides. The
people who serve as guides for commercial and private white-water
trips, for anglers, and for day-use raft trips constitute an
important group of experts on the relationship between flows and
recreation. River guides have first-hand knowledge of the
recreationists, the recreational experience, and the river itself.
Some have years or even decades of experience, and a comprehensive
study_ could not overlook such a rich source of knowledge. We
conducted a formal mail survey of white-water boating guides and,
because of the small number of people involved, made informal contact
with fishing guides and day-use rafting company representatives.
Third, our research depended heavily on the help of resource managers
and other researchers. At each step in the study, a working group
composed of EBRS staff and GCES study managers from the Bureau of
Reclamation and National Park Service reviewed progress and planned
succeeding steps. The evolution of the study was shaped in many ways
by these contacts. ~:embers of the other GCES research teams also
contributed valuable input, particularly those from the Arizona Game
and Fish Department, the U.S. Geological Survey, and the University
of Arizona.
The fourth, and final, step in the research involved
"contingent-valuation surveys" of re creationists to quantify their
preferences for a variety of flow release patterns in monetary
terms. To do this, we applied a technique called "contingent
valuation." Contingent valuation will be discussed often in this
report and it will be convenient to use "CV" as an abbreviation.
CV is a technique developed by econo~ists to measure monetary values
for things that are not typically traded in markets. In the current
case, for example, access to the recreational resources of the
Colorado River is not governed by a market in the traditional sense
of the word. Although recreationists may buy fishing equi~ment in a
sporting goods store or pay a white-water boating firm to be a
passenger on one of its trips, the recreational opportunities
themselves are provided by the public sector and market prices for
access are not available to serve as guides to their economic
values. A CV study is accomplished by using surveys to ask
recreationists what value they would place on access to a
recreational resources if a market or other means of payment did
exist.

•

5
I

In the applications of CV reported here, we first asked
recreationists about the actual costs they incurred to make their
Colorado River trip. It was relatively simple to define the specific
trip to be valued for white-water boaters and day-use rafters, since
it would be very rare for an individual to take more than one such .
trip in a given year. Thus, the first step in asking the CV question
for white-water boaters and day-use rafters was to ask them in a mail
survey about actual expenses for their 1985 trip. Anglers, however,
often take more than one trip per year to fish in Glen Canyon.
Consequently, anglers were interviewed at Lee's Ferry on selected
days and then asked in a subsequent mail survey about this particular
trip.
For all three groups, the next step was to ask respondents whether
they would still take such a trip i f their expenses had increased by
a specified amount. This is the CV question for their actual trip,
and their responses were analyzed to provide an estimate of how much
the trip was worth to them over and above its costs. Finally,
scenarios of recreational experience describing changes in flow
levels or other flow related parameters were presented to respondents
accompanied by a similar CV question. The information in the
scenarios was based on the results of the attribute surveys and
contacts with guides, resource managers, and other GCES scienti~ts.
The outcome of applying CV, then, was a dollar value per trip for the
actual trip, and a dollar value for each of the alternative trip
scenarios. These values were estimated for all three recreation
groups.
The ultimate goal of the study was to assess the impact of
alternative annual flow release patterns or "annual flew regimes" for
Glen Canyon Dam on recreationists in the aggregate. This goal was
achieved by modeling the relationships between flow levels and
monetary values for each of the three groups of recreationists.
Values per trip, based on the results of the CV surveys, gave
estimates of how recreationists would value individual trips under
actual conditions or conditions described in the scenarios. The
evaluation of specific flow release patterns required combining the
values of white-water poaters, anglers, and day-use rafters with the
number of trips taken by each group in 1985 to produce an estimate of
aggregate recreational benefits. A computer model was developed to
calculate aggregate recreational benefits over a wide range of
possible flow release patterns from the dam.
Some readers may find the use of monetary valuation in a recreational
context confusing and perhaps disturbing. A great deal more will be
said about CV and the potential usefulness (and pitfalls) of economic
values for managing natural resources in the next chapter. Our
objective in the current chapter is simply to provide an overview of
our approach to evaluating the impacts of flow levels on recreation
activities and to describe how this research was accomplished.

6

Study Plan
The components of the study are sunmarized in Figure 1-1. This flow
diagram shows how the various parts of the study were integrated to
accomplish the overall goal of evaluating alternative Glen Canyon Dam
releases on Colorado River recreation below the dam and above Lake
Mead. The contacts with guides, resource managers, and other GCES
researchers were inputs for both the attribute surveys and the CV
surveys. These inputs aided the design of these surveys, as well as
the interpretation of the survey results. Surveys of recreationists
were used to collect the necessary data to identify the attributes of
these recreational exp·eriences that are affected by flow levels
(attribute surveys). The flow-sensitive attributes identified were
used to guide the development of the CV surveys used to estimate the
effects of different flow release patterns, as measured in dollar
terms, on the recreational experiences. The values derived from the
CV surveys formed the basis for the Flow Valuation Model. For
specified annual flow regimes within the flow ranges covered by the
model, aggregate recreational values for all white-water boaters,
anglers, and day-use rafters were computed. Aggregate recreational
impacts were calculated for five flow regimes which representatives
of the Bureau of Reclamation specified as being representative of
possible release patterns from q1en Canyon Dam.
The remainder of this report is divided into 8 chapters (Chapters
2-9). Chapter 2 develops the conceptual model for estimating the
relationships between river flows and economic values, and Chapter- 3
describes the research procedures used to collect the necessary data
to identify these relationships. Conclusions regarding the effects
of Glen Canyon Dam releases on white-water boating guides,
white-water boaters, Glen Canyon anglers and day-use rafters are
presented in Chapters 4-7, respectively. Using the results reported
in these chapters, five prototype flow release patterns from Glen
Canyon Dam, which were given to all GCES teams for analysis, are
evaluated and contrasted in Chapter 8 to identify the trade-offs
affecting recreationists from varying operation of the dam. Chapter
9 presents the conclusions from our analysis of the effects of flow
release patterns on the recreational experiences, and discusses the
implications of these results for the management of flow release
patterns from Glen Canyon Dam.
Summarv
The components of this study were specifically designed to identify
the effects of a variety of flow release patterns from Glen Canyon
Dam on Grand Canyon white-water boaters, Glen Canyon anglers and Glen
Canyon day-use rafters. The results of these analyses will help the
U.S. Bureau of Reclamation and the National Park Service evaluate the
key trade-offs between flow levels and users enjoyment of these
recreational activities. The evaluation of such trade-offs, however,

•

-----------,----------------------------------------------

7

Figure

1-1

Diagram of the Relationships
Among Study Components

Contact with Bureau of
Reclamation and National
Park Service Personnel,
and members of other
GCES Research Teams

Attribute Surveys of
White Water Boaters,
Glen Canyon Anglers,
and ~ay-Use Rafters

Contingent-Valuation Surveys
of White Water Boaters, Glen
Canyon Anglers and
Day-Use Rafters

Annual flow
regimes spec_ified
for evaluation by
the Bureau of
Reclamation

Flow Valuation Modeling

Calculated impacts of flow
regimes, in monetary units.
on White Water Boaters,
Glen Canyon Anglers, and
Day-Use Rafters

Formal and informal
contacts with
Professional River
Guides

8

required a substantial data collection effort regarding the
preferences of each user group and the knowledge of professional
guides. Thus, the contributions of this study will extend beyond the
findings presented in this report. Even though the text of this
report only addresses the analyses of trade-offs between flow levels
and user preferences, the data sets that were used in these analy~is
may be useful in analyzing a variety of other types of management
iss-ues.

;

9

••
CHAPTER 2
RECREATION V.ALUF.S .ABD PUBLIC DECISIONS
Introduction
The use of dollars to quantify the effects of Glen Canyon Dam
releases on downstream recreation raises several issues that will be
addressed in this chapter. First, why try to value such effects in
dollars at all? We suggest in the first section that dollars provide
a useful yardstick for measuring the public's preferences for
resource management alternatives. Furthermore, the use of dollars
has some distinct_ advantages when compared to other approaches for
measuring preferences.
The second issue involves the definition of value. Value is a term
that can take on many meanings, even within economics. A
state-of-the-art application of economics to Colorado River
recreation must apply standard, widely accepted definitions of
value. For this purpose, we drew on standard concepts from
benefit-cost analysis. In the middle sections of this chapter, the
concepts of total value and surplus value, as defined in benefit-cost
analysis, Will be explained. These concepts will then be adapted to
the problem of valuing the effects of Glen Canyon Dam releases on
Colorado River recreation.
Toward the end of the chapter, two other issues will be addressed.
As noted in Chapter 1, this study will employ contingent-valuation
(CV). An alternative technique, the travel-cost method, is also used
frequently in recreation valuation studies. A section is devoted to
explaining why the travel-cost method was not used here. Then, given
that the success of the study depends in part on the validity of CV
results, it will be necessary to ask whether CV can be expected to
provide valid dollar values. Selected studies that have investigated
the validity of CV will be summarized.
Why Dollar Values?
Dollar values are used in everyday life to communicate relative
importance. When we read, for example, that videotape recorders have
become a billion dollar a year industry, this tells us something
about how important these devices have become to consumers, and
hence, to the industry and employees who produce them. In a sense,
such dollar values convey something about the priority that our
society is placing on them. Similarly, dollar values can be used to
evaluate social priorities in natural resource management.

10

By definition, resource management involves choices among
alternatives. This is certainly true when choosing appropriate flow
release patterns from Glen Canyon Dam. As discussed in Chapter 1, a
wide variety of daily and annual flow release patterns are
technically feasible within the constraints set by the ~sign
features of the dam, the inflows of water from upstream, and legal
and administrative requirements. Each potential flow release
alternative has its own implications for power revenues; the
well-being of terrestrial and aqoatic ecosystems downstream; various
recreational users of Lake Powell, Glen Canyon, and the Grand Canyon;
water levels at Lake Mead; and the legal requirements for operation
of the dam. Decisions about release patterns, thus, involve a
complex balancing of many social priorities.
The priorities that government officials bring to bear on such
questions are supposed to reflect the wants and needs of the
citizenry. But, how are these wants and needs to be assessed?
Traditionally, we have depended partly on elected officials and their
appointees, operating within the constitutional system, to implement
our preferences. Presumably, elected officials who too often make
decisions contrary to public preferences will not be reelected. The
wise public official has traditionally kept abreast of public·
preferences through contacts with individual citizens, especially key
leaders among his or her constituency, and citizens' groups.
Frustrations among some segments of the public with the apparent
unresponsiveness of government officials (both the elected and
appointed) led to the installation of formal public participation
mechanisms for some decision processes. Public hearings and other
forms of public participation were designed to help decision makers
gauge what the public really wanted, and to give the citizenry an
official forum for expressing their preferences on specific issues.
In a similar vein, there has been a growing interest in using methods
from the applied social sciences, including economics, to assess
public wants, preferences, and priorities in a systematic and
defensible way, and in more depth than is feasible through
traditional public participation techniques. Economic methods have
focused primarily on the measurement of public preferences in
monetary terms.
Dollar valuation as a measure of public preferences, then, may be
thought of as a yardstick for measuring the importance or intensity
of preferences among members of the public for resource management
alternatives. Dollar measures have a couple of desirable features
when employed in this way. First, dollars are a commonly used and
easily understood unit of measure. To say, for example, that Grand
Canyon white-water boaters prefer flow Alternative A to flow
Alternative B is certainly relevant information, but more information
is communicated by saying that white-water boaters gain $100,000 more

..

11

,,

per year in benefits under Alternative A than they do under
Alternative B. Dollars take on extra significance in an absolute
sense because we measure the worth of so many things in monetary
units. This is true for not only mundane things like bus rides or a
can of beans, but for objects of art, classical music recordings,
tickets to the Super Bowl, and vacations in exotic places.
The second advantage of using dollars to measure preferences is that
they are easily added, subtracted, and compared. Other more
qualitative measures of preferences are not easily reduced to a
common denominator. Resource managers must inevitably work with
aggregates of people. When a given dam release alternative affects
both white-water boaters and Glen Canyon anglers, the dollar benefits
accruing to each group can be added to measure the aggregate impact
of that alternative. Perhaps more importantly, when user groups
disagree, dollar values provide a basis for comparison. We can
determine which alternative produces the largest total dollar
benefits for each group separately or for all groups combined.
Methods for measuring these dollar benefits were developed to give
resource managers and public officials an improved understanding of
public preferences regarding resource management alternatives. To
achieve this goal, it was necessary to define carefully what these
dollar values mean so that they could be applied to a broad array of
publicly provided goods, services and amenities in a uniform and
consistent manner.
Value Definitions
It is important to recognize that the concepts of monetary valuation

that are applied to Colorado River recreation in this study are not
new. The general concepts defining monetary values are not only
applicable to recreation, but to all sorts of goods and services.
The use of these monetary values in resource management decisions in
the United States began gaining momentum when-the Flood Control Act
of 1936 declared that Army Corps of Engineers water resource
development projects were to be considered economically feasible only
if "the benefits, to whom so ever they may accrue, are in excess of
the estimated costs." This statement led to the development of a
procedure for evaluating the feasibility of public projects that is
called "benefit-cost analysis."
The concepts and tools of benefit-cost analysis have undergone
continuous refinements over the past 50 years. Whenever an agency
like the Bureau of Reclamation develops a proposal for a water
project, a benefit-cost analysis must be performed. Procedures for
estimating the so-called "National Economic Develoµnent (NED)
Benefits and Costs" of such projects are presented in detail in the
Economic and Environmental Principles and Guidelines for Water and

12

Related Land Resources Implementation Studies (U.S. Water Resources
Council, 1983). The Principles and Guidelines, as this document is
often called, are followed by all federal water resource agencies
including the Corps of Engineers, the Bureau of Reclamation, the
Tennessee Valley Authority, and the Soil Conservation Service. They
are also widely recognized as an authoritative source on benefit-cost _
analysis by economists in m~ny other agencies and outside of
government as well. Valuation procedures for determining the
benefits to society from municipal and industrial water supplies,
agricultural products, flood control, hydroelectric power,
navigation, commercial fishing, and recreation are covered.
Our purpose in introducing the Principles and Guidelines is to assure
that the recreation benefits of alternative flows from Glen Canyon
Dam are measured using standard,. widely accepted definitions and
procedures. We do not intend to imply that a full benefit-cost
analysis is to be performed in this study. The goal here is the much
more modest one of measuring recreation benefits alone.
The Principles and. Guidelines provide a conceptual foundation for the
evaluation of NED benefits based on the concept of willingness to
..2§Y· This concept is applicable to any situation where a public
project increases the output of goods and services. It can also be
applied where quality is improved and where costs of producing
existing output are reduced. Value is defined as the maximtml amount
a consumer would be willing to pay for the outputs or other effects
of a project rather than do without them. To further understand the
willingness-to-pay concept, let us define the total value of the
outputs and other effects of a project as the maximum that all
members cf society combined are willing to pay to obtain them -- or
to avoid them i f negat.:.ve outputs (e.g., pollution) would occur.
Total value can be divided into two part~:: market value and surplus
value. Consider a project that only produces one output. The total
value of that output equals the maximum amount consumers are willing
and able to pay for it rather than do without it. Market value
equals the market price of a unit of output multiplied by the units
of output produced by the project. If, for example, an irrigation
project causes an increase in the nation's supply of carrots, market
value is equal to the price of carrots multiplied by the increased
quantity produced. But some consumers may be willing to pay more for
these additional carrots than they would actually have to pay in the
market. This extra amount over and above what is actually spent is
surplus value. The total value of the carrots equals their market
value plus their surplus value.
In summary, total value can be broken down into market value (the
amount the individual actually pays fer the projects' output) and
surplus value (the additional amount they would pay if they had to to
do so). In economic terminology, surplus value is called "consumer

,,

13

•
surplus" since it accrues to individual consumers. Total value of
the project's output is the sum of the maximum willingness to pay for
all individuals affected by the project, including both market value
and surplus value.
In the carrot example, the emphasis was on quantity, while the
present study emphasizes quality. That is, the water level in the
Colorado River is one of the potential determinants of quality for
white-water boating, fishing, and day-use rafting. The goal is to
measure in dollars people's subjective evaluations of qualitative
changes in their recreational experiences. Still, the principles are
the same whether the changes to be evaluated are quantitative or
qualitative. To the extent that changes in flows enhance
recreational quality, people would be willing to pay more per trip,
i.e., they have larger surplus values. Similarly, if a change in
flows reduces quality, recreationists would be willing to pay less
per tr.ip, i.e., they have lower surplus values.
It is these changes in value "at the margin," to use the economic
term, that are of interest here. Referring again to the carrot
example, notice that the goal was not to measure the value of all the
carrots produced in a given region or the nation as a whole, but
rather the carrots produced, at the margin, by the individual
.irrigation project. Likewise, our goal here is not to measure the
total value of all Grand Canyon recreation. Rather, we will only
consider the values of certain specific user groups for certain
specific effects, at the margin.
Economic terms are easily misunderstood and total value is no
exception. Thus, just to avoid misunderstanding, let us explicitly
state that there is no intention in this study to measure the total
value of Grand Canyon National Park or Glen Canyon National
Recreation Area. To do so would require attention to many other user
groups and to nonuse values as well. Nonuse values, such as option
and existence values, may be important components of the total values
of important environmental assets such as those under study here (see
Krutilla and Fisher, 1975; Smith, 1983; Randall and Stoll, 1983; and
Boyle and Bishop, forthcoming, for further discussion of option and
existence values), but are beyond the scope of the current study.
Here, we will examine only the total values associated with specific
qualitative effects of Colorado River flows on specific user groups:
the white-water boaters, anglers, and day-use rafters.
Two other general definitions will be useful. "Costs" equal the
value of resources required to produce project outputs. They may
include private costs (e.g., farm inputs to grow.the carrots) or
public resources (e.g., costs of a dam). "Benefits" associated with
project output are defined as total value of the outputs minus the
costs directly associated with their production and distribution.

14

..
concepts are readily applied to recreation. In defining the
conceptual basis for recreation valuation, the Principles and
Guidelines state (p. 67), "Benefits arising from recreation
opportunities created by a project are measured in terms of
willingness to pay."
The~e

Consider Glen Canyon fishing as an example. The total value of
access to the Glen Canyon fishery for any given year is the maximum
amount that all anglers combined would be willing to pay rather than
give up access for that year. The concept of market value must be
modified, however. Fishing opportunities at Lee's Ferry do not
directly pass through a market, but are provided subject only to
paying for a proper fishing license. At the same time, any angler
knows that fishing does cost money, not only for the license, but
also for use of a motor vehicle, boat costs, equipment, meals,
lodging or camping, etc. In recreation valuation, "expenditures"
take the place of market value as the first component of total
value. Surplus value (the second component) then equals total value
minus expenditures.
The importance of including surplus values as well as expenditures in
total values can be illustrated using a Glen Canyon fishing example.
Suppose an angler from Page, Arizona, has a total value per fishing
trip of $50, but spends only $10 on a Glen Canyon fishing trip,
leaving $40 in surplus value. If, for example, a ticket booth were
set up at the dock at Lee's Ferry, this angler would continue to come
fishing so long as tickets were less than $40 per trip. At access
fees over $40, he or she would stop coming. If we are going -to
measure the total value of a trip to this person, counting only the
expenditure of $10 would undervalue what it is really worth. As a
measure of the priority that the Page angler places on fishing
access, we must count not. only expenditures, but also surplus value.
Application of the Principles and Guidelines procedures to
white-water boating and ~Hen Canyon day-use rafting also requires the
measurement of total values which ·include surplus values.
The arguments for including surplus values in total recreation values
are relatively straightforward. If dollar values are going to be
used to measure people's preferences, then it seems natural enough to
measure their total values and not just their expenditures. The next
step in the argument is not so easily grasped at first: In
estimating the NED benefits from recreation, only surplus value are
usually counted. Expenditures are usually not counted as part of
national economic developnent benefits. To make the problem even
more difficult, when the viewpoint changes from national economic
developnent to local economic impacts, expenditures play a key role
and surplus values are ignored. It will be helpful to deal with
these issues immediately.

f

15

Benefits Equal Surplus Values, Not Expenditures
Major recreational resources like those found in Grand Canyon
National Park and Glen Canyon National Recreation Area cause money to
flow into local economies as recreationists from outside the area buy
locally sold goods and services. Such expenditures are the "first
round" of local economic effects. Later "rounds" occur as local
businesses and households spend and re~spend this money. Substantial
household income and local employment may be generated in the
process. For example, a Glen Canyon angler from Denver spending
money in a Page restaurant might be the first round. Part of the
money might go to a waitress as wages (the second round). The
waitress in turn might spend the money at the local supermarket
(third round). If the supermarket manager spends the dollar to order
more grocer~es from Phoenix, then the process ends so far as Page's
economy is concerned. The usual way of measuring the effects or such
expenditure flows is through the use of "multi pliers." Local
economic impact analysis attempts to measure the total impacts of
expenditures by outsiders, such as recreationists, on local business
activity, household income, and employment within the local economy.
Measuring local economic impacts can have important policy
implications. Areas like northern Arizona have· limited economic
bases. Providing recreational opportunities that bring people into
the area can make a substantial difference to local incomes and
employment. Bolstering economic activity in such predominantly rura.l
areas has long been a public policy objective.
Interestingly, surplus values have no role to play in analyzing local
economic impacts. Because surplus values involve money that is not
actually spent, it has no effect on local income and employment. For
local impact analysis, only expenditures are relevant.
Turning to the national perspective, expenditures become less
relevant. In fact, expenditures are normally costs rather than
benefits. Returning to our hypothetical Page angler who has a total
value per trip of $50 and s~ends $10 per trip, it costs society $10
worth of gas, fishing lures, etc. to provide a trip to this person.
He or she gets $50 in total value, but incurs costs of $10, leaving
$40 in surplus value.
At first, leaving expenditures out of the analysis of NED benefits
from recreation seems to contradict what has already been said. If
expenditures contribute to local income and employment, shouldn't
these effects be considered benefits? Special exceptions exist, but
in general, they are not included. Unless there are special
circumstances to justify doing otherwise, local effects are assumed
to wash out nationally. To take the present case, suppose that the
recreational opportunities in Glen Canyon and the Grand Canyon did
not exist. Recreationists would spend the money elsewhere and create

16

local economic impacts there. The fact that white-water boaters,
anglers, and day-use rafters are having positive local impacts in
northern Arizona means they are not having local impacts elsewhere.
Unless there is clear evidence to the contrary, local economic
effects are assumed to balance out to zero nationally. Thus, local
impacts are normally not counted in national benefits. The need to
treat national economic develoµnent and local economic impacts
differently is clearly spelled out in the Principles and Guidelines,
Sections 1.7.2 and 1.7.4.
Adaptation of Value Concepts to Colorado River Recreation
The goal of the economic portion of our study was to apply
bene~it-cost concepts to Colorado River recreation.
It should be
explicitly stated again, however, that this study did not attempt to
do a full benefit-cost analysis of alternative Glen Canyon Dam flow
release patterns. To have done so would have required a much larger
project incorporating power benefits, potential impacts on Lake
Powell recreation, and possibly many other implications of flow
release~ from the dam.
Rather, the present project utilizes standard
benefit-cost concepts to quantify one part of the benefits, the
benefits to downstream recreation of varying river flows.
Let us begin with an individual recreationist. The relationship
between stream flows and an individual's surplus values per tr·ip
might look like the curve postulated in Figure 2-1. On the
horizontal axis, the flow rate in cubic feet per second is portrayed,
while the individual's surplus value per trip can be read off the
vertical axis. In this hypothetical example, at very low flows, such
as Point A, surplus value for this recreationist is relatively low.
Increasing the stream flow increases surplus value to point B where
conditions are ideal from the perspective of this recreationist.
Further increases in stream flow cause surplus value to decline
toward point C.
Suppose that Glen Canyon fishing is being evaluated. Point A might
represent a flow rate that is so low that it is difficult to maneuver
boats up the- river from Lee's Ferry without damage to the boat and/or
its motor. Three Mile Bar, which lies upstream from Lee's Ferry, is
difficult or impossible to pass when flows are quite low.
Consequently, low flow levels tend to concentrate anglers in the
three miles of the river above Lee's Ferry and may cause crowding.
Low flows may contribute to reduced fishing quality in other ways as
well. At point B, on the other hand, flows are at a level which
creates ideal fishing conditions for this individual. At higher
flows, such as point C, the current becomes more swift, the fish may
be harder to find, and other effects may also reduce fishing quality
and hence surplus value per trip. We will refer to such curves
depicting the relationship between flows and surplus values as flow
value curves.

17

Figure

2-1

Theoretical Flow Value
Curve for an lndlvldual

Individual's
Surplus Value
Per Trip
($)

AveraQe Flow Level in cfs

..

18

Of course, the specific shape of the flow value curve for any given
recreationist would depend on the recreational activity and the
characteristics and preferences of the individual. ~~ny shapes are
theoretically possible. For example, if there is a white-water
boater who likes big rapids and, over a very wide range of- flows,
feels that the bigger the rapids the better the trip, then rather
than turning down after point B, the curve might rise throughout.
Figure 2-1 simply represents one of many possible shapes for the flow
value function.
The ultimate goal of the CV portion of the analysis is to estimate
the effects of alternative flow release patterns on surplus values
for both constant flows and daily fluctuations in flows. Two methods
of estimating such surplus values are well established: the
travel-cost method and the contingent-valuation method (CV). Since
CV was used in the study, it is worthwhile to briefly consider why
the travel-cost method was not used.
Why Not the Travel-Cost Method?
The travel-cost method treats travel costs and other expenditures
incurred to participate in a r~creational activity as a measure of
the market price for a recreational trip. The responses of people to
these costs (as indicated by whether or not they make a trip) are
used to estimate a relationship between total recreation expenditures
per person per trip and the number of trips an individual chooses to
make. Economists refer to this relationship as a demand function.
This demand function is then used to estimate a surplus value per
trip.
The travel-cost method can sometimes be used to estimate the total
surplus value associated with recreation at a specific site. In
fact, such a model has been developed for Glen Canyon fishing by
Richards (1985). However, it is not clear that such a model could be
successfully applied to white-water boating and day-use rafting. The
vast majority of these recreationists take at most one trip per year,
and many take nnly one trip in a lifetime. This would greatly impede
application of the travel-cost method because the number of trips
taken by any one individual in a given year does not vary. A
so-called zonal model would have to be used and it is questionable
whether the assumptions necessary for a zonal travel-cost model to
work well would hold here. The problem is further complicated for
white-water boaters by the Park Service ceiling on the total nunber
of boaters that can launch each day and for the year as a whole.
Even if these concerns could be overcome, say by using one of the new
probabilistic models like that proposed by McCollum (1986), the
measurement of changes in surplus values due to qualitative changes

19

like those under study here would be difficult using a travel-cost
model. Such qualitative changes would be reflected in travel-cost
analyses only to the extent that recreationists make more or fewer
trips as a response to changes in flow levels. A white-water boating
trip, for example, is planned months or even years in advance, and it
is unlikely that people would change such plans based on the flows
that happened to be in effect when the time for departure arrives,
except under extreme conditions. Also,- the National Park Service
limits on the number and timing of white-water boating trips in the
Grand Canyon further constrains the ability of white-water boaters to
adapt their plans to flow levels. Although we argue that flow levels
do not affect the number of trips taken by white-water boaters and
day-use rafters, they may still enjoy one flow level more or less
than another. If so, however, such preferences could not be
identified with a travel-cost model.
Contingent valuation, on the other hand, is well-suited to evaluating
the effects of such flow changes. It does not depend upon changes in
travel behavior. Rather, it measures changes in recreationists'
surplus values which reflect preferences for qualitative changes in
the experience. For this reason, contingent valuation was used in
this study.
_On the Validity of Contingent Valuation
The contingent-valuation method derives its name from the fact that
individuals are asked. to state their surplus values for an item
contingent on· the existence of a market in which to trade the item or
other means of payment. The elicitation of surplus ·values using CV
is accomplished in a survey setting and no money actually changes
hands. This can either be "willingness to pay" for access to the
item or "compensation demanded" to forgo access. For the current
discussion, we will focus on concerns regarding measurement of
"willingness to pay" using CV questions since this is the approach
used in the current study.
Many questions have been raised about CV in the 25 years since it was
_first applied by Davis (1963; 1964). Concerns have existed about
whether money (surplus values). committed in a survey format can be
taken as an accurate indicator of how people would behave if they
really had to pay their surplus values. To provide the reader with a
context for examining the credibility of the white-water beater,
angler and day-use rafter surplus values, it is helpful to review the
types of concerns that have been expressed and to summarize the
results of several empirical studies designed to address such
concerns.
Concerns about the accuracy of CV are generally treated as various
classes of bias. Examples include strategic bias, information bias,
starting-point bias, vehicle bias, and so on. Biases in general are

20

discussed in Schulze, d'Arge and Brookshire, 1981; Desvousges, Smith,
and McGivney, 1983; and Mitchell and Carson, 1981. We will
categorize these effects under two general headings: strategic bias
and hypothetical bias. Strategic bias would be present if
respondents intentionally expressed surplus values in CV exercises
that were different from what they would pay if real money was
involved (i.e., the values they would express in a well-functioning
market). In the present study,-for example, white-water boaters on
commercial trips might think that the fees outfitters charge could be
affected by their responses to the CV questions. This might lead
them to state values lower than they would actually pay if this
concern was not present. They would do this to keep outfitter.fees
as low as possible. Likewise, if recreationists perceive that they
could affect management decisions about actual stream flows, but not
outfitter fees or other expenses, by stating higher values, then they
might state larger values for preferred flow regimes than they would
if they really had to pay the money. In either case, they would be
intentionally misrepresenting their preferences in order to achieve
what they perceive as a preferred outcome.
To the extent that it exists, hypothetical bias does not stem from
intentional distortions, but rather, arises due to people's inability
or unwillingness to engage in the mental processes necessary to
predict what they actually would pay if required to do so. People
who have never before tried to express their preferences· f.or
environmental assets in dollar terms may have a difficult time doing
so on short n9tice while participating in a survey. Several possible
effects of this uncertainty are plausible. Respondents may exercise
conservatism by stating values that are relatively low compared to
what they might really pay. Or, in CV questions, a means of payment,
such as taxes or entrance fees, is usually specified to make the
hypothetical situation seem realistic. Respondents may use the CV
exercise as an opportunity to protest taxes or fees rather than to
express their true preferences for the resource. If conservatism or
protests regarding the hypothetical payment mechanism or some other
similar reaction is dominant, CV values may turn out to be biased
downward compared to true values. Alternatively, people who are
uncertain about their ~rue values might also express values that are
high relative to what they would really pay in order not to be viewed
as miserly or apathetic in the eyes of the researcher. Hypothetical
bias is used to characterize these and other similar possible
distortions because they all stem from the hypothetical nature of CV
transactions.
In the context of the current study, Colorado River recreationists
may find it difficult to determine just how much more they really
would pay for a trip under specific flow conditions. To the extent
that such difficulties result in an upward or downward distortion of
values expressed in CV surveys, hypothetical bias would be present.

.

21

i

It needs to be emphasized at this point that all we have done so far
is speculate about biases. No empirical evidence has been cited that
indicates whether such biases actually exist, much less their
direction and size. Indeed, most of the concerns raised about CV
have involved only speculation. Empirical evidence, on the o~her
hand, has been rather encouraging. CV measures of willingness to pay
have generally been consistent with other methods of estimating
values, and have been shown to be relatively free from the biases
discussed above.
One empirical approach has been to compare CV and travel-cost
values. The two methods have produced rather similar values in a
number of studies (Knetsch and Davis, 1966; Desvouges, Smith, and
McGivney, 1983; .Bishop, Heberlein, and Kealy, 1983; and Sellar,
Stoll, and Chavas, 1985). Comparison of CV values with the cost and
prices of substitutes has been used to support the validity of CV by
Thayer (1981).
Brookshire, et al. (1982) took a different track in a study of air
pollution in the Los Angeles Basin. They and previous writers (see
Freeman, 1979, for discussion and citations) argued that many peopl~
ought to be willing to pay more for property where the air is less
polluted. Thus, by statistically analyzing data on housing values
and housing characteristics including neighborhood air quality, they
were able to infer willingness to pay for improved air quality using
a statistical model of property values. This is a so-called hedonic
price model. Separate estimates of willingness to pay were made
using CV. The resulting hedonic and CV measures were shown to be
consistent.
Much of the speculation about CV's potential inaccuracy has focused
on the possibility of strategic bias. Yet, empirical attempts to
discern strategic influences in CV results have either failed to show
such influences to be present (see Schulze, d'Arge, and Brookshire,
1981, for discussion) or have shown the effect to be relatively small
(Welsh, 1986).

..

This conclusion is supported by a substantial body of research
results from laboratory experiments involving "public goods." Pure
public goods are a general class of commodities and services where
provision to any one member of society involves provision to all .
Examples include national defense and radio signals. Since
Samuelson's (1954) classic paper on the subject, it has been widely
recognized at a conceptual level that people may have strong
incentives to be "free riders" with regard to public goods. That is,
by not revealing their surplus values, they can theoretically
continue to consume a public good without paying. This would be
strategic behavior much like that which would lead to strategic bias
in CV studies. A number of laboratory experiments have involved
public goods where, theoretically, people should have behaved

22

•
strategically (Bohm, 1972; Scherr and Babb, 1975; Smith, 1977;
Schneider and Pommerehne, 1981; Yarwell and Ames, 1979, 1980; Alfano
and Yarwell, 1981; Brubaker, 1982; Tideman, 1983; Isaac, Walker, and
Thomas, 1984; Kim and Walker, 1984). Strategic behavi~r appeared to
be present to a modest extent in some cases, but was generally much
less prevalent and much less influential than economic theory would
lead one to expect. This suggests that strategic bias is not a major
threat to the validity of CV. See Welsh (1986) for further
discussion of this literature and its relevance to CV.
Further support for the validity of CV results has come from recent
experiments involving "simulated markets." If it were easy to
establish fully functioning markets for environmental assets like
Colorado River recreation, CV would not be needed. However, it is
sometimes possible to simulate markets in laboratory and field
experiments, where actual cash transactions and CV exercises can be
conducted and the results compared for the same applications. Such
experiments provide very valuable evidence about how well CV is
working.
Bohm (1972) conducted a laboratory experiment involving a
closed-circuit television program. Several aspects of public goods
valuation and allocation were studied, including actual cash
transactions for admission to the program and contingent values.
Bohm discovered evidence of hypothetical bias. However, the
magnitude of' the bias was not large (.Bohm, 1972, p. 125).
A recent laboratory experiment was conducted at the University of
Wyoming using undergraduates as subjects (Coursey, Hovis, and
Schulze, forthcoming). The goal was to simulate, in the laboratory,
an unpleasant environmental stimulus. The commodity used to
accomplish this was a bitter, unpleasant, but harmless, substance
called "sucrose octo acetate" (SOA). Part I of the experiment
consisted of asking subjects how much they would pay hypothetically
to avoid tasting the substance. Part II required three steps. In
the first step, each subject tasted a few drops of SOA. In the
second step, each was asked for a revised statement of willingness to
pay to avoid consuming the full amount. In the third step of Part
II, monitors attempted to increase respondents' statements of
willingness to pay by bidding them up in $0.25 increments. In Part
III, a group of eight individuals participated in auctions designed
to elicit actual cash bids to avoid drinking SOA. At the end of the
bidding, those who were the four high bidders actually paid their
bids and did not have to drink the SOA, while low bidders drank. The
results of this study revealed that contingent valuation performed
quite well, with the bids in Parts I, II, and III being quite
consistent.

23
•

i

In another recent experiment, CV was compared with a simulated market
involving the sale of fresh strawberries (Dickie, Fisher and Gerking,
1985). Individuals in one sample of Laramie, Wyoming, households
were contacted at home and given an opportunity to purchase
strawberries. A second sample was contacted in the same fashion,
except that hypothetical bids for strawberries were elicited. Each
household, regardless of the sample, was given a set price for the
strawberries and asked how many pints would be purchased. By varying
the price across households, data were collected to develop a demand
relationship between quantity purchased and price per unit. The
demand function based on CV procedures was not statistically
different from the demand function estimated on the basis of the
actual transactions, indicating that the CV estimate of willingness
to pay was a valid indicator of how subjects would behave when actual
monetary transactions were involved.
Another simulated market experiment dealt with actual permits to hunt
deer. Details are given in Welsh (1986), and Bishop, Heberlein,
Welsh and Baumgartner (1984). By way of a quick summary, the deer
hunting opportunities that were valued in that study took place in
the Sandhill Wildlife Demonstration Area, a wildlife research area in
Wood County in central Wisconsin. To maintain the Sandhill deer
population at desired levels, annual public hunts are conducted. The
economic research focused on one-day hunts for deer of either sex
held in November, 1983 and 1984. Approximately 150 deer hunting
permits were issued to hunters selected by lottery from over 5000
applicants in 1983 and again in 1984. In both years, CV estimates
fer Sandhill permits were compared to values based on actual cash
transactions.
In 1983, only four permits could be sold. Auctions were thus
conducted where the researchers sold permits to the four highest
bidders. Other samples engaged in CV exercises measuring willingness
to pay. The CV exercises were designed to be as much like the actual
auctions as possible, except actual cash offers were not involved.
In 1983, respondents were asked to state their surplus value for a
deer hunting permit using two commonly employed question formats from
CV studies. The same question formats were used in the CV studies
and simulated markets to maintain comparability of the resulting
value estimates. The techniques used to elicit surplus values are
commonly referred to as "iterative-bidding" and "open-ended"
questions (for an explanation of these techniques, see Boyle and
Bishop, 1984; Anderson and Bishop, 1986; and Welsh, 1986).
In 1984, there was no longer a constraint requiring that only four
permits be sold. Again, separate samples were involved in parallel
CV and actual cash exercises so that responses to hypothetical and
real dollar offers could be compared. Rather than using
iterative-bidding or open-ended techniques, the 1984 experiment used

24

"dichotomous-choice" questions. Dollar amounts to purchase a deer
hunting permit were randomly assigned to hunters in advance and each
respondent in the simulated market could either accept or reject the
offer. Hembers of a separate sample were asked in a CV exercise to
imagine they had received such offers and were asked whether they
would accept or reject the specified amount.

•

-

CV yielded relatively accurate values for Sandhill deer hunting
permits. There was not a statistically significant difference
between values derived from the 1984 dichotomous-choice questions
when CV and simulated market results were compared. Furthermore, the
1983 and 1984 CV results were quite consistent. It did not matter
much whether iterative-bidding, open-ended, or dichotomous-choice
questions was used in the CV exercises. Only the 1983 simulated
market values were out of line. They were lower than all the other
results (1983 CV, 1984 CV, and 1984 simulated market) and the
difference was statistically significant. Welsh (1986) attributed
strategic behavior in the 1983 simulated market to the auction
formats that were employed since only four deer hunting permits could
be sold. That is 1 when confronted with an actual cash auction,
respondents intentionally bid lower than their true maximum surplus
values in order to try to get a "good deal." However, strategic bias
apparently did not have the same effect on bids in the CV exercise.
In 1984, participants in the simulated market could only respond to a
single opportunity to accept or reject a predetermined offer. This
effectively eliminated the opportunity to behave strategically. The
close correspondence between all the CV results and 1984 simulated
market results indicated that CV worked well in both years.
By way of conclusions for the Colorado River recreation study, the
weight of the evidence discussed above supports the use of CV
measures of willingness to pay as a reasonably accurate approach to
estimating surplus values. CV is capable, when competently applied,
of yielding values that are comparable to travel-cost values and
simulated-market values. The simulated market experiments provide
particularly compelling evidence. Whether the product was as common
as strawberries or as unfamiliar as SOA, values expressed in response
to CV questions were statistically indistinguishable from values
expressed in actual cash transactions. The deer hunting experiment
provided evidence that this result carries over to recreational
opportunities as well. Strategic and hypothetical bias do not appear
to have substantial adverse effects on final value estimates. This
body of research led us to expect CV to perform well in the study of
Colorado River recreation.

'

25
•

Summary
i

The purpose of this chapter has been to review the conceptual and
empirical basis for our study of Colorado River recreation. As such,
a major goal of the study is to measure recreationists' preferences.
regarding dam operating alternatives in a commonly understood unit of
measure, dollars. However, if dollars are to be used as a value
measure, the definition of value must be consistent with general
benefit-cost principles as recorded in the Principles and
Guidelines. Since National Economic Development (NED) Benefits (and
not regional benefits) are being measured, recreation benefits equal
the surplus value generated by recreation, where surplus value is
defined as the maximum willingness to pay of recreationists for
access to Colorado River white-water boating, angling, and day-use
rafting opportunities over and above actual expenditures.
To estimate surplus values for Colorado River recreation, CV
willingness to pay was adopted. Although researchers continue to
investigate the validity of CV measures, there is already ample
evidence that CV measures of willingness to pay are sufficiently
accurate to yield useful results. Though travel-cost ~odels are
-sometimes used in such cases, constraints on the number of trips
taKen by white-water boaters and day-use rafters and other concerns
make CV a more appropriate technique.
In the next chapter we develop the procedures that were used to
accomplish the goals of the study, especially the developnent of the
CV surveys for each of the user groups. Technical appendices willalso be introduced, as needed, to elaborate on specific aspects of
the analyses.

26

CHAPTER 3
RF.SEARCH PROCEOORF.S

Introduction
The ideal research design for this study would be a carefully
controlled field experiment, where a representative sample of
recreationists from each user group is selected and members of each
sample would be asked to experience the activities under a var·iety of
different flow release patterns. In this within-subjects design,
each individual would experience all relevant flow release patterns.
There were at least two major constraints to the use of a
within-subjects design for this study. First, sampled recreationists
would have been required to spend an unrealistically long period of
time experiencing the various flows. For white-water boaters, in
particular, the experimental treatments could have taken several
months. Second, even if recreationists could have been recruited for
such a study, it is unlikely that the operations of the Glen Canyon
Dam could have been modified to permit the req'Uired variation in flow
release patterns.
An alternative to a within-subjects design wo_uld be to use a
between-subjects design. Two different types of between-subjects
designs were considered. Over the last five years, a wide variety of
flow levels and flow release patterns have been experienced by
white-water boaters, anglers and day-use rafters. The first
between-subjects design would have involved obtaining records of
users from previous seasons and selecting a sample of users from each
group which reflected a variety of flow release patterns. This
design, however, was not used for two reasons. First, there are no
records of Glen Canyon anglers. Thus, it is impossible to identify
users from previous ye~rs. User records from previous years do exist
for white-water boaters and day~use rafters, and samples of these two
user groups from previous years were selected for the initial survey
pretests. Respondents who had taken trips 2-4 years ago, however,
were not able to recall as many of the specific aspects of their
recreational experience as were more recent participants. To avoid
confounding the analysis of the relationship between flows and users'
evaluations of the experience, we decided not to sample
recreationists from previous years.
A second between-subjects design could have been used if
recreationists from a user group had collectively experienced a
variety of flows in recent years. This· design would involve
selecting a sample from each user group to directly evaluate the flow
release patterns they had experienced. Across the entire sample,

27

then, the range of flow levels and flow release patterns would be
evaluated. This method was used where sufficient variations in flows
occurred. However, during the design of the study, we could not be
guaranteed that recreationists would collectively have experienced
sufficient variation in flow levels to make such an analysis
possible. Consequently, an alternative method, using descriptions of
the recreational experiences, was incorporated into the study design
to determine users' evaluations of flow levels.
A two-stage procedure was used to develop and evaluate the
scenarios. The first stage involved "attribute sur·veys" for each
activity. These surveys were used to identify the important
flow-sensitive characteristics or attributes of each experience. In
the second stage, the results of the attribute surveys were used to
construct_descriptions (scenarios) of each recreational experience
under different flows. The purpose of this step was to quantify
respondents' preferences for a variety of flow conditions in monetary
terms. The use of scenarios allowed the evaluation of a variety of
flow release patterns even if the individuals in the samples had not
collectively experienced all relevant flows.
The purpose of this chapter is to explain how each of these steps ln
the research process was accomplished. The chapter is organized
around the four major tasks in this study. The first section
describes the procedures used to administer the attribute and CV
surveys. The second and third sections present discussions of the
key design features of the attribute and CV surveys, respectively.
Finally, the fourth section sUlllmarizes the procedures used in the
analyses of responses to the CV questions and the calculation of
estireated surplus values.
Survey Procedures
Background. A total of seven surveys were conducted for this study.
Since it was not known which aspects of the recreational experiences
were most important to participants and how those aspects were
affected by flows, attribute surveys were conducted with white-water
boaters, Glen Canyon anglers, and Glen Canyon day-use rafters.
Samples of recent visitors from each of these groups received an
attribute survey designed to identify the important characteristics
or attributes of their experience and to identify which of these
attributes, if any, are affected by flow levels. In addition to
these three user groups, a sample of white-water boating guides and
private trip leaders also received an attribute survey designed to
identify the specific ways in which different flows affect the
attributes of white-water boating trips in the Grand Canyon. These
results provided the important information needed to design relevant
CV surveys that would be easily understood by respondents.

28

Formal attribute surveys were not conducted with Glen Canyon fishing
guides or day-use raft trip guides because of the small number of
individuals in each of these groups. Only one company provides
day-use raft trips,. using less than 10 guides, and only 7 Glen Canyon
_fishing guides were identified from National Park Service records in
1985. Instead of a formal survey, input from these two group.s of
guides was obtained through informal interviews and discussions.
Contingent-valuation (CV) surveys were developed for white-water
boaters, anglers, and day-use rafters. CV surveys were not conducted
with any of the guides. The use of CV willingness-to-pay measures
was not appropriate for guides because they are paid (rather than
pay) to take recreationists on the river, and evaluation of the
effects of different flows on guides' earnings from their guiding
activities was not an objective of this study.
Sampling. For the white-water boaters and the day-use rafters, names
and addresses of current-season users were obtained from permits
records and trip rosters filed with the National Park Service and the
commercial rafting companies. No such use records existed for Glen
Canyo~ anglers, so on-site interviews at the Lee's Ferry boat dock
were administered on selected days to identify the anglers' names and
addresses.

For white-water boating guides, the National Park Service roster of
qualified commercial river guides was used to select a sample of
individuals who guide trips for commercial rafting companies. Permit
application forms from private groups were used to identify the most
experienced white-water boater (presumed for our purposes to be the
trip leader) from private trips. Glen Canyon fishing guides were
identified by Glen Canyon National Recreation Area personnel, while
the single rafting company that provides day-use raft trips in Glen
Canyon was contactec for information on that recreational ·e:q;erience.
Survey Techniques. Attribute surveys of white-water boaters, day-use
rafters, and white-water boating guides were conducted using mail
questionnaires. The attribute survey of Glen Canyon anglers was
conducted with an on-site questionnaire administered at the Lee's
Ferry boat -dock and parking lot.

Contingent-valuation surveys followed similar procedures.
White-water boaters and day-use rafters were surveyed using mail
questionnaires. The Glen Canyon Angler CV Survey involved two
steps. First, an on-site questionnaire was used to identify a sample
and collect a limited amount of data specific to the trip when
contact was made. Then, a mail survey was conducted to evalt:ate the
trip when contact was made, as well as the different flow scenarios.
Where possible, mail questionnaires were used for the attribute and
CV surveys for two primary reasons: cost and data quality. ~~il

29

-

questionnaire implementation costs are typically less than those for
telephone and personal interview surveys. This is especially true
when a relatively large amount of data is to be collected. The type
of data required from the attribute and CV surveys also made the mail
questionnaire format more appropriate. The attribute survey reqfilred
a careful consideration of the characteristics of the recreational
experience and decisions about the relative importance of each, while
the CV survey required respondents to carefully evaluate several
detailed scenarios of trips under different flow conditions. Such
thoughtful responses would have been more difficult to achieve u~.ing
a telephone survey.
On-site surveys for the Glen Canyon anglers were required because
there are no existing records with which to identify users names and
addresses. The on-site surveys were feasible because the Lee's Ferry
boat dock and parking lot is the access point used by nearly all
anglers. For the CV survey, a mail questionnaire was used to obtain
anglers' evaluations of the alternative flow scenarios due to the
complexity of the survey. The comprehension and evaluation of the
individual scenarios would have taken more time than anglers would
have been willing to- spend on the dock or in the parking lot at the
completion of a day of fishing, especially in unfavorable weather or
as darkness was approaching.
All of the surveys, both attribute and CV, were
"pretested. 11 Pretesting allows the researcher to refine the surve¥
instrument before a final survey is administered. Both attribute and
CV surveys for white-water boaters and day-use rafters were pretested
by mailing draft questionnaires to a sample of individuals from the
respective sampling frames. The Angler CV Survey was pretested in
this way as well. The Angler Attribute Survey was pretested using a
small number of on-site interviews.
Survey Pretests.

As noted in Chapter 2, there are several techniques of asking
contingent-valuation questions, one of which is dichotomous-choice.
In the design of the study we had concerns about which of the
techniques would be best to use in the context of the current study.
To address this concern, iterative-bidding, open-ended, and
dichotomous-choice questions were used in the Angler CV Pretest
Survey to examine the strengths and weaknesses of each technique.
Respondents to this survey were randomly assigned to three groups and
individuals in each group only answered one type of valuation
question. - Dichotomous-choice was chosen as the technique of asking
the CV questions in the current applications. To address fully the
procedures and results of the comparison of techniques conducted in
the angler CV pretest would be tangential to the body of the report.
However, the interested reader can refer to Appendix K for a complete
discussion of the comparison of CV techniques in the Angler CV
Pretest Survey.

30

-

A formal pretest of the white-water guides survey was not conducted.
Instead, a draft of the questionnaire was evaluated in a focus group
discussion involving commercial white-water guides who had a
substantial amount of experience running the Colorado River between
Lee's Ferry and Lake Mead. A focus group discussion was deemed to be
the appropriate format for refining this questionnaire due to the
number and complexity of the questions to be evaluated.
The primary objectives of the pretests were to assess the clarity of
the questions and the comprehensiveness of response categories, as
well as to "pinpoint" problem questions that respondents might find
difficult to answer: No attempt was made to draw inferences about
the effects of different flows from pretest data due to the small
sample sizes involved. In addition, the pretests of the CV surveys
were used to calibrate the CV questions. The dichotomous-choice
technique, used to ask the CV questions, involved asking respondents
whether they would pay a given dollar amount (offer) above and beyond
their actual trip expenditures to take a specified white-water trip,
fishing trip, or day-use raft trip. Data on the range and
distribution of each group's surplus values were required to select
the specific ~ffers for the dichotomous-choice valuation questions in
the final CV surveys. This calibration process and the selection of
specific offers is discussed ~n more detail later in this chapter.
Survey Implementation. The procedures used to administer the
attribute and CV mail surveys were consistent across the three
recreational user groups. Where mail surveys were used, individuals
in the sample were first mailed an advance letter informing them of
the attribute survey. Five days later they were mailed a cover
letter, a question and answer sheet, and a questionnaire. The
purpose of the question and answer sheet was to respond to some of
the concerns respondents might have regarding why the survey was
being conducted. A postcard was sent three days after the
questionnaire to thank those who had responded and encourage those
who had not responded to do so. Nonrespondents were mailed a second
copy of the questionnaire approximately two weeks after the first
mailing of the survey, and a third copy of the questionnaire was sent
by certified mail about four weeks after the initial survey mailing.

For white-water boating guides a slightly longer interval between
mailings was used, and the third mailing was sent by regular first
class mail. These alterations in the procedures were used because
many commercial river guides move about a great deal, and we were
concerned that the addresses we had obtained were not current for
some individuals. A three week interval between mailings was used so
that the previous mailing would have time to be forwarded to reach a
respondent if necessary. A certified mailing was not used because
there was a higher likelihood that the respondent would not be at the
specified address, mitigating the response inducing qualities of the
certified letter.

31

.
The administration of the attribute survey and the on-site portion of
the CV survey for Glen Canyon anglers involved contacting anglers
according to a predetermined sampling pattern as they returned to the
dock at Lee's Ferry in the late afternoon or as they fished from
shore near the Lee's Ferry parking lot. Procedures here will be
explained further in Chapter 6 and related appendices.
Designing the Attribute Surveys
The objective of the attribute· surveys, as has been previously
stated, was to identify the important aspects of the recreational
activity that are affected by releases from the dam. This was
accomplished by designing survey questions to identify the important
attributes that contribute most (positively or negatively) to the
recreational experience. Informal interviews with resource managers
and members of the appropriate group of river guides were used to
assist in the identification of the particular aspects of each
activity that should be addressed in the attribute surveys. A
pretest was also used to identify appropriate response categories.
Although there were many questions in each of the attribute surveys,
three questions were particularly useful in identifying important
attributes. In the white-water boater and day-use rafter attribute
surveys, three open-ended questions were asked before any specific
attributes had been suggested to respondents. In the first question,
respondents were as~ed about their reasons for taking a trip (see
Figure 3-1). Second, respondents were asked to identify the specific
things (attributes) that would contribute most to an excellent or
perfect trip. An example of this question for white-water boaters is
shown in Figure 3-2. Next, a similar question asked respondents to
list the things that would contribute most to a poor trip. An
example of this question for day-use rafters is shown in Figure 3-3.
The purpose of these open-ended questions was to give respondents an
opportunity to choose their own response categories before being
exposed to the attributes that had been specified by the
researchers. Responses to th~se open-ended questions were examined
to determine if any of the reported attributes could be affected by
releases from the dam.
For the Glen Canyon angler attribute survey, questionnaire length and
time constraints imposed by the use of an on-site interview limited
the number of questions that could be asked, so the open-ended
attribute questions were not used. Instead, a closed-ended question
format was used for rating the importance of the positive and
negative attributes of the Glen Canyon fishing experience. However,
one of the response categories was open-ended so that anglers could
cite their own reasons if they felt the provided answer categories

32
•

Figure 3-1 ·
Question Asking White-Water Boaters About the
Most Important Reason f'or Taking Trip

When you first decided to take a Grand Canyon trip, what was the
ONE thing you looked forward to most?

Figure 3-2
White-Water Boater Positive Attribute Question

What things would contribute most to an excellent or perfect raft
trip in the Grand Canyon for you?

Figure 3-3
Day-Use Ra.f'ter Begative Attribute Question

What things would contribute most to a poor one-day raft trip in
Glen Canyon for you?

'

33

were not appropriate for them. An example
positive attribute question for anglers is
each attribute, anglers were asked to rate
from 1-3, with "not important" rated as 1,
as 2, and "very important" rated as 3.

of the closed-ended
shown in Figure 3-4. For
the importance on a scale
"somewhat important" rated

To determine the "important" attributes of each experience, responses
to the closed-ended questions asking about specific attributes and
the frequency with which respondents cited attributes in responding
to the open-ended questions were analyzed. Attributes with the
highest relative rankings for the closed-ended questions and those
cited most frequently for the open-ended questions were determined to
be important. Convergence of responses from both types of questions
was examined te confirm that the identified attributes were in fact
important. These results were then examined in light of comments by
resource managers from the National Park Service and the findings
from contacts with guides to determine the attributes that are
affected by flow levels. When possible, we also examined
respondents' answers to learn whether responses to the attribute
questions varied with the actual flows ~xperienced. This type of
analysis could help to coP£irm whether attributes are or are not
sensitive to flow levels. Finally, subgroups of respondents were
examined to see i f individuals who had more experience with a variety
of flow levels might answer the attribute questions differently from
those with less experience.
To complement these open-ended questions, a series of questions about
the respondent's experience and specific characteristics of the trip
were asked. These questions were developed using input from guides
and resource managers. Based upon the results of the attribute
surveys for each activity, the important, flow-sensitive attributes
of each activity formed the basis for the construction of the
scenarios for the CV surveys describing the change in the experience
that would occur under different flow conditions.
The results of the attribute surveys and the specific flow-sensitive
attributes identified for each activity are discussed in Chapters 4
through 6.
·
Designing the Contingent-Valuation Surveys
Six issues must be dealt with in designing a CV study: ( 1) Whose
values will be estimated? (2) How will the item to be valued be
defined? (3) What payment vehicle will be used? (4) How will the CV
question be posed? (5) How will the data be analyzed? and (6) What
supplemental data will be required? Let us consider how each of
these issues was resolved in designing the current study.

34

Figure 3-4
Glen Canyon Angler Positive Attribute Question

'

How important would each of the following be in contributing to
an excellent or perfect fishing trip at Lee's Ferry for you?
(CIRCLE ONE NUMBER FOR EACH ITEM)

Not
Important

Somewhat
Important

Very
Important

Catching a trophy fish

1

2

3

Catching your limit

1

2

3

Good weather

1

2

3

High water level

2

3

Low water level

2

3

Camping along the river

2

3

Seeing few others

2

3

c.
"

3

2

3

Rising water level
during the day
Falling water level
during the day
Other

1

35

Values to be Estimated. Concerning whose values will be estimated,
it was decided that white-water boaters, Glen Canyon anglers, and
day-use rafters would be directly affected by flow releases from Glen
Canyon Dam. The final surplus values were based on samples drawn
from 1985 ·visitor rosters.

Still, it is worth noting that Qther people may also be indirectly
affected. In particular, Grand Canyon National Park is a major
national environmental asset. If other sections of the Glen Canyon
Environmental Studies identify clear links between stream flows and
damages to the aquatic or riparian ecosystems, there could be
substantial surplus values associated with the existence of certain
resources in the canyon, as well as additional surplus values
associated with the option of future use which should be included in
a full benefit-cost analysis. (For further discussion of option and
existence values, and relevant citations see Fisher and Raucher,
1984; Randall and Stoll, 1983; Smith 1984; and Boyle and Bishop,
forthcoming.) Because such environmental effects were not documented
when this study commenced, because option and existence values are
more controversial and less well-researched than use values, and
because project resources were limited, it was decided that the
current study would deal only with the use values of the three groups
who would be directly affected by flows.
Definition of Item. When people are asked to state a surplus value
in a CV exercise, they must have a clear idea of what they are
valuing. Thus, a second issue in designing a CV study is how to
define the "item" to be valued. In the present study, three kinds of
"items" were valued. First, all three groups valued an actual trip.
Since day-use rafters and white-water boaters would not, in all
probability, have taken more than one trip during 1985, they were
simply asked to value the trip taken ir. 1985. Many Glen Canyon
anglers, however, take more than one trip per year. This problem was
alleviated by the use of the on-site sample selection so that anglers
could be asked to value the trip taken on that date. To help anglers
recall this trip, we asked them for some information in the on-site
interview about their trip (see Appendix G) and used this information
in the CV survey to jog their memories about the trip (see Appendix
E).

In addition to actual trips, anglers and white-water boaters were
asked to value two types of scenarios: "flow scenarios" and what
might be termed "environmental impact scenarios." After respondents
had answered a CV question for their actual trips, they were asked to
value trips at several alternative flow levels as described by flow
scenarios. Since many, if not most, of the respondents had only
experienced a limited number of flow levels, it was necessary to
describe the implications of flows for the quality of the trip. The
flow scenarios described trips under different flow conditions,
primarily in terms of the changes that would occur in the important,

36

flow-sensitive attributes identified in the attribute survey. These
scenario descriptions were supplemented with the information gained
from contacts with guides and resource managers. A great deal of ·
effort was exerted to insure that the scenario descriptions were
based on documented facts about the recreational experiences and that
they were worded in neutral, matter-of-fact language.
-

Even if scenarios are designed to be meaningful and to give detailed
descriptions, a short description in a questionnaire is not a perfect
substitute for actual experience. Thus, where possible, we also
designed CV questions to assess the economic value of alternative
flows based on the variation in flow conditions actually experienced
by respondents. This was done by developing direct statistical
relationships between flow rates and surplus values. Our ability to
develop this relationship, however, was constrained by the range of
flow levels and types of flow release patterns that recreationists
experienced in 1985, an atypical year of relatively high constant
flows. To fully model the relationships between surplus values and
constant or fluctuating flow levels, it would have been necessary to
manipulate releases from Glen Canyon Dam in ways that would have
allowed adequate subsamples of recreationists to experience both
constant and fluctuating flows across a full range of average daily
flows from perhaps 3,000 to 40,000 cfs. Furthermore, it would have
been desirable to vary flows randomly in order to reduce the risk of
having unrecognized seasonal or other influences in the data.
The Bureau of Reclamation could not guarantee at the outset of the
GCES that they could manipulate the dam in ways that would make it
possible to select samples of individuals who, as a group, would have
experienced all of the relevant flows to be evaluated. After all,
Glen Canyon Dam is a large facility whose operation is constrained by
many requirements dictated by social, legal and governmental
institutions, design characteristics, and the variability of nature.
Consequently, the ability of the Bureau of Reclamation to provice the
flows needed for this study was limited. Thus, the scenarios
provided an alternative way of meeting study objectives if data to
estimate direct relationships between flows and actual trip values
could not be generated.
The second type of scenario evaluated in this study involved direct
environmental impacts. One of the disadvantages of carrying out
various phases of the Glen Canyon Environmental Studies
simultaneously was that the conclusions of other researchers were not
available until near the end of the GCES. For our study, this meant
that the impacts of flows on key recreational parameters, such as the
size and nunber of beaches available for camping on the white-water
trips, were unclear. Nevertheless, many potential environmental
impacts were stated at the outset. To anticipate the implications
for recreation if some of the more relevant potential impacts were

37

verified,· scenarios were designed where conditions were exactly like
those on the actual trip except that one environmental parameter,
important to that experience, was changed. A CV question was
designed to measure the value of this modification in the actual
trip.
Payment Vehicle. The third issue to be dealt with in CV study design
is the determination of the "payment vehicle. II CV questions normally
specify a means or method of payment. In the present study, trip
expenditures were used as the payment vehicle. Respondents were
first asked about their actual trip expenditures. This step was
designed to help them begin to think about their trips in monetary
terms. In the technical language introduced in Chapter 2, the
expenditure question was designed to focus respondents' attention on
the market value portion of the total value. CV questions for the
actual trip and scenarios then asked respondents how much more they
would be willing to pay in additional trip expenditures. Thus,
surplus.value was measured in terms of potential additional
expenditures to meet the cost of recreation.

Trip expenditures were chosen as a payment vehicle because they meet
·the key criteria of being both realistic and neutral (Mitchell and
Carson, 1981 ). The purpose of the payment vehicle is to help make
the CV question seem realistic to the respondent. Trip expenditures
are a very real part of the experience for most adult participants.
Neutrality is important to ensure that the payment vehicle does not
draw too much attention to itself, but allows the respondent to focus
on the item to be valued. For example, taxes are a realistic payment
vehicle for some publicly provided amenities, but using taxes as a
payment vehicle might cause respondents to concentrate on protesting
current tax rates rather than expressing the value they place on the
amenity. Trip expenditures appear to be relatively immune to such
effects.
Choice of" a Contingent-Valuation Technique. The fourth issue in
designing CV studies is to determine how the CV questions will be
asked. In the present study, the technique used is termed
"dichotomous-choice." This method asks whether the respondent would
or would not pay a prespecified amount (the "offer"), an amount in
our case that would be over and above actual trip expenses. The
prespecified amount is varied randomly over respondents. This
technique is dichotomous in the sense that respondents have only two
choices: yes, they would pay the increased expenses, or no, they
would not. The yes/no responses are analyzed along with the amount
of the offer and other variables to estimate surplus values. The
method of analysis will be explained in the next section.
Data Analysis. The fifth issue in designing CV studies regards the
methods used to analyze responses to CV questions to obtain estimates
of surplus values. This· is a somewhat complicated technical problem,

38

~specially for the analysis of dichotomous-choice responses, and the
next major section of this chapter is devoted to an explanation of
the analytic procedures used in this study.

Supplemental Data Required. The sixth issue relates to the types of
data to be obtained in the surveys in addition to CV data.· Normally,
it is possible to augment value information in ways that support
resource management. The present study gathered attribute data not
only to facilitate the formulation of scenarios, but also because
information on what contributes positively or negatively to trip
enjoyment may be directly useful to the National Park Service in
managing river trips. Furthermore, attribute data provided valuable
information for assessing the success of the CV exercise.
Consistency between results from the attribute surveys and CV results
is to be expected. If, for example, flow level A is preferred to
flow level B based on attribute data, flow level A should have a
higher surplus value. If attribute and CV results are contradictory,
.then the interpretation of results would be difficult.

The CV surveys conducted in this study also gathered detailed
socioeconomic data on the user groups studied. These data were
needed to facilitate analysis of the CV data, ·to compare individuals
in the CV samples to those from the attribute survey samples and may
also be of direct interest to resource managers.
The survey design decision just discussed can be illustrated by
presenting several examples of the questions employed in the surveys.
Figure 3-5 shows the expenditure question for anglers followed by· the
CV question for the actual trip. Notice the definition of the "item"
to be valued (actual trip), the expenditure vehicle (increased
expenditures), and the dichotomous-choice format for the CV
question. The blank ~n the CV question was filled in with a randomly
assigned offer. Figure 3-6 is the 5,000 cfs (low water) constant
flow scenario presented to white-water boaters along with the
associated CV question. Figure 3-7 presents an example of an
environmental impact scenario from the angler survey, where the
postulated trip would be comparable to the respondents' actual trip
except that the chances of catching a fish larger than three pounds
would be doubled.
All of the CV questions were designed so that their format would be
consistent across surveys of user groups, and between actual trip and
scenario questions. It is important to note that the CV questions
for each of the scenarios were tied to respondents'. actual trips. To
remind white-water boaters of their actual trip, they were told, in
the survey, the date of their trip and the average flow levels
experienced. In a similar way, anglers were reminded of the date of
their actual trip, as well as the number of fish they reported
catching and the size (pounds and inches) of the largest fish caught
on that day.

...

39

Figure 3-5
Gl:en Canyon Angler Expenditure and Actual Trip CV Question

As near as you can recall for the trip when you filled out our short
survey, about how much was your share of total trip expenses for the
following items? (Include only money you personally spent. I f you
didn't spend money on a certain item, please put $0). [PLEASE
CALCULATE AND FILL IN THE TOTAL ON THE LAST LINE].

•

Gas and Oil for vehicle

$_ _ __

Food and Beverages

$_ _ __

Lodging, Camping

$_ _ __

Fi~hing

$-~~~~~

equipment/bait/license

Guide fees

$

Boat/equipment rental

$

Airfare

$

Car rental

$

Other

$

TOTAL YOU SPENT ON THIS TRIP

$_____

Would you still have gone on that particular trip to Lee's
Ferry if your expenses had been $
aore than the total
you just calculated? (CIRCLE ONE NUMBER)
YES, the trip would still be worthwhile
2

NO, it would not be worthwhile

40

Figure 3-6
White-Water Boater Constant Flow Scenario (5,000 cfs)
and Associated Valuation Question

At a constant flow of 5,000 cfs, the speed of the river is relatively
slow, reducing time for side canyon visits and other attractions.
Boaters must break camp early to stay on schedule. Although rapids
are present at this low water level, the waves are smaller and do not
produce the big "roller coaster" ride created by higher flows. Due
to exposed rocks, some rapids may be so difficult that it is likely
passengers would have to walk around them. However, camping
opportunities are abundant with many large sandy beaches exposed.
We would now like you to imagine that you are presently deciding
whetqer or not to go on a Grand Canyon white-water trip. Imagine
that the trip would be the same as your last trip (e.g., the same
people, same food, etc.) with two exceptions:
The water level would be constant at 5,000 cfs
AND

Your individual costs for the trip increased by $_ __
(over the total cost you calculated on page 8, question
A26)
Would you go on this trip?

(CIRCLE ONE NUMBER)

1 YES, I WOULD PAY THIS AMOUNT TO TAKE THE TRIP
2

NO, I WOULD NOT PAY THIS AM:lUNT TO TAKE THE TRIP

,.

41

!

Figure 3-7
Glen Canyon Angler Environmental Impact Scenario (Larger Fish) and
Associated Valuation Question

A survey of anglers at Lee's Ferry last year showed that about 15
percent of them reported ·catching a fish larger than three pounds,
and only 3 percent reported catching a fish larger than four pounds.
These numbers reflect how an average angler might do on any
particular day at Lee's Ferry. We realize that no one is exactly
average, but we would like you to suppose that the fishery at Lee's
Ferry changed in such a way that your chances of catching one of
these bigger fish were to double. If you feel you are an average
fisher, your chances of catching a fish bigger than three lbs. would
now be about 30 percent, while your chances of catching a fish bigger
than four lbs. would now be about 6 percent. If you think you are
not an average fisherman at Lee's Ferry, your chances would vary
accordingly.
Now we would like you to imagine you are deciding whether or not to
take a trip to Lee's Ferry. On this trip all of the fishing
conditions (water levels, weather, nlUilber of other anglers coming to
Lee's Ferry, etc.) would be the same as the trip when you filled out
our survey with two exceptions:
Your chances of catching a big fish (over 3 pounds) would be
doubled
AND

Your expenses of a trip to Lee's Ferry increased by $ _ __
(over the total you calculated on page 3)

Would you go on this trip?

•

(CIRCLE ONE NUMBER)

1

YES, it would still be worthwhile

2

NO, it would not be worthwhile

42

-

In addition, the valuation sections in both surveys were preceded by
a series of questions asking respondents about the characteristics of
their actual trip and questions asking them to evaluate the trip in
qualitative terms.
The dichotomous-choice question formats generated "yes" or "no"
responses to the CV questions. The offer amounts as specified in
each question, and other data regarding the characteristics of
respondents and their views of the the recreational experience were
used as the basis for estimating surplus values. The nature of the
analyses employed to derive the surplus values are discussed below.
Estimating Surplus Values from Dichotomous-Choice Responses
This section explains the derivation of surplus values from responses
to the dichotomous-choice questions. This process is first discussed
generally, and then in more specific detail using the appropriate
technical terms. Though the more technical discussion may not be
fully understood by some readers, our goal is to give all readers,
regardless of their background, a basic understanding of the
derivation of these surplus values which are discussed in the
succeeding chapter::..
The dichotomous-choice approach has advantages for the current study,
as explained below. However, one of the disadvantages is the
nonintuitive nature of the data analysis. The data analysis would
have been much simpler to explain had we used open-ended or
iterative-bidding questions where respondents were specifically asked
to state their surplus values. With these techniques, a simple
average of the stated surplus values could have been calculated to
determine the average value per trip.
Responses to dichotomous-choice questions are not as easily
analyzed. Suppose, for example, that a respondent answers "yes" to a
dichotomous-choice question involving an increase of $100 in expenses
for an actual trip. This response indicates that the respondent's
surplus value is at least $100. There is no way to tell from simply
observing a "yes" response how much above $100 an individual's
surplus value is. Nevertheless, by analyzing the yes and no
responses of a large number of recreationists, the estimated surplus
value per trip can be computed.
A simple example will illustrate how surplus values are calculated
from dichotomous-choice data. Suppose for the sake of argument that
there is another group of recreationists on the Colorado River who
prefer to travel by innertube. "Tubing," however, is not very
popular, having only four participants, each of whom takes one trip
per year. Suppose that one "tuber" has a surplus value of $100 per

•

43

.

trip, two have surplus values of $200 per trip, and the fourth has a
surplus value of $300 per trip. Suppose also that we ask all four an
open-ended CV question to obtain this information, and they all
truthfully report their full surplus values. The average value is
calculated as follows:
average value
per trip
= $100 + $200 + $200 + $300 = $200
4
The same average can also be constructed as a weighted average.
Twenty-five percent of the tubers (1 out of 4) would pay $100 but no
more; 50 percent, $200, but no more; and 25 percent will pay $300 but
no more. Thus, the following weighed average value can be
calculated:
average value
per trip
=0.25($100) + 0.50($200) + 0.25($300) = $200
This $200 is the same average value per trip as was calculated
before. It was simply calculated in a different way.
While it is less intuitive at first, a third way of calculating the
same average value -- and the way most relevant for understanding ·
dichotomous-choice.values -- is to proceed in $100 increments. All
four tubers ( 100%) are willing to pay $100. Seventy-five percent are
willing to pay an additional $100. These are the people with surplus
values of $200 and $300. Twenty-five percent (the person with a
surplus value of $300) are willing to pay an additional $100
increment above $200. No one is willing to pay more than $300.
Thus, the following weighted average can be calculated:
average value
per trip
= 1.00($100) + 0.75($100) + 0.25($100)

•

= $200

Again, the same $200 weighted average value per trip is obtained •
The reason for proceeding in increments is to avoid double-counting.
For example, in the first step $100 of value was counted for everyone
(100· percent of the tubers). Thus, it was not necessary to recount
the first $100 again when we considered willingness to pay at $200
and $300. The same argument is applied at each increment.

44

In actual CV surveys, any given tuber would normally be asked only
one CV question regarding his or her actual trip, but the principles
would be the same. Suppose that there are 4,000 tubers rather than
four. By asking hundreds of tubers a dichotomous-choice question
with a randomly assigned dollar offer chosen from a broad range of
values, the probability of a respondent saying nyes•1 or "non at
various dollar amounts can be estimated. These probabilities· can
- then be used to calculate an average value per trip using the same
principles as were used in calculating the weighted average based on
increments.
All of this can be readily interpreted in terms of probabilfty
theory. Let x be a random variable representing tubers' surplus
values. Let F(T) be the probability that a randomly selected tuber
would respond "no" to a dichotomous-choice offer of $T. F( ) is a
cumulative distribution function because anyone unwilling to pay $T
should also be unwilling to pay all larger amounts. Thus, F(T) gives
the probability that $T is greater than the maximum willingness to
pay (above and beyond trip expenses) of a randomly selected tuber.
In turn, the probability that this individual will answer yes to an
offer of $Tis [1-F(T)].
Hanneman (1984) has shown that the expected value E(X) of surplus
values, or maximum willingness to pay, for the population can be
calculated from a continuous distribution as:
SCX>

E(X)

[1-F(T)]dt.

('1)

0

For a discrete distribution, equation (1) can be rewritten as:
E (X)

=

n
L: [1-F(T.)](T. - T.)
i
i
J

( 2)

i=l

where n is the number of discrete values taken by T, and T.
corresponds to the i-1 value of T. Thus, T. is always gre~ter than
1
or equal to T .. A more detailed discussion of the technical
consideration~ for specifying the cumulative distribution function
and for dichotomous-choice· estimates of surplus value is presented in
Appendix L.
The application of Equation 2 can be illustrated using the simple
four tuber example. A graphical representation is presented in
Figure 3-8. The horizontal axis shows surplus values per trip and
probability of answering no to a dichotomous-choice offer is shown
on the vertical axis (T =$100, T =$200, and T =$300). The
1
2
3
cumulative density begins at zero, since the probability that a

•

I
I
I

45

member of this population will answer "no" at amounts less than $100
is zero. It jumps to 0.25 at $100, the maximum surplus value of 25
percent of the population. In mathematical terms, F(100)=0.25. At
$200, the F(200)=0.75, since the $100 person and the two $200 people
would start saying no at amounts greater than $200, and F(3p0)=1.00
because everyone would say "no" at amounts greater than $300.
Using equation (2) we can calculate the expected surplus value for
the tuber example.
E(X) = 1.00($100) + 0.75($100) + 0.25($100) = $200.
Note that this calculation is based on one minus the cumulative
distribution function. Using the formula for the area of a
rectangle, the surplus value corresponds to the shaded areas ih
Figure 3-8.
Once again, we have calculated the same estimated surplus value using
a different formula. It is simply interpreted as the expected
surplus value per trip for the population.
In actual applications, the number of observations for any given
dollar amount is generally small due to budgetary restrictions on the
choice of a sample size and the wide distribution of surplus values
that can occur . . To estimate the cumulative distribution, given this
data limitation, we use a legit equation which corresponds to a
logistic distribution (see Hanneman, 1984). The legit function is
fitted to the data using maximum likelihood procedures, and the
estimated logit equation represents a continuous cumulative
distribution function. The calculation of expected surplus values
follows the calculation outlined by equation (1) and the result
(surplus value per trip) is the shaded area in Figure 3-9, which
simply corresponds to the shaded area in Figure 3-8, the surplus
value for a discrete distribution.
Summary

•

The research procedures outlined here describe a two-stage approach
to analyze the effects of different Glen Canyon Dam releases on
downstream recreational experiences. In the first stage, attribute
surveys are used to identify the important, flow-sensitive
characteristics (attributes). The results of these analyses were
used to construct descriptions of the recreational experiences under
different flows (flow scenarios) or changes in an environmental
parameter (such as the number and size of beaches available for
camping). These flow scenarios were then evaluated by recreationists
in the second stage, the contingent-valuation surveys.

46

Figure

3-8

Relatlonshlp Between Tuber Surplus Values and
Answering No to a Dichotomous-Choice ,CV Question

1.00

0. 75

Proportion
of Population
Answering No

F(x)

0. 50

0.25

0.00
$100

$200

$300

Surplus Value Per Trip

..

47

Figure

3-9

Cumulatlve Distribution Function for a
Continuous Distribution of Surplus Values

F(x)
Proportion
of Population
Answering No

0

Surplus Value Per Trip

48

This two-stage process utilizing respondents' evaluations of flow
scenarios was required because few respondents experienced the
necessary variation in flow conditions during the course of their
recreational experience. Most respondents did not have sufficient
experience with the different conditions to allow a reliable
assessment based on data from their actual trips.
Hopefully, the discussion so far has clarified how CV is done,
including how questions are designed and how values are calculated.
At this point, attention will turn to the broader issue of reporting
the empirical results from the surveys.
In the case of white-water boating, results of the attribute and CV
surveys were supplemented by the guide and trip leader survey
results. We will begin the white-water boating analysis by examining
the guide and trip leader results.

•

49

CHAPTER 4
SURVEY OF WHITE-WATER RAFT TRIP COMMERCIAL GUIDES .AND
PRIVATE TRIP LEADERS

Introduction
Commercial white-water guides and private trip leaders are experts on
the impacts of different Glen Canyon Dam releases on white-water raft
trips. Unlike most commercial or private trip passengers, they have
experienced a variety of river conditions and, especially for the
commercial guides, have directly experienced a variety of different
flow conditions on Grand Canyon white-water boating trips. The
purpose of surveying guides and trip leaders was to identify their
flow preferences and their perceptions of the impacts of different
Glen Canyon Dam releases.
In this chapter we first discuss the role of commercial guides and
private trip leaders during white-water boating trips. Next, the
sampling and survey procedures are explained. The results of the
survey and their implications for this study are then discussed.
Background
Approximately 80 percent of the Grand Canyon white-water user days
during a typical year occur on commercial trips. Twenty-one licensed
rafting companies (outfitters) are authorized to operate trips on the
Colorado River in Grand Canyon National Park. Commercial trip
passengers contract with an outfitter to provide a boat, other
rafting equipment, food, and a guide. Commercial trips use both oar
and motor-powered rafts and typically run from 3-4 days for a motor
trip covering only the uppe~ stretch of the river from Lee's Ferry to
Phantom Ranch, to 20 days for an oar-powered trip that covers the
full 250 river miles through Grand Canyon National Park. The
National Park Service (NPS) maintains a roster of qualified Grand
Canyon white-water raft trip guides that is updated annually. Each
outfitter employs a staff of such guides who lead oar trips from
April through October, and motor trips from ~..ay through September.
In addition, a pool of qualified free-lance guides who do not work
for any one outfitter are available to outfitters as needed.
The remaining 20 percent of the recreational user days are allocated
to "do-it-yourself" or private party trips for which individuals
furnish their own boats, rafting equipment, food, and guides or boat
operators. Individuals must apply for a private permit, and permits
are awarded in the order in which applications are received.
Currently, there is a waiting list of approximately 3-4 years for
private permits.

50

1

-To help the National Park Service advise applicants of the types of
equipment and skill that are appropriate for a Grand Canyon trip,
individuals applying for a private permit are required to list the
most experienced white-water boater who will be on the trip. These
individuals were considered, for our purposes, to be trip .leaders,
even if the use permits were not issued in their names.
Commercial guides and private trip leaders are of particular interest
in this study because they have the most first-hand experience with
white-water boating under a variety of conditions, and can provide
accurate and detailed views on the impacts of different Glen Canyon
flows on white-water raft trips in the Grand Canyon.
Survey Procedures

Sampling. A sample of commercial guides was selected from the
. National Park Service's file of qualified guides, and representing
guides currently employed by an outfitter, active free-lance guides,
and guides who were not currently leading trips but had been active in
one or more rafting seasons since 1982. One _hundred-ninety commercial
guides were randomly selected from the 450 names in the NFS guide
file.

Private trip leaders were selected from 198°5 NFS trip launch records
which identified the trip leader (most experienced boat operator) from
each of the private trips scheduled to launch during 1985. A total of
195 private trip leaders were selected from the approximately 223
private parties that received permits to raft the Colorado River
through the Grand Canyon during the 1985 season. Thus, a total of 385
guides and trip leaders were included in the sample.
Response Rate. The first contact with commercial guides and private
trip leaders in the sample occurred in December, 1985. Completed
questionnaires were received from 288 guides, 75 percent of the total
sample. Private trip leaders were slightly more likely to return a
completed questionnaire (78 percent) than commercial guides (72
percent). Fifteen of the questionnaires were returned as
undeliverable.
The response rate as a percent of all deliverable surveys is 78
percent (see Table 4-1). The analyses contained in this report are
derived from 286 completed questionnaires: 134 completed by
11
commercial guides and 152 completed by private trip leaders.-

11 Five surveys were received after the analysis of responses for
this survey was completed. While these are included in the
response rates for Table 4-1, they are not included in the results
reported here.

51
Table 4-1.

Commercial Guide and Private Trip Leader Survey
Response Rate

Surveys
Completed Surveys
Undeliverable Surveys
Surveys Not Returned
Refusals
TOTAL

Percent of
All Surveys

Percent of
Deliverable Surveys*

75%
4
17
4
100%

78%
18

4
100%

* As

noted in the text, 15 questionnaires were returned as
Thus, the percentages in this column are computed
from a sample size of 370 rather than 385.
undel~verable.

Guide Survey Results
The primary objectives of this survey were 'two-fold. First, the
survey provided data on guides' perceptions of the minimum and
maximum flow levels for safely conducting Grand Canyon white-water
boating trips, the flow levels which maximize passenger enjoyment,
and the range of daily fluctuations in flows that they could tolerate
without undue difficulties for themselves or passengers. Second,
respondents were asked to describe the impacts that different Glen
Canyon Dam release patterns have on white-water trips and the steps
that they would take to "buffer" these impacts so that passengers
receive a high quality recreational experience. Different
questionnaires were sent to commercial guides and private trip
leaders to take into account some of the differences between these
two groups (see Appendices A and B).
The data from this survey were used to provide a description of the
context in which Grand Canyon white-water boating trips take place
that can be used to aid in the interpretation of responses to the
questions in the white-water boater attribute and CV surveys.
Results were also used in the design of the white-water boaters'
contingent-valuation survey, especially in the construction of the
alternative flow scenarios which passengers were asked to evaluate.
It will be helpful in presenting the results from the guide survey to
distinguish between commercial guides who lead motorized raft trips
and those who lead oar trips. Thus, in reporting the results, we
will classify respondents as being a commercial motor guide a
commercial oar guide or a private trip leader. Only a few of the
private trips use motor-powered boats, so for this survey all private

52

trip leaders were presumed to be giving responses for oar-powered
boats. These distinctions are useful for drawing contrasts between
the ways guides from these different types of trips responded to
specific questions in the survey.
Evaluation of' Constant Flow Levels. Respondents' evaluatio_ns of
constant flow levels between 2,000 cfs and 80,000 cfs on a
qualitative scale ranging from very unsatisfactory, to very
satisfactory, followed a bell-shaped curve. Responses were similar
across the three groups of guides (commercial motor, commercial oar,
and private trip leaders). As shown in Figure 4-1, all three groups
rated constant flow levels below 5,000 cfs as very unsatisfactory.
At 10,000 cfs, respondents' ratings of the constant flow levels
crossed the neutral line, and ratings peaked at flow levels between
20,000 and 25,000 cfs. Above these optimum flows, ratings decline
and cross the neutral line again at about 40,000 cfs ..

We also asked respondents to state the specific constant flow level
that they would prefer as a commercial guide or private trip leader.
Commercial motor guides reported a mean flow level of 20,622 cfs,
whil.e commercial oar guides and private trip leaders reported
slightly higher mean flow levels of 26,180 and 25,158 cfs,
respectively (see Table 4-2). These.findings correspond closely to
the optimum flow levels displayed for the qualitative rating scales
in Figure 4-1.
Table 4-2.

Pref'erred Constant Flow Levels

Flow
Preferred constant flow:
Mean
Standard Deviation
Sample Size

* Statistics

Commercial
Motor Guides
20,622 cfs a*
6,096
78

Commercial
Oar Guides
26, 180 cfsb
10,583
50

Private
Trip Leaders
25, 158 cfsb
7,515
145

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with differ·er..t superscripts
are statistically differer.t at the 0.05. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents. For example, the two statistics
above with a "b" in their superscripts are not statistically
different from each other. However, they are both statistically
different at the 0.05 level from the statistic with an "a" in the
superscript.

53

Figure

4-1

Guides' Constant Flow Level Preference Ratings

Very
Satisfactory

5

Somewhat
Satisfactory

4

Neutral

Somewhat
Unsatisfactory

3

2

Very
Unsatisfactory

0

10

0

20

30
40
50
Average Flow Level in cfs

60

(x 1 000)

-6-

Commercial
Motor

x- Commercial
Oar

•- Private

70

80

54

In contrast to the optimum constant flow levels, the reported minimum
constant flow level for safely running river trips with passengers
ranged from 8 ,405 cfs for commercial guides on motor trips to 9, 198
cfs for guides on commercial oar trips. Private trip leaders
reported a minimum flow of 9,025 cfs. The minimum flow levels,
reported in Table 4-3 were not statistically different across the
three groups of respondents. About half of each group reported that
the minimum constant flow level requirement was 10,000 cfs or more,
and nearly all respondents reported a minimum flow requirement of
3,000 cfs or more for a safe trip. It is interesting to note that
the average minimum flow levels correspond very closely to the flow
levels at which the preference ratings cross the neutral line in
Figure 4-1.
Table 4-3.

Reported Minimum Constant Flow Levels :for Running River
Trips Sa:fely with Passengers

Flow
Minimum level:
Mean
Standard Deviation
Sample Size
Percent with minimum level
of 3,000 cfs or above
Percent with minimum level
of 10,000 cfs or above

Commercial
Motor Guides
8,405 cfs
3,344
78
1ooa*
42

Commercial
Oar Guides
9' 198 cfs
4,859
52

Private
Trip Leaders
9,025 cfs
4 '271
138

88b

97a

52

48

* Statistics

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
are sta~istically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

The mean maximum constant flow level reported by respondents for
running rapids safely with passengers averaged 59,014 cfs for
commercial motor trip guides, 54 ,910 cfs for coIDiilercial oar guides,

.

55

and 47,210 cfs for private trip leaders (see Table 4-4). Private
trip leaders appear to be more sensitive to very high flows than
commercial guides. Forty-five percent of the private trip leaders
reported that the maximum level at which rapids can be safely run
with passengers is 40, 000 cfs or below. Less than one-third of the
commercial trip guides reported a maximum safe flow level of 40,000
cfs or below.
Table 4-4.

Reported Maximum Constant Flow Level f'or Running River Trip
Saf'ely with Pasaengers

Flow
Maximum flow
Mean
Standard Deviation
Sample Size

Commercial
Motor Guides
a*
59,014 cfs
25,292
69

Commercial
Oar Guides
54,910 cfsa
23,635
50

Private
Trip Leaders
47,210 cfs
16,306
131

Percent with maximum level
of 30,000 cfs or less

9a

12ab

20b

Percent with maximum level
of 40,000 cfs or less

20a

32ab

45b

* Statistics

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

Ef'f'ect of' Constant Flows on Attributes of' a White-Water Trip.
Increases or decreases in constant flows can affect how a guide
manages his or her boat and runs the overall trip. For example, at
very high flow levels guides may be more likely to scout major rapids
before running them. At relatively low flows guides may ask
passengers to break camp early so that they can stay on schedule.
Given these considerations, we asked guides to report how they would
adapt to differing constant flows.

b

56

Rapids are an important attribute of white-water boating trips which
are affected by flows. Constant daily flows affect trip procedures
at major rapids differently for commercial motor, commercial oar and
private trips (see Figure 4-2). Most commercial oar guides report
stopping to scout the major rapids no matter what the flow level. In
contrast, commercial motor guides were more likely to report stopping
to scout major rapids at low flows below 1O,000 cf s and at high flows
- above 50,000 cfs. Private trip leaders were most likely to scout
rapids at moderately high flow levels of 25,000 to 35,000 cfs.
Guides and trip leaders were also more likely to have passengers walk
around major rapids at flows of 10,000 cfs or less and at flows above
35,000 cfs (see Figure 4-3). Commercial oar passengers were.the group
most likely to walk around rapids at all flow levels.
Flow levels can also affect the trip schedule. For example, at low
flows, boat operators may row or run the motor more often to stay on
schedule. As shown in Figure 4-4, commercial guides are more likely
than private trip leaders to attempt to compensate for the speed of
the current at high or low constant flows. Nearly 9 out of 10
commercial guides reported rowing or motoring more at flows of 10,000
cfs or lower, while at high flows (35,000 cfs and above) about 3 out
of 4 commercial guides reported motoring or rowing less.
Stops at attraction sites and time for hiking in side canyons are
important attributes of white-water trips. A serious· i.mplic:.ation of.
low flows is the possibility that both commercial and private trip
passengers may have to miss one or more attraction sites because cf
the additional time needed on the river that is needed to maintain a
trip schedule. Nearly all guides indica fed that certain low constant
flew levels would prohibit stops at certain attraction sites (see
Table 4-5). The minimum flow levels at which white-water boating
trips would not have time for stops at all of their typically
scheduled attraction sites ranged from 8,746 cfs for commercial motor
trips to slightly more than 10,000 cfs for both types of oar power
trips. It is also shown in Table 4-5 that constant high flows in the
30,000 cfs range would allow extra time for scheduled attraction
sites or stops at additional attraction sites.
Similarly, flows cari affect the availability and access to campsites,
another important attribute of white-water boating trips. As
reported in Table 4-6, most guides feel that flow levels do affect
the availability of, and access to campsites. Flow levels at which
it would be difficult to get to campsites on time are similar to
those reported as affecting the time available for attraction sites
and hiking side canyons. In addition, all three groups of guides
indicate that at flow levels above 41,000 cfs, the size and
availability of campsites becomes limited due to high water.

57

Figure

4-2

Proportion of Respondents Who Reported Stopping to
Scout Major Rapids at Constant Flows

1 .0
0.9

T

1---~----x----x---X---X----X----X

0.8

/

0.7
0.6
Proportion
who Stopped
to Scout

·~

0.5
0.4

•

0.3
0.2
0.1
0.0
1-5

5-1 0

1 0-15

15-20

20-25

25-35

35-50

Average Flow Level in cfs
(x 1000)

-t:.-

Commercial
Motor

·X·

Commercial
Oar

•- Private

over 50

58

Figure

4-3

Proportion of Respondents Who Have Passengers Walk
Around Rapids at Constant Flows

1 .0
0.9
0.8
0.7
0.6
Proportion
of Respondents
0.5
Who Have
Passengers Walk
0.4

l
r~x

x

x/

t::i.·

\

x

0.3

•

0.2
0.1
0.0
1 -5

5-1 0

-6-

1 0-1 5
15-20
20-25
25-35
Average Flow Level in cfs
(x 1000)

Commercial
Motor

· x- Commercial
Oar

35-50

·•- Private

over 50

59

.
.
Figure

4-4

Proportlcm of Respondents Who Run Motor or
Row More or Less Than Usual to Compensate
for the River Current at Constant Flows

1 .0
0.9

T

t:,.--~
x~
t,

x

;~·

0.8

x
0.7
Proportion
who
Compensate
for River
Current

xY.
=--·~

0.6
0.5
0.4
0.3
0.2

.---·~-~

•

:~

0.1

Ll

0.0
1-5

5-1 0

1 0-15

15-20

20-25

25-35

35-50

Average Flow Level in cfs
(x 1000)

-t:,.- Commercial
•

Motor

·X·

Commercial
Oar

•- Private

over 50

60

Table 4-5.

Constant Flow Level Requirements 'f'or Use and Access to
Attraction Sites

Fl ow

Commercial
Motor Guides

Percent who feel that certain
flow levels cause problems
with use and access
Mean constant flow below which
there would not be time for
certain attraction sites
Standard Deviation
Sample Size

Private
Trip Leaders

87%ab*

8,746 cfs
3,237

63

~~an constant flow above which
there would be extra time for
attraction sites
29,312 cfs
Standard Deviation
11 '103
Sample Size
64

* Statistics

Commercial
Oar Guides

..

a

10,398 cfs
4,285
49

32,896 cfs
10' 133
48

b

b

1O,156 cf s .
4' 176
115

30,441 cf s
9,392
110

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the·0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

61

"
Table·4-6.

Constant Flow Level Requirements for Access and Availability
of Campsites

..
Flow

Commercial
Motor Guides

Commercial
Oar Guides

Private
Trip Leaders

Percent who feel that certain
flow levels cause problems with
access and use
'tliean constant flow below
which getting to camp on time
is a problem
Standard Deviation
Sample Size
~~an constant flow above
which campsites are limited
Standard Deviation
Sample Size

8' 125 cfs
3,221
56
41,017 cfs
13,640
58

a

9,298 cfsab
3 ,983
47
44,500 cfs
13,902
46

10,025 cfs
4' 642
109
41 ,375 cfs
13,671
112

* Statistics

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

Evaluation of Daily Fluctuations in Fl.ow Level. Prior to the 1983
boating season, daily fluctuations in flow levels were common, often
ranging from 3,000 to 28,000 cfs in a 24-hour period. During the
last three boating seasons, however, a steady high flow of 25,000 cfs
or more has been the most common flow level during the peak summer
boating season. Consequently, not all of the commercial guides or
private trip leaders have experienced daily fluctuations in flow
levels. Two measures of experience were used to identify those
respondents who were qualified to report on the effects _of daily
fluctuations. Respondents who said they could accurately describe
fluctuations and who reported having actually experienced daily
fluctuations of 15,000 cf s or more, were asked to evaluate the
effects of large d~:ily fJuctuaticns.

b

62

·The reported mean tolerable daily change in flow level ranged from
about 3,000 cfs at average daily flow levels of 5,000 - 9,000 cfs to
more than 8,000 cfs at average daily high flow levels of 32,000 cfs
and above (Table 4-7). There were no significant differences in
reported tolerances of daily flow level fluctuations between the
three groups. As the average daily flow level increases, the
reported to~erable range of daily changes also increases, indicating
that at very low flow levels, even small fluctuations make a
difference in the ability to run a white-water raft trip.
Table 4-7.

Reported Mean Tolerable Daily Changes in Fl.ow Levels for
Those Who Have Experienced Daily Fluctuations of at Least
15,000 cfs in the Grand Canyon

Commercial
Motor Guides

Commercial
Oar Guides

Private
Trip Leaders

5,000-9,000 cfs
Mean daily change
Standard deviation
Sample size

3,231 cfs
2,320
52

3,412 cfs
3' 180
34

2,428 cfs
2,233
45

9,000 - 16,000 cfs
Mean daily change
Standard deviation
Sample size

4,706
2,094
52

4,786
2' 311
35

3,936
2,693
47

16,000 - 32,000 cfs
Mean daily change
Standard deviation
Sample size

7' 192
3,404
52

6,708
2,812
36

6,410
3,679
50

32,000 cfs and above
Mean daily change
Standard deviation
Sample size

9,760
5,375
52

7 ,903
2,917
36

8,612
6,470
49

FlOW

The tolerable levels of daily fluctuations reported in Table 4-7 may
represent a "wish" rather than an indication of the largest daily
fluctuations that could actually be ~olerated. Data from informal
interviews and focus groups discussions with commercial guides
suggest that the predictability of the fluctuations, rather than the
range of the daily fluctuation, is the key factor in coping with
daily fluctuations. Evaluations of different flow scenarios later in
this chapter al so support this conclusion.

..

63
•

•

Effect of Daily Fluctuations on Attributes of White-Water Trips. - To
gauge the impact of daily fluctuations on trip procedures,
respondents were given the following hypothetical flow release
pattern to consider:
Assume that flow levels were varying from a low of 3,000 cfs
to a high of 25,000 cfs each day (or within a 24-hour
period). Under these flow conditions, which of the following
would you be likely to do?

Over 60 percent of the commercial oar guides and private trip leaders
indicated that it was very likely they would stop to scout major
rapids more often. Commercial motor guides were less likely to
indicate this (Table 4-8). About half of the respondents in each
group reported they would very likely have to camp above a major
rapid to wait for the water to rise.
Table 4-8.

Impact of Daily Fluctuations in Flow Level on Procedures
at Major Rapids

Procedure
Stop to scout rapids more:
Very likely
Somewhat likely
Not at all likely
Camp above a major rapids to
wait for the water to rise:
Very likely
Somewhat likely
Not at all likely

* Statistics

Proportion Citing Procedure
Commercial
Commercial
Private
Motor Guides
Oar Guides
Trip Leaders

46%a*
40

14

51%
40
9

62%ab
33
5

46%
42

54%
34

12

12

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

64

..
Daily fluctuations appear to have the greatest impact on campsite
selection and the management of rafts at campsites (Table 4-9). Most of
the respondents, especially commercial guides, indicated that they would
select only certain campsites that offered protection against water
level changes. More than 90 percent of the respondents in each group
reported that they would have to check boat moorings during the night to
see if the boat needed to be moved under the daily fluctuations
described.
Table 4-9.

Impact or Daily Fluctuations in Flows on Campsite Selection
and Use

Procedure
Check on boat moorings during
the evening

Proportion Citing Procedure
Commercial
Commercial
Private
Motor Guides
Oar Guides
Trip Leaders
95%

92%

98%

Select certain campsites to
provide protection against
rising water
Spend less time in camp
~ake

camp earlier in the day

* Statistics

sharing a common superscript are not sta~istically
different at the 0.05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

Daily fluctuations in flows might also affect the itinerary of three
out of four commercial and private trips. Almost three-fourths of the
guides and trip leaders reported they would be likely to change their
trip itinerary under the flow release pattern described above in order
to reach certain points at a good time. Respondents reported that; if
anything, they would spend less time rather than more time at
scheduled attractions sites due to fluctuating flows (Table 4-10).

65

Table 4-10.

Impact of' Daily Fluctuations in _Flows on Trip Itineraries

•

Procedure

Proportion Citing Procedure
Commercial
Commercial
Private
Motor Guides
Oar Guides
Trip Leaders

Change trip itinerary

74% *

75%

69%

Spend less time at attractions

42

43

37

Spend more time at attractions

5

12

6

*

Statistics sharing a common superscript are not statistically
different at the 0. 05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when ~tatistical differences are not identified between any
of the three groups of respondents.

Pref'erence Ratings of' Four Flow Release Scenarios. GCES researchers
were given five different flow regimes to evaluate. Each is an
example of a particular type of flow release pattern that satisfies
the legal requirements for releases from Glen Canyon Dam. For
white-water boating, nearly all of which occurs during the months of
May through October, two of the scenarios are identical, so
respondents were asked to evaluate only four scenarios.

The first scenario, Scenario A, represents a constant daily flow
without any fluctuations in the water levels. This scenario was
described to the guides as follows:

.

There would be no daily f'luctuations, but f'lows would change
f'ran one month to the next. Flows during May, June, July and
August would be 10,000, 10,400, 12,750, and 14,400 cf's
respectively, with no daily f'luctuations. The rest of' the
year, f'lows would range f'rom 8,300 cf's to 14,600 cf's, again
with no daily fluctuations.

Scenario B represents a combination of constant daily flows during
the peak rafting season and daily fluctuations in the flow level
during the remainder of the year. The scenario description is as
follows:

66

io

Flows would be constant at 25,000 cf's during June, July, and
August. During the rest of' the year, daily flows could range
f'rom 1,000 to 33,500 cfs.

Scenario C features severe daily fluctuations throughout the year
with the minimum flow level being somewhat higher (3,000 vs. 1,000
cfs) during the prime rafting months of June, July, and August. This
scenario was described in the following manner:
Flows would vary by day, season, and month. During June,
July, and August, daily flows could range :from 3,000 to 33,500
cfs with a major peak at 3:00 p.m. During the rest of the
year, daily flows could range :from 1,000 to 33,500 c:fs.

The final scenario, Scenario D, is comparable to Scenario C in that
fluctuations in daily flows would be the norm throughout the year,
but the daily fluctuations would be moderate throughout the entire .
year. The scenario description is:
Flows could vary by da~, season, and month. Throughout the
year, daily flows could range from. 8,000 to 25,000 c:fs.
During the summer there would be a major peak around 3:00 p.m.

The guides gave Scenarios D and A the highest mean rankings when they
ranked all four scenarios in order of their preference (Table 4-11).
Scenario D was the most preferred by the commercial oar and private
trip leaders, while Scenarios D and A received equal mean ranking
scores from the commercial motor guides.
Respondents were also asked to evaluate the acceptability of each
scenario on a five point scale ranging from completely acceptable to
completely unacceptable. The other points on the scale were somewhat
acceptable_, neutral, and somewhat unacceptable. Scenarios D and A
were the least likely to be rated as unacceptable by respondents in
each of the three groups (Table 4-12). Scenario D was rated as
"Somewhat or Completely Unacceptable" by about one-third of the
commercial oar guides and private trip leader respondents, and less
-than 20 percent of the commercial motor trip guides. The most
unacceptable scenario (C) was rated as "Somewhat or Completely
Unacceptable" by about 9 out of 10 respondents in each of the three
groups. The rankings reported in Table 4-12 correspond closely to
the overall mean rankings presented in Table 4-11.

.

67
•

..

Table 4-11.

Mean Ranking of the Four F1.ow Scenarios

Commercial
Motor Guides

Scenario

D (moderate daily fluctuations)
A (constant daily flows)
B (combination flow pattern)

c

(large daily fluctuations)

1.9 **
1.9a
. 2. 3
3.8

Mean Rankings *
Private
Commercial
Trip Leaders
Oar Guides
1.9b
2.3
2.2
3.7

1.9 b
2. 1a
2.3
3.7

* A score

of 1 was given to the most preferred scenario, while a score
of 4 was given to the least preferred scenario. The mean rankings
were computed averaging the rankings each group of respondents
assigned to a scenario.

**

Statistics sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

Table 4-l2.

Percent Ranking Four Scenarios as Unacceptable

Scenario
D
A
B
C

..

*
**

(moderate daily changes)
(constant daily flows)
(less severe daily changes)
(severe daily changes)

Proportion of Respondents *
Commercial
Commercial
Private
Motor Guides
Oar Guides
Trip Leaders
16"'a **
25,
59
91

34'tb
338.b

53
93

Percent ranking scenario as somewhat or completely unacceptable
on a five-point scale that also included categories of "neutral,"
and "somewhat" or "completely acceptable".
Statistics sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

68

Evaluation of Scenarios. To better understand the guides' preference
ratings of the four flow scenarios, they were given a list of
potential problems and asked to indicate which problems were relevant
to each scenario. Evaluations for each of the scenarios are reported
in.the order in which scenarios are ranked, with the most favorable
scenario (Scenario D - moderate daily changes) described first.

•

Potential problems with Scenario D tended to focus on the daily
fluctuations in the flow level. Fluctuations at camp, difficulty
mooring boats, unpredictable flows and difficulty planning the trip
schedule were the potential problems most often noted with this
scenario (Table 4-13). Commercial motor trip guides were somewhat
less likely than commercial oar guides or private trip leaders to
view these as potential problems with Scenario D.
Table 4-13.

Potential Problems with Scenario D (Moderate Daily
Changes)

Problem

Proportion Citing a Problem
Commercial
Motor Guides

Fluctuations at camp problematic
Difficulty mooring boats
Flow is too unpredictable
Difficult to plan schedule
Problems running rapids
Lowest flows are too low
Unable to avoid other parties
Inadequate flow levels
No flexibility in running trip
Not enough time for attractions
Too much time on the river
Not enough challenge in rapids

* Statistics

72~~

37
38
27a
20a
18a
19
18
15
14
8
oa

Commercial
Oar Guides
76~b
54
50
43ab
'.:19b
39b
24
22
20
20
17 b
4a

Private
Trip Leaders
72~

58
50b
46
29ab
30b
17

16
19
20
15b
8

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

•

69

•

'

Evaluations of Scenario A for the months of May through August
indicate that potential problems focus on the relatively low flow
level (constant daily flows range from 10,000 to 14,400 cfs) during
the major portion of the rafting season (Table 4-14). Not enough
time for stops at attraction sites, inadequate flows, and·too much
time on the river were cited as potential problems with this scenario
by a substantial portion of commercial oar guides and private trip
leaders. Again, commercial motor guides were somewhat less likely to
cite each of these as potential problems, probably due to their
ability to use the motor to make up time.
Table 4-14.

Potential Problems with Scenario A (Constant Daily Flows)
During May through August

Problem

Not enough time for attractions
Inadequate flow levels
Lowest flows are too low
Too much time on the river
Unable to avoid other parties
No flexibility in running trip
Problems running rapids
Difficult to plan schedule
Not enough challenge in rapids
Difficulty mooring boats

Proportion Citing a Problem
Private
Commercial
Commercial
Trip Leaders
Motor Guides
Oar Guides
2ac1a*
16a
19a
24a
26ab
20a·
19
10a
11a
6

54~b
48b
48
39ab
37a
43b
32b
30
19ab
6

- 41 db
36°
31c
40b
21b
26a
23
17a
26b
5

* Statistics

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different
superscripts are statistically different at the 0.05. For example,
i f two statistics both have an "a" in their superscripts, then they
are not statistically different at the 0.05 level. Superscript
notation is not used when statistical differences are not
identified between any of the three groups of respondents .

•

For Scenario B, there would be constant daily flows of 25,000 cfs
from June through August, but during the remainder of the year, flows
could fluctuate daily between 1,000 and 31,500 cfs. For the June
through August period with constant daily flows of 25,000 cfs, no
ootential problems were indicated by more than 11 percent of the
respondents from any of the three groups (Table 4-15).

70

Table JJ-15.

Potential Problems with Scenario B (Combination Fl.ow
Pattern) During June through August

•
Problem

Not enough challenge in rapids
Unable to avoid other parties
Problems running rapids
·Difficulty mooring boats
Difficult to plan schedule
Too much time on the river
No flexibility in running trip
Inadequate flow levels
Lowest flows are too low
Not enough time for attractions

Proportion Citing a Problem
Private
Commercial
Commercial
Motor Guides
Oar Guides
Trip Leaders

11% *
9
1

4%
7
6
4

1
1
1
0

2
2
2

0
0
0

0
0
0

4%
5
2
1
0
0
0
1

1
0

* Statistics

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

For the September through May time period, however, respondents
indicated that there would be a substantial number of problems
associated with daily fluctuations from 1,000 to 31,500 cfs. More
than three out of four respondents indicated that flow levels would
be too unpredictable, fluctuations would cause problems at camp, the
lowest flow levels would be too low, and running rapids would ·be a
problem under this scenario (Table 4-16). Respondents also indicated
difficulty in mooring boats, inadequate flow levels, and the amount
of time they would have to spend on the river would be problems
during the off-season under Scenario B. For problems related to
timing of stops or flexibility in running the trip, commercial motor
guides were less likely to indicate that the severe daily
fluctuations would cause problems, again, probably due to the ability
to use the motor.

•

71

•

-

Table 4-16.

Problem

Potential Problems with Scenario B (Less Severe Daily
Changes) During September through May

Proportion Citing a Problem
Private
Commercial
Commercial
Motor Guides
Oar Guides
Trip Leaders

Flow is too unpredictable
Lowest flows are too low
Fluctuations at camp problematic
Problems running rapids
Inadequate flow levels
Difficulty mooring boats
Difficult to plan schedule
Not enough time for attractions
Too much time on the river
No flexibility in running trip
Unable to avoid other parties
Not enough challenge in rapids

80%a *
84
76a
72
76a
53a
49a
43
38
30a
19
1a

93%b
85b
94
74b
57b
sob
67
48
54b
48
20
6a

91%b
90
83a
74
73a
64a
68b
41
45
32a
19b
21

* Statistics

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different
superscripts are statistically differ·ent at the 0.05. For example,
i f two statistics both have an "a" in their superscripts, then they
are not statistically different at the 0.05 level. Superscript
notation is not used when statistical differences are not
identified between any of the three groups of respondents.

•

Respondents did not feel that the less severe daily fluctuations
during the months of June through August for Scenario C ( a range of
3,000 to 31,500 cfs) would significantly alleviate the potential
problems with this flow scenario. Indications of potential problems
for the two different seasons were nearly identical. Table 4-17
presents the potential problems indicated for the June through August
season, the months with the highest use level for white-water trips.
Responses for September through ~.iay followed a similar pattern with
slightly higher proportions of each group reporting potential
problems. Respondents were most concerned about the effects of the
fluctuations on camps and their impacts on mooring boats and running
rapids. The low flows under Scenario C would also create problems
for raft trips in planning their stops and daily itineraries.

72

Table JJ-17.

Problem

Potential Problems with Scenario C (Severe Daily Changes)
During June through August

Proportion Citing a ProblemPrivate
Commercial
Commercial
Motor Guides
Oar Guides
Trip Leaders

Fluctuations at camp problematic
Flow is too unpredictable
Lowest flows are too low
Difficulty mooring boats
Problems :running :rapids
Difficult to plan schedule·
Inadequate flow levels
No flexibility in :running trip
Not enough time for attractions
Unable to avoid other parties
Too much time on the :river
Not enough challenge in :rapids

85%
80
72 *
59a
66
60
59
44
34ab
32ab
29
9a

91%
91
72
81b
74
74
57
43
50a
44a
43
6a

84%
85
74
68ab
68
66
56
38b
32b
25
36b
19

* Statistics

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of :respondents.

Summary
The optimal flow levels :reported by the commercial guides and private
trip leaders are in the 20,000 to 25,000 cfs :range. The mean minimum
flow level for :running :rapids safely with passengers :ranged from
8,400 cfs for commercial motor guides to 9,200 cfs for private trip
leaders, while the mean maximum safe flow level :ranged from 47,000
cfs for private trip leaders to 59, 000 cfs for commercial motor
guides. Private trip leaders reported the smallest :range of "safe"
flow levels, while commercial motor guides :reported the largest.
Considering all aspects, flows of 20,000 to 30,000 cfs appear to be
optimal for Grand Canyon commercial white-water guides and private
trip leaders. These are the flows at which guides and trip leaders
reported they are least likely to scout :rapids, have passengers walk

•

73

around major rapids, run the motor, and row more often. At these
flow levels campsites are plentiful and there is ample time for side
canyon hikes and stops at attraction sites.
Respondents who had experienced daily fluctuations reported that the
largest tolerable daily fluctuation in flow levels is about 3,000 cfs
with daily flows of 5,000 to 9,000 cfs, and about 8,000 to 9,000 cfs
at higher daily flows averaging 32,000 cfs or more. With ~ release
pattern involving large daily fluctuations, guides indicated they
would scout rapids more often, seek out certain protected campsites,
and change their usual itinerary, including less time at scheduled
attraction sites.
Of the four different flow scenarios outlined by the GCES research
team, the most preferred was one featuring moderate daily
fluctuations in flow levels between 8,000 and 25,000 cfs. This
scenario was slightly more preferred than one featuring moderately
low constant daily flows of 8,300 cfs to 14,600 cfs. The two least
preferred scenarios were a combination flow release pattern featuring
constant daily flows of 25,000 cfs during June through August, and
daily fluctuations from 1,000 - 31,500 cfs during September through
May, and a scenario featuring large daily fluctuations throughout the
·entire year. It should be noted, however, that the constant daily
flow of 25, 000 cfs during June through August un(jer the combination
flow scenario (B) was evaluated most favorably by all three groups.
If these constant daily flows had been extended to include May
through September, this scenario would have been clearly ranked as
most preferred by all three groups of respondents.
In evaluating the four flow scenarios, guides focused on two primary
criteria: the range of daily fluctuations in flow level and the
lower bound of the average .daily flow level during the boating season
of ~~y through October. Both large daily fluctuations in flow levels
and low constant flows wer:e viewed as problematic for running rapids,
selecting campsites and mooring boats, keeping the trip on schedule,
and allowing time for scheduled attraction sites. These problems
were reported to be alleviated by less severe daily fluctuations and
higher average daily flows~

•

74

CHAPTER 5
WHITE-WATER BOATER SURVEYS .ARD RESULTS
Introduction
In this chapter the focus shifts from guides and tr1p leaders to
white-water trip passengers. The results of the white-water boater
attribute and contingent-valuation (CV) surveys are presented.
First, white-water boating on the Colorado River is briefly reviewed
to provide the reader with a perspective for this activity. Sampling
procedures and results of the attribute survey are reviewed next.
These results identify the important attributes of the Grand Canyon
white-water boating experience, and specify those important
attributes that are affected by different releases from Glen Canyon
Dam. This is followed by the results of the CV survey and
respondents' evaluations of white-water boating experiences at
different flows. Respondents' actual trip surplus values, and
surplus values based on responses to six different flow scenarios and
an environmental impact scenario are presented and discussed. Flow
value functions are then derived.
Background
The history of running the Colorado River in the Grand Canyon can be
traced back to 1869, when John Wesley Powell led the first expedition
down the Colorado River through what is now Grand Canyon National
Park. Powell's first expedition, consisting of eight other crew
members (none of them experienced boatmen) and four wooden boats,
actually launched on the Green River in Wyoming. By the time they
reached Lee's Ferry, the starting point for today's Grand Canyon
white-water boating trips, they had already been on the river for
more than 70 days, travelling nearly 600 miles (Crumbo, 1981).
Powell's interests were primarily scientific. Subsequent river
runners were more likely to be trappers, hunters, miners, or
adventurers than scientists. A variety of different river runners
and different types of crafts ran the Colorado River through the
Grand Canyon in the following years, but even as late as 1950, only
about 100 people took the trip (Lavender, 1985).
The first commercial river trips began in 1938, organized by Norman
Nevills, whose Mexican Hat Expeditions was the first commercial
rafting company to operate in the Grand Canyon National Park. The
first motorized trip was conducted in 1949, and, in the 1950's, army
surplus neoprene rafts were introduced as a means to carry a large
number of passengers on a single boat (Stevens, 1983).

--------------~--------

75

•
Today, white-water boating in the Grand Canyon is a $12 million
dollar a year industry, due in part to the construction of the Glen
Canyon Dam (Ridgeway, 1984). Prior to the construction of the dam,
river flows, were dependent upon runoff from snow melting in the
Rocky Mountains, and they varied greatly from low flows of 1,600 cfs
to peak flows approaching 120,000 cfs in the late spring and early
summer. Today, with operation of-Glen Canyon Dam, river flows
typically occur in a much narrower range, from 3,000 cfs to 40,000
cfs, and show less seasonal variation than in pre-dam times. Rafting
occurs during all months of the year, although the great majority of
commercial and private raft trips take place in the months of May
through October.
As no~ed in the preceding chapter, twenty-one companies currently
have permits to conduct commercial raft trips in Grand CanJon
National Park. Each year, approximately 15,000 commercial and
private boaters run the river. Since 1981, restrictions on the
number of white-water boaters have been set by the National Park
Service as a response to rapidly increasing use levels (use had
increased from 547 people per year in 1965 to 16,428 people in 1972)
(Shelby, 1981). Current use restrictions set the number of user days
at 115,500 for commercial trips and 54,450 for private parties.
Motorized trips are currently allowed to launch from mid-May through
mid-September. The season for private and commercial oar-powered
trips is longer.
Commercial white-water boating trips vary on several different
criteria: the size and type of craft; the length of the trip, both
in miles covered and in number of days on the river; and the means of
powering the boat. Boats range from 37-foot G-rigs, consisting of
three smaller rafts lashed together, to single smaller (14-18 foot)
rafts. Power for the larger rafts are supplied by outboard motors
while smaller boats are typically powered by oar. Trips range from a
3-4 day motor trip from Lee's Ferry to Phantom Ranch, a distance of
approximately 88 river miles, to an oar powered trip traversing the
entire 225 mile stretch of river to Diamond Creek, lasting nearly
three weeks.

•

Private (non-commercial) trips running the Colorado River through the
Grand Canyon typically consist of two or more small 18-foot rafts,
although a wide variety of boats including kayaks and canoes often
accompany private trips. Private trips average 14-18 days. Private
permits are issued on a first-come, first-serve reservation basis.
Currently, 223 private party permits are allocated each year, and
applicants may have to wait as long as 3 years to receive one.

76

Attribute Survey Procedures
Sampling. The final attribute survey (see Appendix C) was sent to a
random sample of commercial motor, commercial oar and private trip
passengers selected from the NFS trip launch records for the 1982 and
1984 seasons (April 1 through November 30). Passengers w_ere selected
from three different time periods to ensure that a variety of flow
levels were represented in· the survey. The sampling design called
for the selection of an equal number of passengers from each group
who had experienced high flows (40,000 cfs or more), medium flows
(10,000 - 40,000 cfs), and low flows (less than 10,000 cfs). The
time periods for which passengers were sampled and corresponding
flows are as follows:

May 4 - July 15, 1984
July 24 - September 30, 1984
May 1 - June 30, 1982

High Flow
Medium Flow
Low Flow

The proposed sampling design was to select approxi~ately equal
nUJ!lbers of private boaters, commercial oar passengers, and commercial
motor trip passengers. Commercial outfitters were sent a list of
selected trip dates and asked to provide either names and addresses
of specified passengers (selected at random) or the entire trip
rosters. Commercial trip passengers' names and addresses were
obtained from 19 of the 21 Grand Canyon commercial outfitters. In
cases where the selected passenger number or the trip roster was not
available, ari alternative passenger or trip from a comparable flow
level was substituted. Private trip rosters were obtained from NFS
records of river trips during the 1982 and 1984 seasons. In cases
where a selected private boater's address was not available from Park
Service records or was insufficient for mailing a questionnaire, an
individual from an alternate trip with a comparable flow level was
selected.
In this way, 682 individuals were initially selected to participate
in the study; 2-14 commercial motor passengers, 227 commercial oar
passengers, and 227 private boaters. However, due to language
differences and complications with return postage, individuals who
did not reside in the U.S. were dropped from the sample (12 persons,
1.76 percent). Two other people selected notified us upon their
receipt of the advance letter that they had not actually taken a
rafting trip and were also removed from the sample. These
modifications reduced the overall sample size to 668.
Response Rate. In Yarch, 1985, the sample received their first mail
contact informing them of the attribute survey. Completed
questionnaires were received from 81 percent of the total sample (see
Table 5-1).

77
Table 5-1.

White-Water Boaters' Attribute Survey Response Rate

Completed Surveys
Undeliverable
Surveys not returned
Refusals
TOTALS

*

Percent of
Deliverable Surveys *

Percent
of Total

Surveys

81%
11

91%

7

8
_1
100%

_1
100%

77 surveys were returned as undeliverable. Thus the percentages in
this column are computed from a sample of 591 rather than 668.

Overall, 77 surveys (11 percent) were returned as undeliverable.
Thus, the response rate as a percent of deliverable surveys was 91
percent. The analyses contained in this report are derived from 532
returned surveys: 177 commercial oa1 passengers, 189 commercial motor
1
passengers, and 166 private boaters.Commercial motor passengers
were slightly more likely to respond to the survey, although response
rates for all groups were between 89 and 95 percent (Table 5-2). When
undeliverable surveys are excluded, respondents from 1982 were just as
likely to r·eturn a completed surveys as those from 1984 trips.

Table 5-2.

•

*

White-Water Boater.Attribute Survey Response Rates by
Trip Type and Year

Year

Commercial
Oar Trip

1982
1984

88%
90

Commercial
Motor Trip

95%
94

Private
Trip

86%
92

Response rates calculated as percentages of all deliverable surveys .

ll Six additional surveys were returned after the data analyses were
completed. While these six are included in the response rates
reported in Table 5-1, they are not included in the results reported
in the remainder of this text.

78

It

Attribute Survey Results
Trip Attributes. White-water boaters were first asked, in an
open-ended question format, to report the attributes-they felt would
contribute most to an excellent or perfect Grand Canyon trip.
Answers to this question are summarized in Table 5-3. Good weather,
good social interaction, good guides, an unrushed pace (time for
layovers and stops at attraction sites), and a wilderness experience
were the attributes mentioned most often by the respondents as
contributing to an excellent or perfect trip.
Table 5-3.

Attributes that Contribute Most to
Perf'ect Grand Canyon Raf't Trip

Attribute

An

Excellent or

Proportion Citing Attribute
Commercial
Commercial
Private
Oar
Motor
Boaters

Good weather
Good social interaction
Good guides
Unrushed pace/more layovers
Wilderness experience
Well conducted trip
Good food
Good/exciting rapids
Being in the Grand Canyon
No crowding

* Statistics

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different
superscripts are statistically different at the 0.05. For example,
if two statistics both have an "a" in their superscripts, then they
are not statistically different at the 0.05 level. Superscript
notation is not used when statistical differences are not
identified between any of the three groups of respondents.

Among the attributes listed by 15 percent or more of all respondents,
four are potentially affected by flow levels: time for layovers and
stops at attraction sites, good/exciting rapids, a wilderness
experience, and not feeling crowded. These four attributes are

.

79

related to flow levels in a variety of ways. As reported in Chapter
4, commercial guides and private trip leaders report that the amount
of time for stops at attraction sites, hiking side canyons, and
layovers is significantly reduced at relatively low flows.
Rapids are also flow-related since a number of snall to medium ra~ids
become "washed out" at relatively high flows, while other larger
rapids become more problematic and exciting to run. The feeling of
being in a wilderness area can be affected by fluctuations in daily
flows since changes in flow releases from the dam would have an
obvious visible effect on the recreation environment. Finally,
white-water boaters may feel more crowded at high flows because the
number and size of beaches for camping are significantly reduced. In
addition, during conditions of daily fluctuations in flows, boaters
may become congregated above rapids as they wait for the water level
to rise.
For these four potentially flow-sensitive attributes it is
interesting to note that "good/exciting rapids" is somewhat less
important to private boaters than it is to commercial passengers. On
the other hand, enjoying "a wilderness experience" and "not feeling
crowded" are less important attributes for commercial motor
passengers. These findings would.appear to be related to the types
of trips taken by commercial passengers and private boaters and their
expectations.
Respondents werE? al so asked to identify the attributes they felt
would contribute most to a poor Grand Canyon boating experience. A
question with open-ended response categories was used once again, and
the attributes cited most often were crowding and bad weather (see
Table 5-4). Among the attributes identified by respondents,
crowding, litter, unsafe conditions and low water levels are
potentially affected by flows.
We have already addressed why "crowding" can be a flow-sensitive
attribute. Litter can be related to flows in that high flows and
fluctuating flows with a high upper bound have a tendency to
"flush-out" the canyon, removing debris along the shore, at camping
beaches, and at attraction sites. Flows affect safety primarily
through their effect on the size of rapids, and the difficulty of
running rapids. At high flows; as was previously noted, some rapids
become washed out while others become more problematic to run. In
addition, rapids which are washed out at high flows may be difficult
te run at relatively low flows. "Low water", as revealed by the
results of the guides survey, can affect the ability of trip leaders
to keep the expeditions on schedule and it can also cause problems
with mooring boats at campsites.

80

Table 5-4.

Attributes that Contribute Most to A Poor Grand Canyon
Raf't Trip

..
Attribute

Proportion Citing Attribute
Commercial
Commercial
Private
Oar
Motor
Boaters

Crowding
Bad Weather
Poor guides
Poor Social interaction
Litter
Unsafe conditions
Low water level

* Statistics

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different
superscripts are statistically different at the 0.05. For example,
if two statistics both have an "a" in their superscripts, then they
are not statistically different at the 0.05 level. Superscript_
notation is not used when statistical differences are not
identified between any of the three groups of respondents.

Since commercial trip passengers and private boaters generally have a
limited amount of actual experience with different flow levels,
respondents were also asked to rate a number of attributes of a Grand
Canyon white-water raft trip specified by the researchers (Table 5-5).
The most important attributes in Table 5-5, as designated by the
overall ratings, correspond closely to respondents' answers to the
open-ended attribute questions. In addition, there is substantial
agreement among the three types of boaters on the important attributes
of a Grand Canyon raft trip. Of the ten attributes listed, five may
be affected by flow levels: being in a natural setting, stopping at
side canyons or creeks, hiking in side canyons, large rapids, and
observing wildlife.

81
•

Table 5-5.

Attribute

Importance Ratings of" Grand Canyon Raf"t Trip Attributes

Overall Ratings*
Private
Commercial
Commercial
Oar
Motor
Boaters

Being in a natural setting
Stopping at side canyons or creeks
Relaxing, getting away from it all
Hiking in the side canyons
Confidence in my guide or trip leader
Large rapids
Observing flora, fauna, and geology
Being on the Colorado River
Learning about the history of the
Grand Canyon
Seeing Wildlife

3.ob

2.9a
2.8

2.1

2.7a
2.8a

2.1

2.7a
2.6
2.7b
2.5

b

2.9
2.8b
2.8b
2.5
2.6b

2.1

2.6
2.6ab
2.6

* Overall ratings are calculated by assigning values of 1 through 3
to responses of "not important" through "very important",
respectively, and computing weighted averages.

**

.

Statistics sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

Attributes of" Rapids. We presumed at the outset of this study that
rapids would be an important attribute. Thus, we wanted to identify
the specific aspects of rapids that could increase or decrease
respondents' enjoyment of their trips. To accomplish this objective,
respondents were provided with a list of 12 attributes that are
specific to rapids and they were asked to indicate whether each one
"increased," "decreased," or "had no effect" on their enjoyment. The
five most important and the five least important attributes from this
list are reported in Table 5-6. The numbers reported here record the
percentages of respondents indicating that the specific attribute could
either increase or decrease their enjoyment of a Grand Canyon
white-water trip.

82
Table 5-6.

Important Attributes of Rapids Increasing or Decreasing
Enjoyment by User Group

Attribute

Citing Attribute
Commercial
Private
Motor
Boaters

Pro~ortion

Commercial
Oar
Increasing Enjoyment:
Rapid with large waves
Roller coaster ride
Long rapid
Large number of rapids
Learning how to "read" rapids
from the guide or trip leader
Decreasing Enjoyment:
Having to walk around a rapid
Waiting at a rapid for other
trips to run it
Fear of falling out of boat and
being in the water for long time
Concern a bout damage to
personal equipment
Rocks sticking out of water

91
92
91

90l·

95~b
95
94
92

946
98
89
87

87a

85a

76b

81%

83%

83%

48

51

47

45a

33b

50a

36a
31

26b
25

51c
29

* Statistics

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different
superscripts are statistically different at the 0.05. For example,
if two statistics both have an "a" in their superscripts, then they
are not statistically different at the 0.05 level. Superscript
notation is not used when statistical differences are not
identified between any of the three groups of respondents.

The relative rankings of both positive and negative attributes are
quite consistent across the three groups of respondents. Large waves,
a roller coaster ride through rapids, and length and the number of
rapids were cited most often as increasing enjoyment. Having to walk
around a rapid was noted most often as decreasing enjoyment.
Private boaters were less likely to indicate that learning how to
"read" rapids from a guide or trip leader increased their enjoyment.
Private boaters were more concerned, however, with the chances of
falling out of the boat than were commercial motor passengers, and
they were also more concerned about damage to personal equipment than
were commercial passengers.

•

..

83

Attributes of Campsites. Campsites were also presumed to be an
important attribute. To identify the aspects of campsites that are
important, respondents were asked to evaluate a list of 26
campsite-specific attributes and to indicate whether they felt each
attribute was "very important," "somewhat important," "not important,"
or "had no effect." The ten most important attributes of campsites
are listed in Table 5-7. The relative rankings of attributes are
similar across the three groups, and clean campsites and a natural
appearance are the most important attributes.
Table 5-7.

Rating of Attributes of Campsites

Attribute

Clean, unlittered campsites
Natural appearance
Scenic view
Isolation from other groups
Side canyon for hiking
Few flies or mosquitoes
Nearness to river
Clear water in side canyons
Place to dock boats
Flat area for sleeping

Overall Ratings*
Private
Commercial
Commercial
Oar
Boaters
Motor

3.0

3.0

2.9 H
2.7a
2.7a
2.6a
2.6
2.5
2.5
2.4a
2.5

2.9
a
2.7b
2.6b
2.5
2.6
2.5
2.5
2.4a
2.4

* Overall ratings are calculated by assigning values of 1 through 3
to responses of "not important" through "very important,"
respectively and computing weighted averages.
** Statistics sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
~ statistically different at the 0.05.
For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

Sensitivity of Selected Attributes to Flows. The sensitivity of four
selected attributes to flow levels were also examined by grouping
respondents according to the average flow levels they actually
experienced. The average flow levels were categorized as: low (less

84

.
than 16,000 cfs), medium (16,000 to 32,000 cfs), and .h!.E;h (greater than
32,000 cfs). Responses to items measuring walking around rapids,
amount of time for hiking and attraction sites, reported crowding on
the river, and reported crowding at campsites were examined by trip
type and flow category. Commercial oar passengers were the most likely
group to report having to walk around a rapid at any flow level, while
commercial motor passengers were the least likely (Table 5-8). For
commercial oar and private trip respondents, the probability of having
to walk around a rapid increases as flow level increases. For
commercial motor trip passengers, however, there is no relationship
with flow level.
Table 5-8.

Respondents Va1king Around Rapids by Flow Level
Experienced

Trip Type

Commercial Oar

Proportion of Respondents Walking Around
Rapids
High Flow
Low Flow
Medium Flow
16%a*

36%b

47%b

Commercial Motor

3

0

0

Private Boater

5a

6a

26b

* Statistics

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different
superscripts are statistically different at the O. 05. For
example, if two statistics both have an "a" in their superscripts,
then they are not statistically different at the 0.05 level.
Superscript notation is not used when statistical differences are
not identified between any of the three groups of respondents.

Respondents' evaluations of the adequacy of time available for hiking
side canyons and stopping at attraction sites were not affected by
the average daily flow level experienced. Most commercial trip
passengers felt they had enough time, but only 1 out of 2 private
boaters felt they had enough time (Table 5-9). This is probably due,
in part, to a difference in expectations since private boaters
generally have more experience in the canyon and may be more aware of
attraction sites and hiking opportunities than are commercial
passengers.

85

•
Table 5-9.

Amount of Time for Hild.ng and Attraction Sites

Response

Commercial
Oar

Commercial
Motor

Private
Boaters

There was enough time

81%a*

69%b

52%c

There was not enough time

18a

30b

47c

1

1

1

There was too much time

* Statistics

sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different
superscripts are statistically different at the 0.05. For example,
if two statistics both have an "a" in their superscripts, then they
are not statistically different at the 0.05 level. Superscript
notation is not used when statistical differences are not
identified between any of the three groups of respondents.

Respondents were also asked to indicate on a nine-point scale how
crowded they felt the river was on their trip. Responses to this
question are summarized in Table 5-10. Commercial motor and oar
passengers' perceived level of crowding did not differ by flow
levels. Private passengers, however, felt more crowded at low and
high flows than at moderate flows.
Since crowding can manifest itself at camping beaches, especially at
high flow levels, we also asked respondents whether they felt crowded
at campsites during their trip. Responses to this question reveal
that private passengers who experienced high flow levels were most
likely to perceive crowding at campsites (Table 5-11).

Table 5-10.

Reported Crowding by Flow Level Experienced

ProDOrtion Feeling Crowded

Level of
Crowding

Commercial Oar

Commercial Motor

Private Boaters

Low

Medium

High

Low

Medium

High

Low

Not at all crowded
( 1, 2)

39%

33%

39%

44%

50%

45%

29%a

50%b

19%c

Slightly crowded
(3,4)

37

40

47

40a

36ab

24b

35

30

33

Moderately crowded
(5,6, 7)

23ab•

27a

10b

15a

11a

31b

35a

18b

46a

1

0

4

1

3

0

1

2

2

3.3

3.4

3.2

3 .1

2.9

3.3

3.aa

3.ob

4.4a

Extremely crowded
(8, 9)
Mean Rating

•

•

Medium

High

Mean ratings calculated by assigning values of 1 through 9 to responses of
"Not at all crowded" through "Extremely crowded", respectively, and
calculating a weighted average.

•• Statistics sharing a common superscript are not statistically different
at the 0.05 level. Statistics with different superscripts are
statistically different at the 0.05. For example, i f two statistics both
have an "a" in their superscripts, then they are not statistically
different at the 0. 05 level. Superscript notation is not used when
statistical differences are not identified between any of the three groups
of respondents •

..

"

00

(j'I

~

•

87

..
Table 5-11.

Reported Crowding at Campsites by Flow Level Experienced

Proportion Feeling Crowded
High Flow
Low Flow
Medium Flow
Commercial Oar

30%

40%

39%

Commercial Motor

28

24

40

Private Boater

38a*

37a

70b

* Statistics sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different
superscripts are statistically different at the 0.05. For example,
if two statistics both have an "a" in their superscripts, then they
are not statistically different at the 0.05 level. Superscript
notation is not used when statistical differences are not
identified between any of the three groups of respondents.

Respondent Perceptions of Flow Levels. Given a choice of flow
levels, about one-half of the respondents indicated they preferred
the flow level to be about the same as it was during their trip.
Examination of the responses of those who expressed a preference for
a flow different than they actually experienced, however, reveals a
pattern favoring medium flow levels (Table 5-12). That is, those
experiencing a low flow were more likely to prefer higher flows, and
the converse holds for those who experienced high flows. Those who
experienced medium flows and expressed a preference for a different
flow were evenly split regarding their preferences for a higher or
lower flow.

With respect to fluctuations in daily flow levels, substantial
fluctuations in excess of 12,700 cfs are required before nearly all
respondents' reported noticing changes in the water level.
Respondents' awareness of flow fluctuations are reported in Table
5-13. These responses are categorized by trip type and the level of
fluctuation actually experienced. Overall, private boaters were the
most sensitive to fluctuations and commercial motor passengers were
the least sensitive.

Table 5-12.

Flow Level Pref'erences by Flow Level Experienced

Preferred Flow

Proportion Preferring Specified Flow Level
Low

Commercial Oar
High
Medium

O%a*

Lower

13%b

23%b
53

Same

47

56

Higher

36a

13b

Don't Know/Care

17

18

6b
18

Commercial Motor
Low Medium High
7%a

8%a

30%b

Private Boaters
Low Medium Hil<:h
5%a

10%a

50%b

50

52

35

49ab

66b

40a

23

16

15

34a

12b

2b

20

24

20

12

12

8

* Statistics sharing a common superscript are not statistically different at the
0.05 level. Statistics with different superscripts are statistically different at
the 0.05. For example, if two statistics both have an "a" in their superscripts,
then they are not statistically different at th.e 0.05 level. Superscript notation
is not used when statistical differences are not identified between any of the
three groups of respondents.
CD
CD

..

•·

..

•

89

•
Table 5-13.

Respondents Reporting Awareness of Changes In Flow Level

•
Average Daily Fluctuations Experienced
Less than
12,700 cfs
2,5002,500 cfs
12,700 cfs
or more
Commercial Oar
Commercial Motor
Private Boaters

37~a*

65~b

18
59a

65
76a

85Gc
79b
98

* Statistics sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different
superscripts are statistically different at the 0.05. For example,
if two statistics both have an "a" in their superscripts, then they
are not statistically different at the 0.05 level. Superscript
notation is not used when statistical differences are not
identified between any of the three groups of respondents.

The majority of respondents felt that daily fluctuations in flow
levels would make the river seem "somewhat less" or "much less" like
a natural setting (Table 5-14), although nearly one in four
respondents felt that fluctuations "would not have any effects," and
11 percent overall did not know what the effect would be. Private
boaters were much more sensitive than commercial passengers to
fluctuations, with nearly three out of four saying that fluctuations
would make the setting seem less natural.
Contingent-Valuation Survey Procedures
The White-Water Boaters' Contingent-Valuation Survey (see
Appendices D and E) was sent to a sample of 598 individuals who took
a Grand Canyon trip during the 1985 rafting season (February 26
through November 6). Names and addresses for these individuals were
obtained from National Park Service launch records and commercial
outfitters. The sample was stratified into three user groups: 1)
passengers from commercial oar trips; 2) passengers from commercial
motor trips; and 3) individuals who took private trips. Surveys were
sent to 195 commercial oar passengers, 191 commercial motor
passengers, and 212 private boaters.
Sampling.

..

90
Table 5-111.

Evaluation or Efrects or Daily Fluctuations
Perceptions or a Batural Setting

Commercial
Oar

Evaluation of
Fluctuations
Much more like a natural setting
Somewhat more like a natural setting
Wouldn't have any effect
Somewhat less like a natural -setting
Much less like a natural setting
Don't know

2%
7ab*
24
25
28a
14a

•

OD

Commercial
Motor
3%
6
28a
25
22a
16a

Private
Boaters

•

1%
4
19b
22b
52b
2

* Statistics

sharing a common superscript are not statistically different
at the 0.05 level. Statistics with different superscripts are
statistically different at the 0.05. For example, if two statistics
both have an "a" in their superscripts, then they are not statistically
different at the 0.05 level. Superscript notation is not used when
statistical differences are not identified between any of the three
groups of respondents.

The flow levels experienced by respondents for their 1985 trips are
recorded in Table 5-15. The average flow levels for private boaters are
somewhat lower than those of commercial passengers, due to the extended
private rafting season, and the fact that a larger percentage of commercial
passengers take their trips during the months of June through August when
average flow levels are generally higher. The highest average daily flow
level experienced by commercial passengers was 44,400 cfs and the low was
10,500 cfs. The comparable high and low flows for private boaters were
43,200 cfs and 10,700 cfs, respectively.
Table 5-15.

Daily Flow Levels Experienced by Respondents to the
Grand Canyon Boater Contingent-Valuation Survey

Flow

Commercial
Passengers

Private
Boaters

Average Flow

28 '900 cfs

26,000 cfs

Average High Flow

31,600

29 ,200

Average Low Flow

252200

21 2800

.

91

"

..

Response Rate. The CV surveys were mailed out in May, 1986.
overall, 508 usable questionnaires were returned (87 percent of the
total sample). Eleven additional questionnaires were returned but
were not used because the respondents were under 18 years old. The
response rate as a percent of all deliverable questionnaires was 91
percent (Table 5-16). The results presented in this report are based
on the responses of 506 Grand Canyon boat trip passengers: 170
commercial oar p~~sengers, 167 commercial motor passengers, and 169
private boaters..
Table 5-16.

Vhite-vater Boater Contingent-Valuation Survey Response Rate

Surveys
Completed surveys
Undeliverable *
Not applicable
Surveys not returned
Refusals
TOTALS

Percent of
All Surveys
87%
4

Percent of
**
Deliverable Surveys
91%

***

1

8

9

_o

_Q

100%

100%

*This includes 11 questionnaires returned but not included in the
data analysis since the respondents were less than 18 years old.
** The percentages in this column are computed from a sample size
of 560 rather than 598. The undeliverable and not applicable
surveys have been excluded.
*** Two of these surveys were returned after the data analysis was
completed.

Contingent-Valuation Survey Results

..

Actual Trip. Respondents spent a relatively large amount of money,
on average, for their Grand Canyon white-water trips. Average
reported total expenditures ranged from about $557 for individuals on
private trips to roughly $1,406 for i:;assengers on commercial trips.

l:/ Two surveys were received after the data analyses for this report
were completed. While these two are included in the response
rates reported in Table 5-16, they are not included in the results
reported in the remainder of this report.

92

These differences in total expenditures are primarily due to payments
to commercial rafting companies and greater expenditures for
transportation to the Grand Canyon on the part of commercial trip
passengers.
Respondents also placed a substantial surplus value on their actual
trip, above and beyond their actual expenditures. We found that
surplus values vary with the average flo~/level experienced as well
as type of trip (commercial or private).
Commercial passenger
surplus values for constant flows rise from $47 per trip at 1,000 cfs
to a maximuw of $898 at 33,000 cfs, and then decline to $732 at
45,000 cfs.-1 Private boaters' constant flow surplus values follow
a similar pattern, rising from $21 per trip at an average flow of
1,000 cfs to a maximum of $688 at 29,000 cfs, and then declining to
$376 at 45,000 cfs. Thus, surplus values of commercial passengers
are higher than thos5 for private passengers at all constant flow
levels (Figure 5-1).-1 This is not surprising, however, since
their actual expenditures per trip are also higher.
The magnitude of the difference between private boater and commercial
passenger surplus values increases from $26 at 1,000 cfs to $356 at
an average flow of 45,000 cfs.
The optimum flow level for commercial passengers, as indicated by
surplus values, occurs at 33,000 cfs. This is 4,000 cfs higher than
the optimum flow level for private boaters (29,000 cfs), and private
boater surplus values decline faster at flow levels above their
optimum than do those for commercial passengers. These results may
be due to the fact that commercial passengers have a professional

3/ The statistical results from the analysis of each of the
contingent-valuation questions (for the actual trip and the
seven scenarios) are reported in Appendix L.
~/

The average flow levels experienced by respondents for their
actual trips ranged from 10,500 to 44,400 cfs. Thus, we
extrapolated the flow value function for white-water boaters down
to 1,000 cfs to cover the full range of constant flow levels the
GCES research teams were asked to evaluate.

2./ The values plotted in Figure 5-1 and summarized above can be
thought of as surplus values associated with constant flow
levels. The average daily fluctuation experienced by respondents
was 6,700 cfs, which is probably not readily noticeable to most
white-water boaters. Only 12 percent of the respondents
experiencing daily fluctuations in excess of 10,000 cfs, so we did
not have a sufficient number of these types of observations to
make any inferences about fluctuating flow surplus values from the
actual trip data.

93

•
Figure

•

5-1

Relationship Between Surplus Values and
Flow Levels for Respondents' Actual Trip
($ Per Trip)

$900
800
700
600
Surplus
Value
Per Trip

500
400
300
200
100
0
0

5

10

20

15

25

30

35

40

45

Average Flow Level in cfs
(x 1000)

·•- Private Boater Values

.

.[] • Commercial Passenger Values

50

94

guide who is familiar with the river and, as a result, they can enjoy
the larger "roller coaster" ride through rapids at high flow levels
without the concerns of handling a boat. Private boaters, on the
other hand, may have to give more consideration to the problem of
maneuvering their boats through rapids at higher flows since they
typically do not have the experience with Grand Canyon boating that
commercial guides do.
Several other types of variables were also examined to determine if
they significantly affected actual trip surplus values. The type of
boat used (motor raft, oar raft, paddle raft, kayak, or dory) did not
affect surplus values. Private boater surplus values, however, were
significantly reduced if they had to share beaches for camping or
felt crowded on the river. These variables did not significantly·
affect surplus values for commercial passengers. This result could
be due to differing expectations between these two groups.
For a full discussion of the variables analyzed in conjunction with
the actual trip valuation question, the reader should refer to
Appendix L.
Scenarios. Respondents were asked to evaluate a total of six
different flow scenarios as well as one environmental impact scenario
(see Table 5-17). As with the analysis for the actual trip valuation
question, several variables were tested to determine if they had a
statistically significant effect on surplus values. The size of the
increase in trip expenses presented to respondents had a significant
effect on surplus values at the 0.10 level or better in all seven
cases. The variable representing actual trip expenditures had a
significant effect at the 0.10 level in five of the seven cases and,
once again, there was a positive relationship between expenditures
and surplus values, a result that was unanticipated. We found no
evidence of strategic bias or hypothetical bias in respondents'
answers to the scenario valuation questions. A full description of
the analyses of responses to the flow scenario CV questions is
contained in Appendix L.

With so many scenarios, one concern was whether respondents' answers
to the associated valuation questions might be affected by the order
in which the scenarios were presented. Therefore, the sample was
randomly split in half and each group received a different sequence
of scenarios in their questionnaires. All questionnaires had the
actual trip question first. Half then began with the 5,000 cfs
constant flow scenario and moved up through successively higher flows
with constant flows always preceding the corresponding fluctuating
flows. For example, the 5,000 cfs constant flow scenario preceded
the 5,000 cfs average flow with daily fluctuations scenario. The
environmental impact scenario, which postulated a significant
reduction in beaches over time, was the last scenario valued.

95

•

For the other group, the order of the scenarios was reversed, except
that for purposes of scenario wording, the constant flow scenarios
had to precede the corresponding fluctuating flow scenarios at each
level. Statistical analysis showed no effect of the scenario order
on surplus values.
The surplus values for each of the scenarios are summarized in Table
5-17. For low flows, large fluctuations around an average flow of
5,000 cfs significantly increased surplus values, relative to a
constant flow, in the case of commercial passengers, but did not
affect private boater values. At a moderate flow of 22,000 cfs,
large daily fluctuations significantly reduced surplus values for
both groups of respondents. A high constant flow of 40,000 cfs
produced lower surplus values than moderate constant flows of 13,000
and 22,000 cfs, but were higher than those for a low constant flow of
5,000 cfs. Finally, a substantial reduction in the number of sand
beaches for camping would substantially reduce surplus values, and
only the 5,000 cfs scenarios recorded lower surplus values. We will
briefly discuss each of these values.

Table 5-17.

Estimated Scenario Surplus Values :f'or White-Water Boaters

($

PER TRIP)

Commercial
Passengers

Private
Boaters

$176

$233

5,000 cfs with fluctuations

226

241

13,000 cfs

488

504

22,000 cfs

602

525

22,000 cfs with fluctuations

467

384

40,000 cfs

439

434

Beaches reduced

413

377

Scenario
5,000 cfs

96

Case 1 - Constant now of" 5,000 cf's.
survey as follows:

This flow was described in the

At a constant f'low of 5,000 cf's, the speed of" the river is
relatively slow, reducing time f"or side canyon visits and
other attractions. Boaters •ust break camp early to stay
on schedule. Al though rapids are present at this low
water level, the waves are mal.ler and do not produce the
big "roller coaster• ride created by higher f'lows. Due to
exposed rocks, some rapids aay be so dif"f"icult that it is
likely passengers would have to walk arotmd them.
However, ca11ping opporttmities are abundant with many
large sandy beaches exposed.
Over 90 percent of all respondents felt that this scenario
represented a trip that would be worse than the one that they
actually experienced (Table 5-18). This feeling is represented by
the surplus values of $233 and $176 per trip that private boaters and
commercial passengers, respectively, assigned to this scenario.
These values are substantially less than the maximum surplus values
for these two groups derived from the actual trip data.
Table 5-18.

Rating of" the 5,000 cf's Constant now Scenario Relative
to Actual Trip

Rating

Better
About the Same
Worse

Proportion of Respondents Citing Rating
Private
Commercial
Passengers
Boaters

0%
4

96

4%
3
93

Case 2 - Average now of" 5,000 cf's with Daily nuctuations.
scenario description was as follows:

The

With f"lows f"luctuating daily f"rom 1,000 to 17,000 cf"s, arotmd
an average daily now of" 5,000 cf's, most people are aware of"
changes in the water level. Trip speed is relatively slow,
reducing time f"or side canyon visits, and boaters must break
camp early to stay on schedule. Large sandy beaches are
generally abtmdant, but boatmen must take care selecting
mooring sites. Occasionally, due to low water in the
morning, gear will have to be carried a long ways (perhaps

•

97

•

across slippery rocks) to be loaded on the boats. Boataen
may have to wait above certain rapids tor the water to rise,
or hurry to get to a rapid before the water falls. Due to
exposed rocks, s0111e rapids may be so difficult that it is
likely passengers would have to walk around them. At other
rapids, however, higher flows may produce large waves and a
bigger •roller coaster• ride than at a low constant flow.

Respondents were first asked whether they would prefer a trip with
low water and large daily fluctuations, as described above, or low
water with small daily fluctuations. Given these two alternatives,
private boaters were more likely to prefer low water with small
fluctuations, while commercial passengers indicated a preference for
low water with large fluctuations (Table 5-19). These findings
appear to be consistent with the type of trips each group
experienced. Commercial passengers do not need to be concerned with
the management of a boat so that they may enjoy the large
fluctuations at a low flow level because they can get a bigger
"roller coaster" ride when rapids are reached at the high end of the
fluctuation. On the other hand, private boaters must consider the
effect that fluctuations have on their trip schedule and the care of
their boat (s).
Table 5-19.

Respondents' Preferences tor Fluctuations at Low Flow
Levels

Preference
Low water/small fluctuations
Low water/large fluctuations
Makes no difference

..

Proportion of Respondents
Stating Preference
Commercial
Private
Passengers
Boaters

49%

30%
60

42

10

9

Commercial passengers' preferences for low water with large daily
fluctuations relative to constant low flows is consistent with the
surplus value they assigned to this scenario. The surplus value
commercial passengers assigned to this scenario is $226 per trip.
This is significantly different from the value of $176 they placed on
the 5, 000 cfs constant flow scenario. The surplus value for private

98

boaters is $241 per trip, which is not statistically different from
the surplus value 7eported for the 5,000 cfs constant flow scenario
6
of $233 per trip.Case 3 - Constant Flow of 13,000 cfs.
evaluated the following experience:

In this scenario respondents

At •oderate water levels (around 13,000 cfs), the pace of
the river is slightly faster than at lov flows, leaving a
little •ore tille for hild.ng in side canyons and stops at
attractions. Most boating groups will not have a problem
staying on schedule. Rapids tend to have larger waves and
provide a little •ore of a •roller coaster• ride than at low
water. Passengers may have to walk around only a few
rapids. Campsites are still large and plentiful.

Most respondents
worse than their
surprising given
most respondents
derived from the

Table 5-20.

indicated that this trip would be about the same or
actual trip (Table 5-20). This result is not
that 13,000 cfs is considerably below the flow level
experienced as well as the optimum fl ow 1 evel s
actual trip valuation data.

Bating of The 13,000 cfs Constant Flow Scenario Relative
to Actual Trip

Rating
Better
About the Same
Worse

Proportion of Respondents
Citing Ratings
Commercial
Private
Passengers
Boaters

15%
32
53

25%
36
39

~/The Chi-square statistics for these tests are 7.46 and 0.08,
respectively, with two degrees of freedom. These statistics indicate
that the null hypothesis of no difference can be rejected at the 0.10
level for commercial passengers, but cannot be rejected for private
boaters.

•

99
...

•

The surplus values assigned to this scenario are $504 per trip for
private boaters and $488 per trip for commercial passengers. These
values are both significantly larger than the respective f}1r'Plus
values reported for the 5,000 cfs constant flow scenario.
Case 4 - Constant Flow of 22,000 ct"s. Case 4 is another constant
flow scenario which was described as follows:
At moderately high water levels (around 22,000 ct"s),
the pace of the river is faster than at lower flows,
leaving more time for side canyons and stops at
attractions. Boating groups do not have a problem
staying on schedule. Rapids have larger waves and
provide a bigger •roller coaster• ride than at moderate
water. Only a few passengers choose to walk around
some of the bigger rapids for their safety. Some
potential campsites are under water in some areas of
the canyon, but generally campsites are plentiful
although a bit smaller in size.
A majority of the respondents felt that a trip under these conditions
would be about the same as the trip they actually experienced (Tabl"e
5-21). This is to be expected since this scenario comes the closest
to describing the actual flow levels experienced by most of the
respondents.
Table 5-21.

Rating of The 22,000 cfs Constant Flow Scenario Relative
to Actual Trip

Rating
Better
About the Same
Worse

Proportion of Respondents
Citing Ratings
Commercial
Private
Passengers
Boaters
22%
67

30%
66

11

1

1/ The Chi-square statistics for these tests are 88.42 for commercial
passengers and 31.60 for private boater's with two degrees of freedom,
indicating that the null hypothesis can be rejected at the 0.10 level.

100

Overall, private boaters assigned a surplus value of $525 per trip to
this scenario, while the surplus value for commercial passengers is
$602 per trip. The surplus value for commercial passengers is
statistically larger than the respective value reported for the
13,000 cfs constant flow scenario of $488. The same comparison for
private boaters did not reveal a significant difference in surplus
values g;tween the 13,000 cfs and 22,000 cfs constant flow scenario
values.Case 5 - Average Flow oC 22,000 cf's With Fluctuations. Case 5 is
similar to Case 4 except that fluctuations were introduced. This
scenario was described in the following manner:
With large daily fiuctuations Crom 10,000 cf's - 31,500 cf's,
around an average daily now oC 22,000 cCs, most people are
aware oC water level changes. The boatmen will have to take
more care in selecting mooring and camping sites. Due to
low water levels in the morning, gear may have to be carried
(perhaps across rocky areas) to be loaded on the boats.
Boatmen may decide to wait above certain rapids Cor the
water level to rise or may have to hurry to get to a certain
rapid beCore the water level Calls. In addition, sane ·
rapids may be diCCicult due to exposed rocks at low water
levels and other rapids might be quite large at high water
levels, and it is likely that passengers may have to walk
around a Cev rapids. When the water is high or rising,
however, the standing waves in some oC the major rapids
become larger, resulting in a bigger •roller coaster• ride.
The majority of respondents, regardless of trip type, said they would
prefer to experience moderately high water with small fluctuations
rather than moderately high water with large fluctuations (Table
5-22) •

.§_/ The Chi-square statistics for these tests are 8.08 and 0.90,
respectively, with two degrees of freedom. These statistics indicate
that the null hypothesis of no difference can be rejected at the 0.10
level for commercial passengers and cannot be rejected for private
boaters.

•

101

..
Table 5-22.

Respondents' Preferences ror Fluctuations at Moderately
High Flow Levels

•

Preference

Proportion of Respondents
Citing Preference
Commercial
Private
Passengers
Boaters

Moderately high water/
small fluctuations

81%

89%

Moderately high water/
large fluctuations

11

7

8

4

Makes no difference

The surplus values for this scenario are $384 per trip for private
boaters and $467 for commercial passengers. These values are both
significantly lower than the respectiv,2/surplus values reported for
the 22,000 cfs constant flow scenario.
Case 6 - Constant Flow or 40,000 cf's. Case 6, the final flow
specific scenario respondents were asked to evaluate, was described
in the following manner:
At high water levels (around JJ0,000 crs), the current is
rast. Trips are able to stop at additional side canyons
and spend additional time at attraction sites. Fewer
rapids are present, as some or the smaller rapids are
"washed out.• In other rapids, however, the waves are
very large and some passengers, especially those on oar
powered trips, race an increased likelihood or having to
walk around one or aore or the major rapids ror their
saf"ety. Campsites become aore scarce as sandbars and
shore areas are flooded, and cam.psites are much smaller.
In SOiie areas of" the Canyon, there is an increased chance
of cam.ping with or near other groups.

The Chi-square statistics for these tests are 7.59 for private
boaters and 12.21 for commercial passengers, with 2 degrees of
freedom, indicating that the null hypothesis of no difference can
be rejected at the 0.10 level.

102

Most respondents felt that this scenario described a trip that would
be about the same or worse than their actual experience (Table
5-23). This result is consistent with the previously reported
findings in that 40,000 cfs is a higher flow level than most
respondents experienced and is also higher than the optimum flows
derived from the actual trip valuation data. The surplus values
respondents assigned to this scenario, $343 per trip for private
boaters and $439 for commercial passengers, reflect these feelings.
The value for commercial passengers is significantly lower than the
respective surplus value reported for the 22,000 cfs constant flow
scenario. However, a statistically significant difference does not
exist between the 22 1990 and 40,000 cfs constant flow scenario values
for private boaters.~
Table 5-23.

Rating of' The Ji0,000 cf's Constant Flows Scenario Relative
to Actual Trip

Rating
Better
About the Same
Worse

Proportion of Respondents
Citing Rating
Commercial
Private
Passengers
Boaters

7%

27
66

8%
36
56

Change 1 - Beaches Reduced. This scenario is not anchored at a particular
flow level but to the flow the respondent actually experienced.
Respondents were asked to evaluate a scenario where the number of sand
beaches available for camping are substantially reduced. The scenario
description is as follows:

j.Q./ The Chi-square statistics for these tests are 21.62 and 3.21,

respectively, with 2 degrees of freedom. These results indicate
that the null hypothesis of no difference can be rejected at the
O. 10 level for commercial passengers and cannot be rejected for
private boaters.

•

103

•

There are indications that certain types of flow
patterns in the long run may reduce the mnber of sandy
beaches in the Grand Canyon. At present, the area
between Hance Rapids and Havasu has fewer beaches than
other parts of the canyon. Trip leaders must plan
schedules very closely to ensure a good campsite in
this area. As beaches disappear, this careful planning
would have to be extended to other parts of the canyon.
This planning might aean missing some attraction sites
to get to camp early or longer stops at some attraction
sites. Fewer beaches would increase the likelihood of
camping near other parties and perhaps sharing a beach
with other parties. .SC.e caaps might have to be made
in areas without any sand.

Private boaters placed a surplus value of $377 per trip on this
scenario and the value for commercial passengers is $413 per trip,
indicating that a reduction in the number of beaches would
substantially decrease the surplus value that boaters place on their
Grand Canyon white-water trips. Only the constant flow and
fluctuating flow scenarios at 5,000 cfs have lower surplus values.
Summary
The attribute survey revealed that several important attributes of
Grand Canyon white-water trips are affected by flow levels: being in
a natural setting, stopping at attraction sites and hiking side
canyons, and rapids. Analysis of these attributes for respondents'
actual trips reveal that these attributes are generally sensitive to
release patterns from Glen Canyon Dam.

•

The optimum constant flow levels, according to the analysis of the
actual trip data, occur at constant average daily flows of 29,000 and
33,000 cfs for private boaters and commercial passengers,
respectively. The highest scenario surplus values, however, occur at
a constant flow of 22,000 cfs. We believe this difference is simply
due to the fact that we did not anchor a scenario in the flow range
from 29,000 to 33,000 cfs. If we had selected a scenario which was
anchored at an average flow of 31,000 cfs, we strongly suspect that
this would have been the scenario with the highest surplus value.
Despite the fact that we did not select a scenario that was anchored
at a moderately high flow, the surplus values for constant flow
scenarios show a great deal of consistency when plotted against the
actual trip surplus values. This comparison is done for commercial

104

passengers in Figure 5-2 and for private boaters in Figure 5-3.11/
The scenario values, for both groups of respondents, are somewhat
higher than the actual trip values at flows below 15,000 to 20,000
cfs. At higher flows the direction of the difference is reversed.
Given these graphic representations of the relationships between
average flow levels and surplus values, one can see why we believe
that a scenario anchored at a constant flow of 31,000 cfs may have
resulted in the highest surplus value across all scenarios.
Only 12 percent of the respondents experienced daily fluctuations in
flow levels in excess of 10,000 cfs. Thus, we did not have enough
observations to draw any inferences about fluctuating flow surplus
values from the actual trip data. The potential for this type of
problem was anticipated in the study design and the scenarios were
developed to describe Grand Canyon boating.experiences under a wide
range of flow regimes. From the analyses of responses to the
scenario valuation questions, we found that fluctuations in daily
flows significantly increase commercial passenger surplus values at a
low average flow of 5, 000 cfs, while private boater surplus values
were unaffected by these same conditions. At a moderate flow of
22,000 cfs, however, fluctuations in daily flow levels significantly
reduce surplus values for both commercial passengers and private
boaters. The flow value functions for fluctuating flows are
presented in Figure 5-4, and are derived by linear interpolation
between the 5,000 and 22,000 cfs fluctuating flow scenarios. The
functions are extended to 3,000 and 25,000 cfs to cover the full
range of average flows that can occur with daily fluctuations in
excess of 10,000 cfs.
Given the findings reported above, we would conclude that the Grand
Canyon White-Water Boater Contingent-Valuation Survey was quite
successful. The results are internally consistent and match well
with the preferences expressed by white-water boaters in the earlier
attribute survey, as well as the flow preferences of the commercial
white-water guides and private trip leaders collected in a separate
survey and summarized in the preceding chapter. The resulting
surplus values, therefore, seem to be adequate for the next stage in
the analysis, the evaluation of alternative annual flow regimes,
reported in Chapter 8 •

. 11/ The flow value functions for the scenarios were derived by
linear interpolation between the 5,000 and 13,000 cfs, the 13,000
and 22,000 cfs, and the 22,000 and 40,000 cfs constant flow
scenario values.

•

105

Figure

5-2

II

Commercial Boater Surplus Values for
Constant Flow Scenarios and Actual Trip
($ Per Trip)

$900
800
700
600
Surplus 500
Value
Per Trip 400

300
200
100
0
0

5

10

15

20

25

30

35

40

Average Flow Level in cfs
(x 1000)

{]- Actual Trip Values

•- Scenarios Values

45

50

106

..
Figure

5-3

Private Boater Surplus Values for
Constant Flow Scenarios and Actual Trip
($ Per Trip)

$900
800
700
600
Surplus
Value
Per Trip

500
400
300
200
100
0
0

5

10

15

20

25

30

35

Average Flow Level in cfs
(x 1000)

·D- Actual Trip Values

·•·· Scenario Values

40

45

50

107

..

Figure

5-4

Commerclal Passenger and Private Boater
Surplus Values for Fluctuatlng Flow Scenarios
($ Per Trip)

$900
800
700
600

Surplus 500
Value
Per Trip 400

..
..
..
..
.. .. ..

300

.. •.a
..
..
.... ••

•

-~

200

D

100
0
0

5

10

15

20
25
30
35
Average Flow Level in cfs
(x 1000)

·•- Private Boater Values

-a·

40

45

Commercial Passenger Values

50

108

CBAP'l'ER 6

GLEN CDYOH OGLER SURVEYS ARD RF.SUI.TS

Introduction
In this chapter we present the results of the attribute and
contingent-valuation (CV) surveys of Glen Canyon anglers. The
attribute survey was designed to provide a detailed understanding of
the fishing experience and to identify the important attributes of
the experience which are affected by flows. The results of the
attribute survey, particularly with respect to the identification of
flow-sensitive attributes, were used in designing the flow scenarios
to be evaluated in the Angler CV Survey. The results of the Glen
Canyon Angler CV Survey include an average actual trip surplus value,
a flow value function based on anglers' surplus values for a variety
of different flow scenarios, and values for two environmental impact
scenarios.
Background
As noted in Chapter 1, the current Glen Canyon trout fishery is a
by-product of Glen Canyon Dam. The dam provides more stable flows in
this section of the Colorado River than had existed previously. The
water is also colder and carries less silt. This new environment is
ideal for trout, and the Arizona Game and Fish Department began a
stocking program in 1964. As many as 100,000 rainbow trout have been
stocked in some years, and in more recent years brook trout have been
stocked as well. Fresh water shrimp were introduced to provide a
forage base for trout in 1968 to provide forage for trout. They have
flourished, providing ample support for the fishery.
Janisch (1985) has summarized the history of the fishery in four
stages. The period 1964 to 1971 was the "put-and-take" era;
catchable-sized trout were stocked and most were caught within a few
months. The average weight of the rainbow trout taken was less than
0.75 pounds during this period, and fishing pressure was relatively
light compared to more recent times.
Around 1971, shrimp started playing a major role in the trout's diet,
and the growth rate apparently increased. This resulted in the
trophy fishery era from 1972 through 1978. The average weight of
fish harvested peaked at 3,53 pounds in 1978. Bag limits of 10 fish
weighing a total of 40 pounds were not unusual during this period.
Anglers responded to the trophy fishing opportunity, with the number
of angler days growing rapidly. The presence of larger fish led

109

Arizona Game and Fish managers to suspect that substantial natural
reproduction was occurring, and the stocking strategy shifted from
introducing catchable-sized trout as practiced during the
put-and-take era to stocking fingerlings as a supplement to natural
reproduction. Research subsequently showed that the fishery is
heavily dependent on stocking, and that only a limited amount of
natural reproduction is taking place.
In 1978, the bag limit was reduced from ten to four trout in an
attempt to protect the resource from ever increasing fishing
pressure. In 1980, a rule was enacted requiring that trout either be
released or killed immediately after being caught. This rule was an
attempt to discourage people from keeping fish alive for extended
periods and then releasing them if a larger fish was taken, a
practice resulting in a high mortality rate for the released fish.
Even though the fishery has declined in productivity since 1978,
fishing pressure continued to escalate until 1984. Janisch has
termed the period from 1978 to 1984 the "quality fishery" era. Creel
census reports still showed a very respectable average weight of 2.79
pounds for fish caught and kept through this period. However, the
days of the trophy fishery were ending and the average weight of fish
taken declined steadily.
Janisch characterized the current era, beginning in 1985, as a period
of "something less than quality but not put-and-take." Catch rates
are still relatively high with some large fish taken, but most fish
are small in comparison to the trophy era. Anglers have responded by
greatly reducing participation from the peak of 52,000 angler days in
1983 to only 15,000 angler days in 1985. To further reduce fishing
pressure, the Arizona Game and Fish Department enacted a lures only
regulation which took effect on January 1, 1986. Preliminary
indications are that this has further reduced participation in the
fishery. Many anglers appear to have discontinued fishing in Glen
Canyon rather than use artificial lures. The current management
objective is to reduce fishing pressure and to continue stocking
trout to restore the fishery to the "quality," if not the "trophy, 11
level.
Attribute Survey Procedures
Sampling. No convenient list of names and addresses of Glen Canyon
anglers is available, so the only way to identify users is to contact
them on-site. Since the attribute survey was relatively short (see
Appendix F), it was conducted on-site. Two trained interviewers
administered the attribute survey to anglers.

110

The sampling strategy for the attribute survey was constrained by the
overall study deadline. It would have been desirable to sample
anglers randomly across a full year to be sure that there were no
seasonal biases in the responses. However, to keep the project on
schedule, sampling for the attribute survey was limited to November
and December, 1984. These are relatively high use months, especially
November, because fishing is often good. Unfortunately, Glen Canyon
Dam flow releases were stable at about 25,000 cfs for the entire
sampling period, so we could not sample anglers over a variety of
flows. While not ideal, this sampling scheme was adequate to
identify major attributes of the fishing experience.
The goal was to sample 200 anglers. Anglers were sampled on November
23, 24, 26, 29 and 30, and December 1, 2, 4, 6, 11, and 14. The
sampling strategy involved contacting anglers between 3 PM and dark,
when they were coming off the water after a day of fishing. Boat
anglers were contacted as they arrived at the Lee's Ferry boat dock.
Bank anglers were approached as they fished along the bank in the
dock area or as they were leaving through the Lee's Ferry parking
lot. One half of the adult members from each boat or party of bank
anglers were asked to fill out a one-page questionnaire.
Response Rate. A total of 211 completed on-site questionnaires were
obtained. This represents 92.5 percent of all the anglers contacted
(7.5 percent of the sampled anglers declined to complete the
attribute survey). Congestion at the boat dock and the high use
level on Thanksgiving weekend may have resulted in our missing as
many as five boats on November 23 and four boats on November 24.
When these missed parties are counted, the response rate for the
attribute survey was 87.8 percent. The reasons most often cited by
those refusing to participate were that it was too dark and that
there was insufficient time. A large share of the nonresponses
occurred during the congested period on Thanksgiving weekend.

Attribute Survey Results
Glen Canyon Angling Experience. In order to better understand the
effects of flow levels on fishing, it will be helpful to segment
results based on the amount of experience respondents had with
alternative water flows prior to the interview. Two measures of
experience will prove useful. First, we asked study participants how
many years they had fished in Glen canyon. More than one-third (37
percent) reported that 1984 was their first year, while 12 percent
had only fished at Lee's Ferry for two years. Twenty-seven percent
of the respondents had fished there for three or four years, and only
21 percent reported having fished at Lee' Ferry for more than five
years. The highest level of experience reported was 25 years.

111

Respondents were also asked about their experience with various
flows. Results of this question are summarized in Table 6-1. During
the sampling period, the river flow was constant at approximately
25,000 cfs, so all respondents had experienced relatively high steady
flows. Forty-four percent reported having fished in Glen Canyon at
medium flows (9,000 - 16,000 cfs), while 38 percent reported fishing
experience at low flow levels (9,000 cfs or less). More than half,
54 percent, reported having fished in Glen Canyon during periods of
daily fluctuations in flows. It is interesting to note, however,
that about one of three sampled anglers indicated that they didn't
know if they had fished under the various flow levels. For two of
the three water levels (low and medium), more than 50 percent of the
respondents either had not fished or didn't know if they had fished
under those conditions.
Table 6-1.

Respondent Reported Fishing Experience At Specif'ied Flow
Levels

Responses

•

Proportion Reporting Experience With Flow
Medium Flow
Low Flow
Fluctuating
(9,000-16,000 cfs) (Less than 9,000 cfs)
Flow

Yes, Experienced

44%

38%

No, Did Not Experience

20

27

16

Don't Know

36

35

30

When classified by reported historical experience with water levels,
48 percent of the anglers had experienced only high flows, 5 percent
reported experience with high flows and at least one other flow level
(medium or low), and 48 percent reported experience with high flows
and at least one other flow level, as well as fluctuating flow levels
(see Table 6-2). This bimodal distribution of reported experience
with flow levels is not surprising given that 49 percent of the
sample reported having fished at Lee's Ferry for two or less years, a
period during which high flows were the rule rather than the
exception.
Since flows were relatively constant during the attribute survey, it
was not possible to directly measure their impact on the perceived
quality of respondents' most recent fishing trips. Inferences about
these impacts were made from answers to questions about the expected

112

impacts of flow levels. Responses to these questions were examined
to determine if they are systematically linked to respondents'
historical experience with various flow levels.
Table 6-2.

Respondents Having Experience With Various Flow Levels

Flow
High Flows Only (greater than 16,000 cfs)
High and Low or Medi-um Flows
High, Low or Medium, and Fluctuating Flows

Proportion
With Experience

48%
5

48

Both years of fishing experience at Lee's Ferry and the range of
flows experienced were used to analyze the effects of experience on
anglers' responses. While these two measures of experience may be
important in the analysis of responses to questions about the impaots
of water level, it is important to note two facts. Of the anglers
having two or less years of experience at Lee's Ferry, 18 percent
reported having fished at Lee's Ferry during medium flows
(9,000-16,000 cfs), 8 percent reported fishing at Lee's Ferry during
low flows (9,000 cfs or less), and 22 percent reported fishing at
Lee's Ferry during fluctuating flows. Given that the two years prior
to the attribute survey had seen fairly constant and high flows, some
anglers may have answered the water level experience question
incorrectly. Secondly, the years of experience classification may
not provide a subsample of anglers that have experienced all flows.
For example, an angler with more than two years of experience may
still have fished only during certain water levels.
At several points, respondents' answers are broken down by experience
level. Anglers with more than two years of experience, as well as
reported exposure to daily fluctuations in flow level were classified
as "experienced."
Important Attributes. To identify the attributes that contribute
most, either positively or negatively, to the Glen Canyon fishing
experience, respondents were first asked the importance of various
reasons in their decision to fish at Lee's Ferry rather than
somewhere else. Table 6-3 shows that the two most important reasons
in the decision to fish in Glen Canyon are the size and the number of
fish the respondent expected to catch. This result holds true for
experienced and inexperienced anglers and for boat anglers as well.

113
The least important reasons for fishing at Lee's Ferry are the lack
of other trout fishing areas and its location relative to home,
indicating there are substitutes for the Glen Canyon fishery.
Table 6-3.

Reasons For Fishing in Glen Canyon

Reason

Proportion of
Res2ondents Stating Reason Is Im2ortant *
More than
Experience With
Boat
Two Years
All
Anglers Ex2erience Fluctuating Flows Anglers

Thought I would catch a
large fish

86%

82%

84%

88%

Thought I would catch
a lot of fish

75

73

75

77

Wanted to fish in
Glen Canyon

57

50

51

65

Few other trout fishing
areas available

41

37

40

45

Close to home

33

39

42

28

* These

percentages are derived from respondents' an5'Wers to questions
asking whether the reason is very, somewhat or not important.
Percentages reported here record the proportion of respondents
saying the reason is somewhat or very important.

•

Respondents were also asked to rate the importance of various factors
in contributing to a 2erfect or excellent fishing trip. As shown in
Table 6-4, the two most important attributes of an excellent or
perfect Glen Canyon fishing trip are "catching a trophy fish" and
"good weather." "Camping along the river" is the least important
attribute for all groups of respondents. The relative rankings of
attributes are quite consistent for both categories of experience and
for boat anglers .
Respondents were al so asked to rate the importance of a list of
factors that might contribute to a 2oor fishing trip. As shown in
Table 6-5, the most important factor contributing to a poor trip was

114

"Catching no fish." Other important factors were, in order of
importance, "Not being able to get upstream to fish," "Boat/motor
trouble due to water level," "Poor weather," and "Seeing many others,"
all of which were rated as important by 64 percent or more of the
respondents from each of the groups.
Table 6-4.

Attributes Contributing to an Excellent or Perrect Glen
Canyon Fishing Trip

Attribute

Proportion of
ResQondents Stating Attribute Is ImQortant *
More than
All
Two Years
Experience With
Boat
Anglers ExQerience Fluctuating Flows Anglers

Good weather

84%

80%

80%

84%

Catching a trophy fish

79

79

85

83

Catching your limit

68

59

64

68

Seeing few others

67

66

76

70

Low water

60

66

70

59

High water

54

54

59

54

Falling water level

52

55

59

47

Rising water level

51

56

54

49

Camping along the river

26

27

34

27

* These

percentages are derived from respondents' answers to questions
asking whether the attribute is very, somewhat or not important.
Percentages reported here record the proportion of respondents saying
the attribute is somewhat or very important.

•

115

Table 6-5.

Attributes Contributing to a Poor Glen Canyon Fishing Trip

Attribute

•

Proportion of
*
Respondents Stating Attribute Is Important
More than
All
Two Years
Experience With
Boat
Anglers Experience Fluctuating Flows Anglers

88%

Catching no fish

89%

85%

Not being able to get
upstream

79

70

76

87

Poor weather

78

76

75

77

Boat/motor trouble due
to low water

74

67

72

81

Seeing many others

71

64

71

74

Not catching your limit

63

58

60

64

High water

61

66

71

61

Low water

61

69

75

61

Falling water level

58

54

67

54

Rising water level

57

61

56

55

Not catching a trophy fish

57

53

56

59

Not being able to camp
along the river

31

29

33

33

* These percentages are derived from respondents' answers to questions
asking whether the attribute is very, somewhat or not important.
Percentages reported here record the proportion of respondent saying
the attribute is somewhat or very important.

116

In addition to important attributes of a Glen Canyon fishing
experience, anglers were also asked about their knowledge of the flow
level on the day of their trip. Anglers were the recreation group
most likely to have made several visits and, consequently, may have
been most inclined to find out about the flows at the time of their
trip.
The majority of anglers (75 percent), however, did not try to find
out the expected flow prior to arriving at Lee's Ferry. Twenty-five
percent attempted to learn the expected flow level, and about 70
percent of these anglers were successful. Experienced anglers were
more likely to try to find out the expected flow level before the
trip, but even among those who had experienced a variety of flows,
only 38 percent reported seeking information prior to arriving at
Lee's Ferry. This result does not necessarily imply that anglers are
uninterested in flow levels or that their fishing experiences are
unaffected by flows, but rather, that there may be other more
constraining factors which determine the days on which they go
fishing. For example, most people only have a limited amount of free
time and it may be unrealistic to expect that an angler will only go
fishing on his or her free days when preferred flows exist. It may
be that most go fishing when they can and then adapt their fishing
techniques to the flow levels they experience.
Since nearly all anglers sampled for the attribute survey experienced
relatively high flow levels on their trip, we also asked respondents
to assess the impact of a lower water level on several attributes of
a Glen Canyon fishing trip. Among anglers with an opinion, there was
a general consensus that water levels lower than 25,000 cfs would
improve the chances of catching fish as well as the chances of
catching a trophy fish (Table 6-6). Experienced anglers expressed a
belief that lower water would be more likely to increase the
probability of catching fish. However, these proportions were only
slightly higher than those for catching a trophy fish. Experienced
anglers also were more likely to think that low water would increase
the chance to fish in certain preferred areas.
In summary, the results of the attribute survey indicate that two
factors dominate the evaluation of fishing trips at Lee's
Ferry--catching fish and the weather. Other secondary (but still
important) factors include the degree of crowding, the ability to get
upstream, and boat/motor trouble. Respondents, especially
experienced respondents, felt lower water would improve the chances
of catching a fish in general and increase the chances of catching
trophy fish in particular. They also felt lower flows would improve
their chances to fish certain preferred areas.

117
Table 6-6.

Perceived Eff'ects of' Low Flow Levels on a Glen canyon
Fishing Experience

Attribute

Proportion of Respondents
*
Stating Low Water Would Increase Activity
More than
Experience With
Boat
All
Two Years
Anglers
Experience Fluctuating Flows Anglers

Chances of catching a
trophy fish

43%

49%

54%

40%

Chances of catching
fish

43

54

56

40

Chance to fish certain
preferred areas

36

45

50

32

Amount of time spent
fishing

25

29

35

24

Problems with boat
or motor

20

21

26

23

* These

percentages are derived from respondents' answers to a question
in which they were asked whether a low flow level would increase,
decrease, have no effect, or don't know effect on specified
activities. The percentages reported here record the proportions of
respondents saying a low flow level would increase the activity.

Contingent-Valuation Survey Procedures
The sampling frame for the Glen Canyon Angler
Contingent-Valuation Survey consisted of anglers at Lee's Ferry on 75
selected days between April 29 and December 19, 1985. A sampling
period of this length was chosen to minimize the potential for a
seasonal bias in the types of anglers selected to participate in the
study. The same procedures used for the angler on-site attribute
surveys were used to contact anglers and to solicit their names and
addresses for the CV mail survey.
Sampling.

•

Our field personnel estimated that 986 anglers were eligible for
selection on the 75 specified sampling days, and they were able to
contact 900 (91 percent). Some anglers were missed during busy times
at the dock, while others had not returned by dark. For the 900

118
anglers contacted, 774 completed the on-site questionnaire (86 percent)
and provided a usable name and address for the mail survey. The
remaining 126 anglers either declined to complete the on-site
questionnaire or listed insufficient or illegible address information.
The sampling frame, then, contained 86 percent of the anglers contacted
and 78 percent of the estimated total number of anglers.
Three hundred of the anglers who provided usable address information
during the period April 29 through July 29, 1985 had been previously
selected to participate in the Glen Canyon Anglers CV Pretest Survey
(See Appendix K for the goals and results of the pretest). Of the
remaining anglers in the sampling frame, 298 were randomly selected to
participate in the final CV survey. The results of this latter survey
are presented in this report.
Response Rate. The Glen Canyon Angler CV mail survey was conducted
during January-March, 1986. Overall, 237 completed questionnaires were
returned, 80 percent of the total sample. Two of these responses were
subsequently excluded from the data analyses; one individual responded
to a faulty questionnaire and the other was only 13 years old. Thus,
the results presented in this report are based on 235 usable
responses. Summary statistics for the response are presented in Table
6-7.
Table 6-7.

Glen Canyon Angler CV Survey Response Rate

Surveys
Completed Surveys
Undeliverable *
Not applicable
Surveys not returned
Refusal
TOTAL

*

Percent of
All Surveys
79%
3
1
17 ***

_Q

100%

Percent of
Deliverable Surveys **
82%
18
0
100%

Two surveys were returned, but were not included in the data
analyses. As noted in the text, one of these was from an individual
who was only 13 years of age and the other was returned by an
individual who responded to a faulty survey.

**

The percentages in this column are computed from a sample size of
288 rather than 298. The undeliverable and not applicable sur·veys
have been excluded.

***

Indicates a number which is less than one percent of the sample,
i.e., only one angler refused to participate in the survey.

•

119
Contingent-Valuation Survey Results
Actual Trip. The valuation section of the Angler CV Survey was
initiated with a question about the actual expenditures incurred on
their trip (see Figure 3-5 for the specific question format). On
average, respondents reported expenditures that totaled about $189 for
the trip (Table 6-8). This is the actual sum of average expenditures
for individual items. However, when asked to add up their
expenditures, respondents reported a slightly lower average total. of
$156 for the trip. This difference may be due to computational
errors, or it may simply reflect respondents' best guess as to how
much they really spent in total. Regardless of the reason for this
difference, respondents were instructed to refer to perceived total
expenditures (the mean of $156) when responding to the subsequent CV
questions.
Table 6-8.

Average Expenditures Reported by Anglers on the Trip When
On-Site Interview Completed

Item

Amount Spent

Gas and Oil for Vehicle
Food and Beverages
Lodging, Camping
Fish Equipment/bait/license
Guide Fees
Boat and Equipment Rentals
Air fare
Car Rental
Other

$ 40

TOTAL

$189

40
24
20

13
7
9
1

35

After determining their total expenditures for the actual trip,
respondents were presented with a dichotomous-choice CV question in
which they were asked if they would have gone on this trip if their
expenses had increased by a specified amount. The specified amount
presented to each angler was a randomly selected amount (offer) from a
range of values determined by the results from the pretest survey.
The first step in the analysis of 7esponses to the valuation question
was to estimate a logit equation.-1 Several variables in addition to
the offer amount were used as independent variables to explain

ll Maximum likelihood procedures were used to estimate legit equations.

120

responses. For the actual trip valuation analysis, independent
variables included expenses actually incurred, whether or not a guide
was employed, the season of the year when the trip took place~ length
of trip in days, and the number of fish caught, among others.-1
The probability of any given respondent answering yes to the valuation
question decreased as the magnitude of the dichotomous-choice dollar
offer increased. In addition to the offer amount, actual trip expenses
and respondents' reported concern about the use of the CV results to
increase license fees were significant at the 0.10 level in explaining
responses to the dichotomous-choice question for the actual
trip.
The sign of the coefficient on expenses indicated that those with
higher expenses would also have higher surplus values. Economists,
however, would ordinarily expect the opposite: other things being
equal, the higher the expenditures, the smaller the surplus values.
The license fee variable was significant and the sign for this variable
indicated that those who were concerned. about fishing fees had lower
values. This was interpreted as an indicator of strategic bias. The
variable was coded as "1" if the respondent was concerned about
increases in fishing license fees when answering the CV questions and
zero otherwise. In calculating values, this variable was set equal to
zero in the logit equation to eliminate the effects of strategic bias
from the final results.
Most notable among the nonsignificant variables was the flow level
experienced on the actual trip. However, further investigation of this
result produced a reasonable explanation. About half of our sample
experienced constant flows while half experienced fluctuating flows,
where the latter was defined as a fluctuation of more than 10,000 cfs
in the course of a 24-hour day. The sample was split on this basis and
separate logit equations estimated for each. Coefficients for the two
equations were significantly different at the 0.05 level, indicating
that surplus values pej_ trip are significantly different for constant
and fluctuating flows. 1 Actual trip surplus values were estimated
to be $130 for constant flows and $104 for fluctuating flows (Table
6-9).

5=_/

•
The statistical analyses of the respondents' answers to the actual
trip and each of the scenario valuation questions are summarized in
Appendix N, and only the salient results with respect to estimated
surplus values will be reported.

11 The Chi-square statistic for this test is 8.53 with 3 degrees of
freedom, indicating that the null hypothesis of no difference can be
rejected at the 0.05 level.

121

Table 6-9.

Estimated Glen Canyon Angler Surplus Values {$ Per Trip)

ll

Type of Surplus Value
Actual Trip:
Constant Flow
Fluctuating Flows
Scenarios:
3000 cf s, constant
3000 cfs, with fluctuations
10,000 cfs, constant
10,000 cfs, with fluctuations
25,000 cfs, constant
25,000 cfs, with fluctuations
40,000 cfs, constant
Chances of catching a trophy fish doubled
Chances of catching no fish doubled

Surplus
Values
$ 130

104
$

60
77
126
87
94
68
52
139
64

Interestingly, the constant flows tended to occur at relatively high
levels, averaging 28,800 cfs, while the fluctuating flows occurred at
an average daily flow of only 11,900 cfs with an average daily
fluctuation of 15,500 cfs. Of those who had experienced a constant
flow level, none experienced an average flow below 10,000 cfs, and 75
percent experienced an average daily flow between 20,000 and 30,000
cfs. In contrast, only two of the respondents who experienced a
fluctuating flow level were on the river when the average flow
exceeded 20,000 cfs. In fact, 75 percent of these respondents
experienced an average daily flow between 5,000 and 15,000 cfs. This
unequal distribution of flow levels experienced by anglers confounded
our efforts to isolate the effects of differences in the average
daily flow on actual trip surplus values.

•

Although the data did not permit quantification of the relationship
in the actual trip equation, the importance of flows to fishing
quality is borne out in the attribute survey data by the significant
difference between the responses to the actual trip valuation
question between those who experienced constant flows and those who
experienced fluctuating flows, and the scenario values to be
presented below. With adequate data, the relationship between flows
and surplus values would, we believe, be statistically significant
for actual trip data. However, given inadequate data relating flows
to actual trips, the flow-value curve for anglers will be based on
responses to the scenarios through which all anglers evaluated a
variety of different flow conditions.

122

Two other variables which were not significant in the logit analysis
are worth noting. One was an indicator of respondents' self reported
confidence about their responses to the CV questions. This finding
suggests that hypothetical bias was not a problem in this analysis.
The other variable was constructed to assess the possibility of
seasonal bias. This was a concern because our sample was not fully
representative of anglers across the entire year of 1985. The
sampling period extended over almost eight months, but sampling did
not occur on every day during this period, and 300 of the anglers
contacted early in the year were used for the pretest. The season
variable was not significant, indicating that value estimates for
actual trips are not affected by season.
Scenarios. Respondents were asked to evaluate a total of nine
scenarios (see Table 6-9). As with the analysis for the actual trip
valuation question, several variables were tested to determine if
they had a statistically significant effect on surplus values. The
amount of increase in expenses was significant at the 0.10 level or
better in all nine cases. The variable measuring actual trip
expenditures was significant at the 0.10 level in six of the nine
cases and, once again, there was a positive relationship between
expenditures and expected values, a result that was unanticipated.
We found no evidence of strategic bias or hypothetical bias in
respondents' answers to the scenario valuation questions. See
Appendix N for details of the analyses for scenario valuation
questions.

With so many scenarios, one concern was whether respondents' answers
to the associated valuation questions might be affected by the order
in which the scenarios were presented. As respondents worked through
the questionnaire, they might have become bored with the process or
later responses might have been influenced by how they responded to
earlier scenarios. To test for the possibility of an "ordering
effect," two different survey booklets were designed, differing only
in the sequence in which the scenarios were presented. Sampled
anglers were randomly assigned to one of the two survey groups.
Statistical analysis of responses to these two sequences of scenarios
failed to reveal an ordering effect, indicating that the order in
which the scenarios were evaluated by respondents did not affect
surplus values. See Appendix N for the full analysis.
The estimated surplus value per trip for each scenario is summarized
in Table 6-9. Because these values will be used to derive the
estimated flow value function and for other purposes, each scenario
is discussed briefly below.
Case 1 - Constant Flow of 3,000 cfs. This first scenario represents
a Glen Canyon fishing experience at a low flow level that is constant
at 3,000 cfs. The fishing experience was described to respondents in
the following manner.

•

123

Boat anglers have said that getting upstream to :fish can
S011etimes be a problem at low water (3,000 cf"s or less). At
a constant now of' 3,000 cf"s, large boats can't get past the
sand and gravel bar three miles upstream :frooa Lee's Ferry,
while even very small boats may have to be dragged over
slippery rock gravel bars. Consequently, nearly all of' the
:fishing would occur in the three miles just above Lee's
Ferry. In addition, damage to boats and •otors is somewhat
more :frequent than at higher water levels. However, low
water tends to concentrate :fish, and bank anglers can :find
large areas of' exposed gravel and rocks leaving a great deal
of' space between the water and the edge of' the vegetation.
The majority of respondents (64 percent) felt that this
represented an angling experience that would be "worse"
actual trip, while 23 percent said that this would be a
experience (Table 6-10). The surplus value respondents
this scenario is $60 per trip.
Table 6-10.

Rating of' the 3,000 cf"s Constant Flow Scenario Relative
to Actual Trip

Rating
Better
Worse
About the same

..

scenario
than their
"better"
assigned to

Proportion
Citing Rating
23%
64
13

Case 2 - Average Flow of' 3,000 cf"s with Fluctuations. For this
scenario respondents were asked to evaluate an angling experience with
daily fluctuations in flow levels from 1,000 cfs to 15,000 cfs, with
an average flow of 3,000 cfs. This scenario was described to
respondents as follows •
The questions above asked about a relatively constant :flow of
3,000 cf"s. Daily changes in the water now may have other
e:f:fects, in addition to those described in Case 1, on the
quality of' fishing. With nows changing daily from a low of
1,000 cf"s to a high flow of 15,000 c:fs (around an average
now of 3,000 cf"s), boats may get swamped i f they are tied
too tightly to the bank during a :fluctuation. There is also
a chance of getting stranded above 3 mile bar if' the water
drops substantially. On the other hand, biological studies

124

give some indication that rising water may cause the fishing
to improve as fish begin to feed on the debris stirred up by
the rising water.

After reading this description, respondents were asked whether daily
fluctuations around an average flow of 3 ,000 cfs would "hurt,"
"improve," or "make no difference" in a Glen Canyon fishing
experience relative to a constant flow of 3,000 cfs (Case 1). The
largest group of respondents (44 percent) said that fluctuations
would hurt the fishing experience, while 36 percent said that
fluctuations would improve the fishing experience (Table 6-11).
Table 6-11.

Rating of the 3,000 cf's with Fluctuations Scenario
Relative to the 3,000 cfs Constant Flow Scenario

Rating
Improve the fishing experience
Hurt the fishing experience
Make no difference

Proportion
Citing Rating

36%
44
20

The surplus value assigned to this scenario by respondents is $77 per
trip. This value is not statistically different from the $60 surplus
value res~9ndents assigned to the 3 ,000 cfs constant flow
scenario.Thus, based on surplus values, we would conclude that
large daily fluctuations around a low average flow of 3,000 cfs would
neither hurt nor improve a Glen Canyon fishing experience, on
average, relative to a constant flow of 3,000 cfs.
Case 3 - Constant Flow of 10,000 cf's. This scenario describes an
angling experience at a moderate flow that is constant at 10,000
cfs. The scenario description provided to respondents is as follows:

.!±_/The Chi-square statistic for this test is 3.18 with 2 degrees of
freedom, indicating that the null hypothesis of no difference
cannot be rejected at the 0.10 level.

125

Boat anglers seea to experience fever problems with damage to
their boats and motors when the water is at least 10,000 cfs,
and boats can get up and downstream with no difficulty. At a
f'l.ow of 10,000 cfs, bank anglers would still find exposed
gravel and rock bars and some ro<m between the water's edge
and shore vegetation. In previous studies, about 40 percent
of the anglers have said that they feel the fishing is
generally better at constant f'l.ovs of 10,000 cfs than when
the water level is higher.
A majority of respondents (56 percent) felt that this scenario
represented a "better" fishing experience than their actual trip,
while only 13 percent felt that the experience would be "worse"
(Table 6-12). The 56 percent who indicated that this experience
would be better than their actual trip is significantly larger than
the 23 percent who said that Case 1 (3,000 cfs) would be better than
their actual trip, indicating that respondents prefer the 10,000 cfs
flow level to the 3, 000 cfs flow level described in Case 1.
Table 6-12.

Rating of the 10,000 cfs Constant Flow Scenario Relative
To Actual Trip

Rating
Better
Worse
About the Same

Proportion
Citing Rating
56%
13
31

The surplus value assigned to this scenario by respondents is $126 per
trip. This surplus value is significantly different from the $6g
surplus value reported for the 3,000 cfs constant flow scenario.-1
Thus, an increase in the flow level from 3,000 to 10,000 cfs
significantly increases angler surplus values.

2.I The Chi-square statistic for this test is 38.02 with 2 degrees of
freedom, indicating that the null hypothesis of no difference can be
rejected at the 0.01 level.

126

Case .II - .Average Flow of 10,000 cf's with Fluctuations. This scenario
represents an angling experience with daily flows ranging from 1,000
cfs to 22,000 cfs, with an average daily flow of 10,000 cfs. The
scenario description is presented below.
The questions above asked about a relatively constant flow
of 10,000 cf's. Daily changes in the now may have other
effects on the quality of fishing in addition to those
described above for Case 3. At moderate nows, large daily
fluctuations frc:a a low now of 1,000 cf's to a high flow of
22,000 cf's (around an average of 10,000 cf'.s) may contribute
to the swamping of boats tied to the bank or dragging
anchors. There would still be a chance of getting stranded
above 3 mil.e bar. .Again, however, there is some indication
that the rising water may improve fishing as fish begin to
feed on debris stirred up by the rising water.
After reading this scenario, respondents were asked whether the
described conditions would "hurt," "improve," or "make no
difference" in a Glen Canyon fishing experience relative to a
constant flow of 10,000 cfs. A majority of respondents (53
percent) said that fluctuations would hurt the fishing
experience, while 30 percent said that fluctuations would
improve the fishing experience (Table 6-13). The 53 percent
indicating that fluctuations would hurt the fishing experience
at 10,000 cfs is significantly larger (p<0.05) than the 44.
percent of respondents who said fluctuations would hurt the
fishing experience at 3,000 cfs. This result suggests that
anglers feel fluctuations have a more adverse effect on a
fishing experience at 10,000 cfs than at 3,000.cfs. The surplus
values reported below support this conclusion.
Table 6-13.

Rating of the 10,000 cf's with Fluctuations Scenario
Relative to the 10,000 cf's Constant Flow Scenario

Rating
Improve the fishing experience
Hurt the fishing experience
Make no difference

Proportion
Citing Rating
30%
53
17

127

The surplus estimate for this scenario is $87 per trip. This value
is statistically different from the 6 ~urplus reported for the 10,000
cfs constant flow scenario of $126.In contrast to the finding
that fluctuations neither hurt nor improve a fishing experience at an
average flow of 3,000 cfs, we found here that daily fluctuations in a
moderate flow level (10,000 cfs) significantly reduces angler values.
Case 5 - Constant Flow of 25,000 cfs. For this scenario respondents
evaluated a moderately high constant flow level of 25,000 cfs. The
scenario description provided for respondents is presented below.
Next, consider a constant now of 25,000 cfs.

There is no

minimllll motor horsepower restriction, although motors with

10 h.p. or less may have problems getting upstream. The
chance of damage to boats and motors due to obstructions in
the water are small, but the high nows of 25,000 cfs may
increase the chances of swamping a boat while dragging an
anchor, especially for inexperienced boaters.
Fish may be less concentrated at this higher flow level.
Bank anglers may have less space between the waters edge
and bank vegetation, but eddies along the shoreline are

often larger and more pronounced. About 12 percent of the
anglers in a recent survey felt that fishing was better at
25, 000 cfs than at lover water levels in Glen Canyon.
The largest group of respondents (44 percent) felt that the
conditions presented in this scenario were "about the same" as they
had experienced for their actual trip (Table 6-14). This result is
not surprising given that the average flow experienced by all
respondents was 20,500 cfs.
Table 6-1-.

Rating of the 25,000 cfs Constant Flow Scenario Relative
to Actual Trip

Rating
Better
Worse
About the same

Proportion
Citing Rating
23%
33

44

E_/ The Chi-square statistic for this test is 10.60 with 2 degrees
of freedom, indicating that the null hypothesis of no difference
can be rejected at the 0.01 level.

128
Respondents assigned a surplus to this scenario of $94 per trip.
This surplus value is significantly different from i?e value of $126
reported for the 10,000 cfs constant flow scenario.
Thus, for a
large increase in the average daily flow level from 10,000 to 25,000
cfs, angler surplus values decline.
Case 6 - Average Fl.av o:f 25,000 cf's with Fluctuations. This scenario
presents an angling experience with fluctuations in flow levels from
12,000 cfs to 32,000 cfs, with an average daily flow of 25,000 cfs.
The scenario respondents were asked to evaluate is as follows.
The impacts o:f large daily :fluctuations are somewhat
dir:ferent at higher water than at lower water levels. With
:fluctuations :from low :flows o:f 12,000 c:fs to high :flows o:f
32,000 cf's (around an average now o:f 25,000 cf's), it is
very unlikely boats would get stranded above 3 E.l.e bar.
Boats tied too tightly to the shoreline, however, may be
nooded. Rising water might also trigger more feeding by
:fish, but :fish becane more dirf:"icult to find because o:f the
higher water and faster current. In the long run, large
daily nuctuations at this now level may wash away many o:f
the campsites upstream :from Lee's Ferry.
The majority of respondents (65 percent) felt that fluctuations at an
average flow of 25,000 cfs would hurt the fishing experience (Table
6-15). This proportion is significantly different, at the 0.01
level, from the 53 percent who said fluctuations would hurt fishing
at 10,000 cfs and the 44 percent who said fluctuations would hurt
fishing at 3,000 cfs. These results indicate that fluctuations are
perceived as being more problematic by Lee's Ferry anglers at higher
flow levels.
Table 6-15.

Bating of:" the 25,000 cf's with Fluctuations Scenario
Relative to the 25,000 cf's Constant Fl.av Scenario

Rating
Improve the fishing experience
Hurt the fishing experience
Make no difference

Proportion
Citing Rating
16%
65
19

II The Chi-square statistic is 6.91 with 2 degrees of freedom,
indicating that the null hypothesis of no difference can be rejected
at the 0.05 level.

.

129

The surplus value respondents assigned to this scenario is $68 per
trip. This value is statistically different from the value reported
for the 25,000 cfs constant flow scenario ($94), indicating that
fluctuating flows also have a si§?ificant negative impact on angler
values at moderately high flows.-

Case 7 - Constant Flow of' ll0,000 cf's. This scenario describes a high
flow level that is constant at 40,000 cfs. The scenario was
described in the following manner.
At constant f'lows of' 40,000 cf's, the current is swif't and the
Park Service requires all boat motors to have at least a 25
horsepower motor. Large boats can get up and down the river
more easily than smaller boats. The chances of' damage to
boats and motors due to obstructions in the water are smaller
than at lower flows. However, f'or inexperienced boaters the
high water increases the chances of' boats being swamped while
dragging anchors. Eddies along the shoreline are larger and
well def'ined, but bank anglers f'ind the water is up into the
bank vegetation and this may make bank f'ishing more dif'f'icul t
f'or them. At these high f'lows, f'ish f'eeding patterns may
change since f'ish would generally stay out of' the main
current. Fish may be harder to f'ind.
The majority of respondents (79 percent) said that this scenario
described an angling experience that would be worse than their actual
trip, and only 5 percent rated the conditions as being better than
their actual trip (Table 6-16).

Table 6-16.

Rating of' the 110,000 cf's Constant Flow Scenario Relative
to Actual Trip

~

Rating
Better
Worse
About the same

Proportion
Citing Rating

5%
79
16

.§_/The Chi-square statistic for this test is 8.16 with 2 degrees of
freedom, indicating the null hypothesis of no difference can be
rejected at the 0.05 level.

130

The surplus value for this scenario is $52 per trip. This value is
significantly diffe§_7nt from the value of $94 for the 25,000 cfs
constant flow trip.
This result indicates that surplus values
continue to decline at a very high constant flow of 40,000 cfs.
Change 1 - Bigger Fish. This scenario was not anchored to a
particular flow level, but to the flow level experienced on the
actual trip. This scenario was explained to respondents as follows:
A survey of' anglers at Lee's Ferry last year showed that
about 15 percent of' them reported catching a f'ish larger
than three pounds, and only 3 percent reported catching a
f'ish larger than f'our pounds. These numbers ref'lect how an
average angler might do on any particular day at Lee's
Ferry. We realize that no one is e:xactly average, but we
would like you to suppose that the f'ishery at Lee's Ferry
changed in such a way that your chances of' catching one of'
these bigger f'ish were to double. If' you f'eel you are an
average f'isher, your chances of' catching a f'ish bigger than
three lbs. would now be about 30 percent, while your chances
of' catching a fish bigger than four lbs. would now be about
6 percent. If' you think you are not an average f'isherman at
Lee's Ferry, your chances would vary accordingly.
The estimated surplus values for this scenario is $139 per trip,
indicating that increases in the probability of catching a larger
fish is an important attribute of a Lee's Ferry fishing experience.
Change 2 - Chances of' Getting Skunked. Here respondents were asked
to evaluate a Lee's Ferry fishing experience where the chances of
catching no fish were doubled, regardless of the flow level. The
scenario description is presented below.
We are sure that almost every angler has experienced, at one
time or another, •getting skunked• (catching no f'ish at all).
In f'act, about 20 percent or our respondents to a previous
survey at Lee's Ferry reported they had not yet caught a
f'ish. This number renects how an average angler might do on
any particular day at Lee's Ferry. Bo one is e:xactly average,
but we would like you to suppose that the f'ishery at Lee's
Ferry changed in such a way that your chances of' getting
skunked were to double. In other words, if' you feel you are

28.99 with 2 degrees
of freedom, indicating that the null hypothesis of no difference
can be rejected at the 0.01 level.

!)_/ The Chi-square statistic for this test is

'

.--------------------------------------------

-~---

131

an average angler and came repeatedly to Lee's Ferry, you
could expect that on f'our out of' ten trips you would catch no
f'isb at all. If' you were an above average angler, your
chances of' getting skunked would be less, and if' you were
below average, the chances would be greater than 4 out of 10
tiaes.

The estimated surplus values for this scenario is $64 per trip. This
indicates that doubling the chances of being skunked would
si.gnificantly diminish Glen Canyon anglers' surplus values. The
surplus value for this type of angling experience is comparable to
the surplus value of $60 per trip for the 3,000 cfs constant flow
scenario.
Summary
The results of the attribute survey revealed that catching fish and
good weather are the dominant attributes of a Glen Canyon fishing
experience. Other important attributes that could be affected by
flow levels included the degree of crowding on the river, the ability
to get upstream, and boat or motor trouble due to low water. These
last two attributes are especially important to boat anglers.
The flow-sensitive attributes identified by the attribute survey were
used in the design of flow scenarios to be evaluated by anglers. In
addition, given anglers' concerns regarding catching fish, and
catching a trophy fish in particular, two environmental impact
scenarios were designed to assess the effects of long-run
environmental changes on these attributes.
Respondents placed the highest surplus value ($139) on the scenario
describing a fishing experience in which the chances of catching a
fish greater than three pounds would be doubled. This result is not
surprising given that Glen Canyon has become known as a trophy
fishery and, on average, respondents had several years of experience
with this fishery. At the other end of the continuum, respondents
assigned the lowest surplus value ($52) to the 40,000 cfs constant
flow scenario. This result indicates that very high flow levels in
Glen Canyon significantly reduce angler surplus values.
Examining only the flow specific scenarios, respondents assigned the
highest surplus value to the 10,000 cfs constant flow scenario
($126). The value for constant flow scenarios rose from $60 at 3,000
cfs to $126 at 10,000 cfs, and then declined to $94 at 25,000 cfs and
$52 at 40,000 cfs. We found that all of the pairwise comparisons of
surplus values from 3,000 to 10,000 cfs, 10,000 to 25,000 cfs, and
25,000 to 40,000 cfs were significantly different at the 0.05 level.

132

While these scenarios cannot be used to identify a specific optimum
flow, the optimum would almost certainly fall in the 8,000 to 15,000
cfs range.
At a low flow of 3,000 cfs, there was not a statistically significant
difference in surplus values for the constant flow and fluctuating
flow scenarios. Given this finding, we will use a combined value of
$68 for each of these scenarios in future analysis of flow release
patterns from Glen Canyon Dam (see Figure 6-1). At higher flow
levels, 10,000 cfs and 25,000 cfs, respondents placed significantly
lower values on the fluctuating flow scenarios than they did on the
corresponding constant flow scenarios. These findings indicate that
fluctuations at low flows neither improve or hurt the experience for
the average angler, but 1 ~1ve a significant negative impact on surplus
values at higher flows.~
Thus, Figure 6-1 maps out the flow
value curves for constant and fluctuating flow levels. These surplus
value-flow relationships are based on the analyses of response to the
flow-specific scenarios. The constant flow function was derived by
linear interpolation between the 3,000 and 10,000 cfs, 10,000 and
25,000 cfs, and 25,000 and 40,000 cfs constant flow scenario values.
The 25,000 to 40,000 cfs line segment was extended to 45,000 cfs to
cover the full range of flows to be evaluated and to maintain
consistency with the white-water boater flow value functions. The
fluctuating flow function was also derived by linear interpolation
using the fluctuating flow scenario values.
Final Observations
Two caveats need to be expressed regarding the use of the angler flow
value functions. First, the change in regulations requiring that
only artificial lures be used in Glen Canyon upstream from Lee's
Ferry is a basic change in the parameters of the fishery. The effect
on values is unclear. Among some segments of the angling public who
prefer lures, this may enhance surplus values. We would speculate
that for most anglers, however, the effect would be in the opposite
direction. There are numerous unofficial reports of substantial
reductions in participation and some displacement to sites downstream
where access is more difficult but natural bait is legal.

1.Q/

A scenario presenting a fluctuating flow around an average flow
of 40,000 cfs was not evaluated because any flow level above
33,500 from Glen Canyon Dam can only occur as a constant flow,
since the bypass tubes are open. In addition, the highest
average flow that is possible under conditions of fluctuating
flow is 25,000 cfs.

133

..
Figure

6-1

Glen Canyon Angler Flow Value Functions for
Constant and Fluctuating Flow Levels
($ Per Trip)

$140

120

100

Surplus
Value
Per Trip

80

60

."'-•

40

20

0

0

5

15

10

20

25

30

35

Average Flow Level in cfs
(x 1000)

·O-

Fluctuating Flow ·•- Constant Flow

40

45

50

134

The results reported here, and the history of the fishery, however,
caution against assuming this is a permanent state of affairs. If
the regulation, as intended, helps restore the trophy fishery of a
few years ago, angler surplus values could be greatly enhanced.
A second caveat relates to the emphasis throughout this chapter on
value per trip. Less attention has been given to number of trips
taken per year by individual anglers. This concern stems from the
history of the fishery. The number of angling days has dramatically
changed in response to changing conditions over the years. Ideally,
we would have estimated the effects of flows not only on surplus
value per trip but also on the number of trips taken by anglers under
the new conditions. However, this would have required many more
resources than were available for this study. For example, if
fishing conditions improved, not only would it be necessary to
estimate the effect on the number of trips by current participants,
but also the number of trips taken by new participants who were not
in the user population in 1985. To gain information about these
people may have required a general population survey of Arizona, Utah
and California. Even with funds for such a survey, we are not
convinced that the results would be reliable. Once fishing quality
improved and use rates responded, additional issues such as the
effect on fish populations and potential crowding might affect
participation.
Given the unpredictable future of the fishery and, consequently,
participation, we will value alternative operating plans for Glen
Canyon Dam using 1985 values estimates and 1985 estimates of trips
taken. This is a conservative approach since 1985 was a rather poor
year for fishing and a year of steady high flows. ~~ny of the flow
regimes that will be evaluated in Chapter 8 have flows more favorable
to fishing than those experienced by most of the 1985 anglers. Thus,
the use of 1985 trips is probably an underestimate of the total
number of trips that would be taken under more favorable conditions.
The ultimate effect of the lures regulation remains to be seen. To
the extent that fishing improves, surplus values will be enhanced.
If the management program fails, a return to 1985 conditions should
be feasible by lifting the ban on natural bait. Thus, given the
uncertainty surrounding the future of the fishery and participation
relationships, use of 1985 values and participation rates is a
reasonable way to represent the economic effects of alternative flow
management regimes on the fishery. However, our computer model for
evaluating flow regimes does have the option for the operator to
input new (or revised) use rates if such information becomes
available in the future and is determined to be more reliable than
1985 figures.

135

aIAP'l'ER 7
DAY-USE RAFTER SURVEYS ARD RESULTS
Introduction
The organization of this chapter follows the same format as was used
in previous chapters for white-water boaters and anglers. Background
information about rafting in Glen canyon is present first, and a
discussion of the results of the attribute survey of Glen Canyon
rafters follows. The subsequent discussion of the
contingent-valuation survey builds on these findings and includes an
explanation of the relationship between respondents' surplus values
and flow levels. The chapter closes with a summary of the major
findings, highlighting ·the implications of these findings for future
analyses of flow release patterns from Glen Canyon Dam.
Background
A Glen Canyon raft trip is a one-day flat water trip on a 15-mile
section of the Colorado River through Glen Canyon. At low to
moderate flow levels, generally less than 29,500 cfs, raft trips
depart from a dock near the Glen Canyon Dam and float or motor
downstream to Lee's Ferry. The motor is used primarily to maneuver
the boat and to stay on schedule at relatively low flow levels. At
flow levels above 29,500 cfs, which typically occur when the jet
tubes at Glen Canyon Dam are open, trips depart from Lee's Ferry and
motor upstream before floating back downstream. An alternate trip
route is used at these high flow levels because it is not safe to
depart from the base of the dam due to the volume and turbulence of
the water released from the jet tubes. Most trips departing from
Lee's Ferry do not go all of the way up the river, and passengers do
not get a view of Glen Canyon Dam from the river.
All Glen canyon raft trips have a professional guide to run the raft
and explain the attractions along the river to passengers. Rafts
departing from the clam have a capacity of 20 passengers, while rafts
leaving from Lee's Ferry can only carry 10 passengers. This
difference in raft capacity occurs because the outer pontoons are
removed from the rafts on trips departing from Lee's Ferry to reduce
the resistance from the water while motoring upstream.
Wilderness River Adventures is the only concessionaire authorized by
the National Park Service to provide Glen Canyon raft trips. The
Fred Harvey Transportation Company also participates in the provision
of these trips by busing raft passengers from the South Rim to Glen
Canyon for a raft trip.

136

Commercial Glen Canyon raft trips were provided from the late 1960's
through 1982 by the Fort Lee Company. In Wilderness River
Adventures' first year of operation (1983), they carried about 500
passengers through Glen Canyon. Since that time the number of
passengers has increased to 8,469 in 1985, and as of August 31, 1986,
they had carried 9,497 passengers for the current calendar year.
Wilderness River Adventures' sales brochure describes a Glen Canyon
raft trip in the following manner.
"Along the 15-mile route, you'll float leisurely past the
multi-colored sandstone cliffs of Glen Canyon. Then take
time out to explore ancient Indian petroglyphs on the
canyon's sandy beaches. And·enjoy a tasty picnic lunch
along the shores of the Southwest's most celebrated river.
End your float adventure at historic Lee's Ferry.
Wilderness River Adventures includes everything you need for
a safe and pleasurable raft trip. Our tour buses take you
from Page to the launch site and pick you up again that
afternoon. Our skilled, well-versed crew guides you in
spaciously comfortable and sturdy neoprene rafts."
One might classify a Glen Canyon raft trip, therefore, as a leisurely
one-day float trip on the Colorado River.
Attribute Survey Procedures
The final attribute survey (see Appendix I) was sent to a
sample of 300 individuals who took a Glen Canyon raft trip during the
months of April through October, 1985. Names and addresses for the
sample were obtained from Wilderness River Adventures and the Fred
Harvey Transportation Company. The lists of names provided by these
two companies, however, do not represent all of the 1985 Glen Canyon
raft trip passengers.
Sampling.

Wilderness River Adventures only keeps a record of the name and
address of the one individual per group that registers for a trip.
This listing, then, represents only group leaders or heads of
households, depending on the composition of the group taking the
trip. Consequently, not all individuals from each group were
eligible for selection in the sample.
The Fred Harvey Transportation Company provided a list of names and
addresses for individuals who made advance reservations for their
raft trip or who filled out a comment card after their trip. The
advance registration list, like the list of names obtained from

137

Wilderness River Adventures, only represented group leaders and heads
of households. The comment cards were simply a form for customers to
fill out on which they indicated their satisfaction with the Fred
Harvey Transportation Company bus ride to and from Glen Canyon and
the Wilderness River Adventures raft trip through Glen Canyon. The
list of individuals who filled out the comment cards was not
restricted to heads of households and group leaders, however, they
were a self-selected group and not necessarily representative of all
passengers.
Despite its shortcomings, the sampling procedure provided a list of
Glen Canyon raft trip passengers covering an entire rafting season.
Sampled passengers would have the necessary experience to evaluate
such a trip. Though not ideal, we felt this sample was adequate to
identify the important attributes of a Glen Canyon rafting
experience.
The sample was stratified into two groups according to the place of
departure for the raft trip (the base of Glen Canyon Dam or Lee's
Ferry). The purpose of this stratification was to identify whether
large differences in flow level, as determined by place of departure,
affected respondents' satisfaction with their Glen Canyon raft
trips. All respondents, regardless of the origin of their raft trip,
received the same survey.
A summary of the flow levels experienced by attribute survey
respondents is presented in Table 7-1. Overall, 1985 was a year of
medium to high flows, and as a result, none of the respondents
experienced an average daily flow of less than 10,000 cfs. Glen
Canyon raft trips left from the dam during all months of the sampling
period (April through October) except June. Lee's Ferry launches
occurred only in May and June, 1985.
Table 7-1.

Daily Flow Levels Experienced By Respondents to the
Glen Canyon Ha.f'ter Attribute Survey

Flow
Dam
Average flow
Average high flow
Average low flow

Trip Origin
Lee's Ferry

25,500 cfs
27,700
22,400

39, 100 cfs
40,700
36.900

138

..
Response Rate. In November, 1985, respondents in the sample received
their first survey contact. Overall, 220 completed questionnaires
were returned, 73 percent of the total sample. Thirty questionnaires
were returned incomplete because the recipient had registered in
advance but did not take a Glen Canyon raft trip, and an additional
11 questionnaires were returned as undeliverable. The response rate
as a percent of all deliverable questionnaires and individuals who
had actually taken a Glen Canyon raft trip was 85 percent (see Table
7-2). The response rate did not differ by the origin of respondents'
raft trips. The data presented in this section are based on the
responses of 215 Glen Canyon raft trip passengers: 108 passengers
from trips leaving from the 9am, and 107 passengers from trips
1
departing from Lee's Ferry.Table 7-2.

Glen Canyon Rafters' Attribute Survey Response Rate

Surveys
Completed survwys
Not applicable
Undeliverable
Surveys not returned
Refusals
TOTALS

Percent of
All Surveys

Percent of
••
Deliverable Surveys

73%

85%

10
4
10

11

--3

100%

4
100%

• These are the 30 respondents who made advance rese_rvations, but
did not take the trip.
••As noted in the text, 11 questionnaires were returned as
undeliverable, and an additional 30 questionnaires were returned
because the individual had not taken the trip. Thus, the
percentages in this column are computed from a sample size of 259
rather than 300 •

.1/ Five surveys were received after the analysis of responses for
this survey was completed. While these five are included in the
response rates for Table 7-2, they are not included in the results
reported here.

139

Attribute Survey Results
The ten most frequently cited attributes that would contribute to an
excellent or perfect one-day raft experience are listed in Table
7-3. A good guide, rapids and faster water, good weather, and good
food were the attributes listed most often. It should be emphasized
that this question asked which factors would contribute to an
excellent or perfect trip, and it did not measure whether respondents
actually experienced the attribute on their trip.
Table 7-3.

Attributes That Contribute Most To An Excellent or Perfect
One-Day Raf't Trip

Attribute
Good guide
Rapids and faster water
Good weather
Good food
More stops and activities
Fewer people
Nice co-travelers
Shorter bus ride
Full r~ to dam
Shade at picnic area

Proportion Citing Attribute
Dam
Lee's Ferry
28%
23
15a **
10a
10
8
7
5a
2
,a

•

37%
25b
26b
20
6
6
8
6b
9b
7

* Some

respondents listed as many as four attributes that would
contribute to an excellent or perfect trip for them. Thus, the
percentages listed reflect the relative number of respondents
citing a particular attribute and, as a result, the columns do not
add up to 100 percent.

**

Statistics sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different
superscripts are statistically different at the 0.05. For
example, if two statistics both have an "a" in their superscripts,
then they are not statistically different at the 0.05 level.
Superscript notation is not used when statistical differences are
not identified between any of the three groups of respondents.

140

Only two of the attributes listed in Table 7-3 appear, at face
value, to be sensitive to flow levels: rapids and faster water and
a full run up to the dam. However, further consideration suggests
that rapids and faster water are not affected by flow levels. Host
of the respondents experienced relatively high flow levels, 21,000
to 45,000 cfs on average, and it is unlikely that there would be an
opportunity for faster water in most years. In addition, there are
no rapids in Glen Canyon. Iri short, there is not much that can be
done to provide a white-water rafting experience in Glen Canyon.
The desire by some passengers who took a trip starting at Lee's
Ferry for a full run upriver to the dam is a flow-sensitive
attribute. Glen Canyon raft trips are required to depart from Lee's
Ferry at relatively high flow levels, and most of these trips· do not
motor far enough upstream for passengers to view the dam from the
river.
The ten most frequently cited attributes that would contribute to a
poor raft trip are listed in Table 7-4. Five of these attributes
may be related to flow levels: too many people, boredom, no
white-water, no stops, and too much motoring. Contacts with too
many people may be related to flows in that rafts leaving from the
dam have a capacity of 20 passengers while the rafts departing from
Lee's Ferry have a capacity of only 10 passengers, requiring twice
as many boats to carry the same number of passengers. Because we
did not find a significant difference in the proportion of
respondents citing this attribute between the two types of trips, we
concluded that "too many people" was not related to flow levels.
Boredom appeared to be an attribute that would be more of a concern
to passengers from Lee's Ferry trips since they saw the same scenery
twice. An analysis of the percentage of respondents citing this
attribute, however, revealed that those leaving from the dam, rather
than those leaving from Lee's Ferry, were more likely to cite
boredom as a negative attribute. As a result, we concluded that
this attribute is also not related to flow levels.
For reasons already stated, the absence of white-water or rapids
should not be considered as a flow-sensitive attribute. Finally, no
stops and too much motoring were not considered to be flow-sensitive
attributes. Our understanding, from discussions with
representatives of Wilderness River Adventures, is that all raft
trips make the same number of stops and do approximately the same
amount of motoring, regardless of the origin of the trip.

"

141

..
Table 7-11.

Attributes That Contribute Most To A Poor One-Day Raft Trip

Attribute
Poor weather
Poor guide
Poor food or no food
Disagreeable people
Boredom
No white-water
Too many people
No stops
Too much motoring
Poor boat condition

Pro12ortion Citing Attribute
Lee's Ferry
Dam
3ei ..
22
14
9a
9
9
8
6a
5
3

*

3eG
32
11
9b
2
4
12b
1
1
6

* Some

respondents listed as many as four attributes that would
contribute to a poor trip for them. Thus, the percentages
listed reflect the relative number of respondents citing a
particular attribute and, as a result, the columns do not add
up to 100 percent.

**

Statistics sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different
superscripts are statistically different at the 0.05. For
example, if two statistics both have an "a" in their superscripts,
then they are not statistically different at the 0.05 level.
Superscript notation is not used when statistical differences are
not identified between any of the three groups of respondents.

The attribute survey dealt with respondents' knowledge of the flow
level on the day of their raft trip. Nearly everyone who responded
(94 percent) did not know the expected water level on the day of
their trip (see Table 7-5). Of those who did know the expected water
level, none said that it influenced their decision to take the trip.
These results are exactly the same across trip origins.

142

Table 7-5.

Respondents' Knowledge of' the Expected Water Level

Response

Proportion
Who Knew Water Level
Dam
Lee's Ferry

Did not know expected water level

94%

94%

Knew expected water level
- influenced decision to
take raft trip
- did not influence
decision to take raft trip

0

0

6

6

Some respondents were able to express a preference for a general flow
level. These preferences are reported in Table 7-6. Note, however,
that nearly half of all respondents (45 percent) did not state a
preference. An interesting result presented in Table 7-6 is that
only three percent of the respondents leaving from the Dam would have
preferred a lower water level than they experienced, while 16 percent
of those leaving from Lee's Ferry would have preferred lower water.
Table 7-6.

Respondents' Evaluation of the Water Level During Their
Raf't Trip

Response

Proportion
Checking Response Category
Dam
Lee's Ferry

Preferred iower water
Preferred current level
Preferred higher water
Don't know or doesn't matter
• Statistics sharing a common superscript are not statistically
different at the 0.05 level. Statistics with different superscripts
are statistically different at the 0.05. For example, if two
statistics both have an "a" in their superscripts, then they are not
statistically different at the 0.05 level. Superscript notation is
not used when statistical differences are not identified between any
of the three groups of respondents.

143
These two proportions are statistically different. In addition,
respondents from the dam were significantly less likely to state a
preference for a flow level. Respondents who left from Lee's Ferry
were also more likely to prefer higher water than were those leaving
from the dam. These results suggest that respondents whose trips left
from the dam may have been somewhat more satisfied with the flow level
they experienced than were respondents whose trips left from Lee's
Ferry, but that respondents from Lee's Ferry trips were not unanimous
in their preferences for a flow level.
In summary, we concluded that none of the attributes of a Glen Canyon
raft trip listed by respondents were sensitive to river flow levels.
However, there was some tentative evidence that the origin of a raft
trip may be a flow-sensitive attribute. That is, re~pondents whose
trips left from the dam appeared to be slightly more satisfied with
the flow level they experienced than were those who took a trip
starting at Lee's Ferry. The primary objective of the
contingent-valuation survey of Glen Canyon rafters, therefore, was to
test for a statistically significant difference in surplus values
according to the origin of the raft trips.
Contingent-Valuation Survey Procedures
Sampling. The sampling frame used for the Glen Canyon Day-Use Rafter
CV survey was the same as was used for the attribute survey. The same
sources were used to identify users from 1985, so the same limitations
on the generalizability of survey results also exist for this sample.
The final contingent-valuation survey (see Appendix J) was sent to a
sample of 300 passengers.

As in the attribute survey, the sample was stratified into two groups
according to the point of departure for the raft trip. This
stratification allowed us to test for differences in respondents'
surplus values according to the origin of their raft trips. A summary
of the flow levels experienced by respondents to the
contingent-valuation survey is presented in Table 7-7. The flow
levels are comparable to what was reported in Table 7-1 for
respondents to the Glen Canyon Rafters' Attribute Survey since the
same sampling frame was used for both surveys.
Table 7-7.

.

Daily F1.ov Levels Experienced by Respondents to the
Glen Canyon Raf'ter Contingent-Valuation Survey

Flow
Average flow
Average high flow
Average low flow

Trip Origin
Dam
Lee's Ferry
25,600 cfs
28,000
21'9 00

39,200 cfs
41'000
36,300

144
Response Rate. For this survey, respondents were first contacted in
July, 1986. Overall, 211 completed questionnaires were returned, 70
percent of the total sample. Twenty-seven questionnaires were
returned incomplete because recipients made reservations in advance
for a Glen Canyon raft trip, but did not take the trip. The response
rate as a percent of all deliverable surveys and all individuals who
had actually taken a Glen Canyon raft trip was 82 percent (See Table
7-8). The response rate did not differ significantly by the origin
of respondents' raft trips.
Table 7-8.

Glen Canyon Raf'ters CY Survey Response Rate

Percent of
Total Surveys
Completed surveys
Undeliverables*
Not applicable
Refusals
Surveys not returned
TOTALS

* Respondents

71%

Percent of
**
Deliverable Surveys

82%

4
9
2
~

100%

2

--1§.

100%

who pre-registered but did not actually take a raft

trip.

**

As noted in the text, 27 questionnaires were returned because the
individual had not taken the trip. In addition, 12 questionnaires
were undeliverable. Thus, the percentages in this column are
computed from a sample size of 261 rather than 300.

The CV results presented in this section of the report are based on
the responses of 200 Glen Canyon raft trip passengers: 104
passengers from trips leaving from the Dam, and 96 passengers from
trips departing from Lee's Ferry. This number differs from the
number of completed questionnaires received because the responses of
three respondents who were younger than 18 years of age were not
used, and eigh~ questionnaires were received after the data analysis
was completed.-1

~/

Eight surveys were received after the analysis of responses for
this survey was completed. While these eight are included in the
response rates for Table 7-8, they are not included in the results
reported here.

145

Contingent-Valuation Survey
Given that the point of trip departure was the only important
attribute identified by respondents as being potentially flowsensitive, the contingent-valuation survey was designed to determine
whether respondents' surplus values would vary according to the
origin of their raft trip. Respondents were only asked to value
their actual trip. Scenarios were not used because the attribute
survey results indicated that Glen Canyon rafters were relatively
insensitive to flow levels, and a sufficient number of observations
on individuals from both types of trips could be obtained to make
inferences about differences in surplus values according to this
dichotomy.
All respondents were asked to value their actual trip using the same
question format. Responses were then tested to determine whether a
statistically significant difference existed between Lee's
Ferry passengers and individuals who left from the dam. If a
statistically significant difference was identified, it would be
possible to conclude that large differences in flow levels do affect
respondents' surplus values.
The mean actual trip expenditures across all respondents was $59.
Average expenditures for passengers on trips departing from the dam
and for Lee's Ferry passengers were $58 and $60 per trip,
respectively, an insignificant difference. These expenditures
included the payment to the rafting company, the purchase of any food
and beverages for the trip not provided by the rafting company, and
money spent on personal items such as suntan lotion or film for a
camera.

•

The calculated surplus value for all respondents, in excess of the
$59 they spent for their raft trip, is $26. This result implies
that, on average, the cost of a Glen Canyon raft trip would have to
increase by at least $26 before a typical respondent would choose not
to take the trip. Analysis of the surplus values indicated that
there was not a statistically signifi~nt difference depending upon
the origin of trip (See Appendix 0).
Thus, we concluded that
surplus values for a Glen Canyon raft trip are not affected by gross
differences in flow release patterns from Glen Canyon Dam. This
result indicates that if flow levels are such that raft trips in Glen
Canyon are feasible at all, surplus values are constant across a
variety of flow regimes.

1/ The calculated Chi-square statistic is 2.88 with 2 degrees of
freedom, indicating that the null hypothesis of no difference can
not be rejected at the 0.10 level.

146

Summary
The results of the attribute survey provided tentative evidence that
the only flow-sensitive attribute of a Glen Canyon raft trip may be
the origin of a respondents raft trip, i.e., whether an individual's
trip departs from the base of Glen Canyon Dam or Lee's Ferry. Given
this finding, the CV survey was designed to test for differences in
respondents' surplus values according to the origin of their raft
trip. The difference in surplus values was not statistically
significant. Based on these analyses, we have concluded that flow
levels, whether constant or fluctuating, do not affect surplus values
for a Glen Canyon rafting experience.
It is our understanding, from discussions with representatives of
Wilderness River Adventures and the Glen Canyon National Recreation
Area, that the rafts used in Glen Canyon require two feet of water to
run the river. This minimum water level would require a flow of
1,000 to 3,000 cfs. At the other end of the continuum, the maximum
flow level for Glen Canyon raft trips is 46,000 cfs. This maximum is
set by determining a break even number of passengers and then
calculating a flow level beyond which the boat motors would not have
enough power to carry this number of passengers upstream from Lee's
Ferry. These minimum and maximum flows, from 1,000 to 46,000 cfs,
cover the range of flow levels which the GCES were assigned to
evaluate. Therefore, for all flow regimes that will be evaluated in
the subsequent analyses in Chapter 8, Glen Canyon rafters' surplus
values will be constant at $26 across all relevant flow levels.

147

CHAPTER 8
EVALUATION OF ABBUAL FLC7i REGIMES FROM GLEB CABYON DAM
Introduction
Each GCES research team was asked to evaluate five annual flow regimes
specified by the Bureau of Reclamation. The objective was to examine
the effects of the specified release patterns on the resources being
studied and to provide each research team with a basis for proposing
alternative flow regimes. These analyses also help tie together the
various components of the GCES by giving all teams the same flow
regimes to assess.
The five flow regimes all involve 8.25 million acre-feet per year of
water to be released. Actually, both recent experience and
expectations regarding water use in the Upper Colorado River Basin
states over the next several decades indicate that annual releases
exceeding 8.25 million acre-feet will be common in future years. In
addition to the five regimes with releases of 8.25 million acre-feet,
two-high water flow regimes will be considered in this chapter. The
goal will be to gain insights to the effects of high flows on
recreation benefits. Finally, a flow regime that maximizes recreation
benefits will be designed as a benchmark for comparison with more
realistic flow regimes.
The objective of this chapter is to report how each of the flow
regimes affect Glen Canyon fishing and Grand Canyon white-water
boating. Since Glen Canyon day-use rafters' surplus values are
constant across the flow patterns to be evaluated by the GCES, they
will not be considered in this analysis.

•

At the beginning of this chapter, we review the flow value curves for
constant and fluctuating flows for both anglers and white-water
boaters. This review is followed by a section describing 1985 use
rates for Glen Canyon anglers and Grand Canyon white-water boaters.
The flow value functions and use rates are used to compute annual
recreation benefits for each of the flow regimes. The results of this
analysis are discussed and the effects of these flow regimes on angler
and white-water boater surplus values are identified. Finally, we
close with some suggestions regp.rding flow regimes that would be most
conducive to Glen Canyon angling and Grand Canyon white-water boating
experiences.
Flow Value Functions
In the preceding chapters we developed flow value functions for Glen
Canyon anglers and Grand Canyon white-water boaters. For the latter

148

group, separate functions were computed for commercial passengers and
private boaters. For each group, a constant-flow function and a
fluctuating-flow function were presented. A constant flow was defined
as any average daily flow where the daily fluctuations would be less
than 10,000 cfs. A fluctuating flow, in contrast, is an average daily
flow which is associated with daily fluctuations of 10,000 cfs or
more. This distinction is based on the results of the attribute
surveys which indicated that a majority of respondents notice daily
fluctuations in flow levels only when they exceed 10,000 cfs.
Moving to the flow value functions, Figure 8-1 shows that constant
flow surplus values for commercial white-water passengers range from
$47 per trip at 1,000 cfs to a maximum of $898 at 33,000 cfs, before
declining to $732 at 45 ,000 cfs. Priva·te white-water boater surplus
values rise from $21 per trip at 1,000 cfs to a maximum of $688 at
29,000 cfs, and then de.cline to $376 at 45,000 cfs. Thus, commercial
passenger and private boater preferences for constant flow levels, as
measured by surplus values, are quite similar, with maximum values
occurring at 33,000 and 29,000 cfs, respectively. Private boater
surplus values, however, decline faster at flow levels above their
optimum.
Figure 8-1 also shows the white-water boater flow value functions for
fluctuating flow levels. Surplus values range from $198 and $224 per
trip at 3,000 cfs to $446 and $373 at 25,000 cfs for commercial
passengers and private boaters, respectively. At relatively low flow
levels, less than an average daily flow of 10,000 cfs, fluctuating
flow surplus values are higher than the corresponding constant flow
values, while the converse holds at moderate flow levels above 10,000
cfs. Daily fluctuations of 10,000 cfs or more do not occur with
average daily flows of less than 3,000 cfs or more than 25,000 cfs.
Flow value functions for anglers are shown in Figure 8-2. For
constant flow levels, surplus values rise from $56 per trip at 1 ,000
cfs to a maximum of $126 at 10,000 cfs and then decline to $94 at
25,000 cfs and $38 at 45,000 cfs. The angler flow value function for
fluctuating flows rises from $77 per trip at 5,000 cfs to $87 at
10,000 cfs, and then declines to $68 per trip at 25,000 cfs.
Figure 8-3 examines the flow value functions for constant flows for
all recreationists simultaneously. This figure reveals the
differences in flow preferences between white-water boaters and
anglers. While anglers attain a maximum surplus value per trip at
10,000 cfs, the white-water boaters attain their maximums at 29,000
and 33,000 cfs, respectively, for private boaters and commercial
passengers. Similar differences in preferred flows exist for
fluctuating flows (Figure 8-4). It is also worth noting that surplus
values per trip are much higher for white-water boating than fishing,
particularly at moderate and relatively high flow levels. This

149

..

Figure

8-1

White-Water Boater
Flow Value Functions

$900
800

,, ,, ----

700

,,···

-- --

.. -······ ........ .

... ..

······ ...

/
600

/,/

Surplus
Value
500
Per Trip

400

,! , ',

'

300

'
'
''

'

,'

.ef"/
'
'
,• /

200

0.

100

•'

,. •'

0
0

5

10

15

20

25

30

35

40

45

Average Flow Level in cfs
(x 1000)
Private
Boater
Constant Flow
Values

•

Commercial
Passenger
Constant Flow
Values

·•- Private
Boater
Fluctuating
Flow Values

·D • Commercial

Passenger
Fluctuating
Flow Values

50

150

Figure

8-2

Angler
Flow Value Functions

$900
800
700
600
Surplus
Value
Per Trip

500
400
300
200

...

100

~~···

.....
•·· ....... ················•
-o

0 ............. .

···········•········

0
0

5

10

15

20

25

30

35

40

Average Flow Level in cfs
(x

·O- Fluctuating Flow Values

1000)
·•·· Constant Flow Values

45

50

151

Figure

8-3

Flow Value Functions
(Constant Flows)

$900
800
700
600
Surplus 500
Value
Per Trip 400

300
200

-------..------·-

100
0
0

5

10

15

20

25

30

35

40

•
45

Average Flow Level in cfs
(x

Private Boater
Values

•

1000)

·•- Angler Values
Commercial
Passenger Values

50

152

Figure

8-4

Flow Value Functions
(Fluctuating Flows)

$900
800
700
600
Surplus
Value
Per Trip

500
400

...

300

.. --··
..
..
..
.. ..

.o
..
..
..
..

.

200
100
0
0

5

10

15

20

25

30

35

40

45

Average Flow Level in cfs
(x 1000)

·•- Private Boater
Values

-a·

Commercial
Passenger Values

·0 - Angler Values

50

153

reflects in part the larger number of days per trip for white-water
boaters (8.8 days per trip for commercial passengers and 16.4 days per
trip for private boaters) compared to fishing (2. 5 days per trip).
The relationship between surplus values and average flow levels is one
component of an analysis of the effects of annual flow regimes on
angling and white-water boating. A second and equally important
component of this analysis is the number of trips for each group.
That is, in order to calculate the total recreation benefits for each
user group as well as all recreation groups combined, it is necessary
to know how many trips were taken in 1985.
White-Water Boater and Angler Use Rates
Information on 1985 use rates was obtained from Grand Canyon National
Park for white-water boaters and from Glen Canyon National Recreation
Area for anglers. These figures, as shown in Table 8-1, indicate that
the majority of white-water trips in 1985 occurred in the months of
May through September, while the primary angling seasons occurred in
the fall (September thru November) and winter (January thru April).
Patterns observed in 1985 were fairly typical, although as noted
previously, total participation in fishing was much lower in 1985 than
in previous years.
The difference in seasons of peak use are important given the
substantial disparity in flow preferences of the two user groups.
Moderate constant flows around 10,000 cfs would be conducive to
angling in the fall and winter months and would not adversely affect
the great majority of white-water boaters. In contrast, higher flows
of 25,000 to 30,000 cfs or more during the summer months would be
conducive to white-water boating without affecting the great majority
of fishing trips •

.

154

Table 8-1.

Month
January
February
March
April
May
June
July
August
September
October
November
December
TOTALS

Grand Canyon White-Water Boating and Glen Canyon Angling
1985 Use Rates (Trips)

a
White-Water Tri12s
Commercial Passengers
Private Boaters

Angleb
Trips

0
0
73
205
1, 445
3,027
2,528
2,684
1,218
184
18
0

0
21
79
240
339
289
385
355
356
233
57
14

952
546
652
697
426
298
273
266
591
583
515
270

11,382

2,368

6,069

~/ Derived from National Park Service launch records.
_QI Glen canyon National Recreation Area estimates the number of

angler-days of participation each year. To get angler trips, we
divided angler-days by 2.5 days per trip, the average trip length as
determined in the Angler CV Survey.

Annual Flow Regimes Involving Releases of 8.25 Million Acre-Feet
The five annual flow regimes devised by the Bureau of Reclamation each
specified average monthly flow levels and the ranges of daily
fluctuations in flows. These five alternative flow regimes were
designed to represent a variety of possible flow patterns that adhere
to the physical, legal and administrative constraints of operating
Glen Canyon Dam. Current water law requires that, on average, at
least 8.25 million acre-feet per year must be released from Glen
Canyon Dam. All five flow regimes involve annual releases of 8.25
million acre-feet. To simplify the analysis, flow patterns were
assumed to be the same for all days within each month, but were
allowed to vary from month to month.

155

The first three alternatives represent low to moderate average flows
in all months with alternative types of fluctuations in daily flow
levels (Table 8-2). Alternative 1 represents constant monthly flows
ranging from 8,300 to 14,600 cfs. Such a flow pattern would result if
the power plant was operated to generate base-load power. This
alternative would be quite conducive to angling in Glen Canyon since
angler surplus values are highest at constant flows of about 10, 000
cfs, and values decline relatively slowly as flows increase or
decrease from this point (see Figure 8-2).
The second flow regime also represented low to moderate average daily
flows from 8,300 to 17,000 cfs, but with severe daily fluctuations.
During June through August, daily flows could range from 3,000 to
33,500 cfs, and during the rest of the year daily flows could range
1,000 to 33,500. This is the type of daily flow pattern that would
occur if the power plant were to be fully utilized for peak-power
generation. Alternative 3 presents average daily flows that are the
same as Alternative 2, but the fluctuations in daily flows would be
moderate, ranging from 8,000 to 25,000 cfs, throughout the year.
Alternative 3 represents a compromise between base-load and peak~load
generation. For purposes of analysis, however, Alternatives 2 and 3
are exactly the same since any flow level with daily fluctuations in
excess of 10,000 cfs is classified as a fluctuating flow for purposes
of evaluating recreationists' preferences.
The fourth alternative represents extremely low average daily flows
with large daily fluctuations throughout most of the year. The
exception is high constant flows of 25,000 cfs during the prime
white-water boating months of June through August. This regime mixes
base-load generation in the summer with peak-load generation for the
rest of the year. Of the four alternatives we have discussed to this
point, it would appear that this alternative would present the best
flow regime for white-water boaters.
The fifth, and final, alternative has relatively low average daily
flows with daily fluctuations throughout the year. However, the range
of fluctuations only exceeds 10,000 cfs in the months of April through
October. Thus, anglers would receive constant flows throughout most
of the prime fishing months.

"

In the next section, we will combine the surplus values and use rates
for each user group to evaluate each of these five annual flow
regimes. This will be done to determine which specific regime
provides the highest recreation benefits to each group individually
and all groups combined.

Table 8-2.

ilteroatiYe Annual

Month

now

Alternative 2

Alternative 1
Average
flow

Regimes

Range
High
Low

Average
flow

Alternative 3

Range
High
Low

Average
flow

Alternative 5

Alternative 4

Range
High
Low

Average
flow

Range
High
Low

Average
flow

Range
Hildl
Low

January

14,600

14,600

14 ,600

14,500

33,500

1,000

14,500

25,000

8,000

11,400

33,500

1,000

8,000

10,000

6,000

February

12,200

12,200

12,200

9,500

33,500

1,000

9,500

25,000

8,000

5,400

33,500

1,000

8,000

10,000

6,000

March

8,300

8,300

8,300

8,300

33,500

1,000

8,300

25,000

8,000

4,900

33,500

1,000

8,000

10,000

6,000

April

10,300

10,300

10,300

9,700

33,500

1,000

9,700

25,000

8,000

5,000

33,500

1,000

12,600

33,500

1,000

May

10,000

10,000

10,000

9,000

33,500

1,000

9,000

25,000

8,000

4,900

33,500

1,000

12,200

33,500

1,000

June

10,400

10,400

10,400

9,800

33,500

3,000

9,800

25,000

8,000

25,000

25,000

25,000

13,400

33,500

3,000

July

12, 750

12,750

12,750

17 ,000

33,500

3,000

17,000

25,000

8,000

25,000

25,000

25,000

17 ,000

33,500

3,000

August

14 ,400

14 ,400

14,400

16,500

33,500

3,000

16,500

25,000

8,000

25,000

25,000

25,000

16,500

33,500

3,000

September

10,000

10,000

10,000

10,200

33,500

1,000

10,200

25,000

8,000

8,400

33,500

1,000

13 ,400

33,500

1,000

October

9,900

9,900

9,900

9,200

33,500

1,000

9,200

25,000

8,000

4,900

33,500

1,000

9,300

33 '500

1,000

November

9,800

9,800

9,800

9,400

33,500

1,000

9,400

25,000

8,000

5,000

33,500

1,000

8,900

10,000

6,000

December

13,600

13 I 600 __13_._QQ_0_13_,4Q()__ - 33' 2QQ __ J. QOO

13,400

25,000

8,000

11,400

33.500

1,000

8,000

10,000

6,000

......
V1
(j\

i·

;.

....

157

Evaluation of Annual Flow Regimes for 8.25 Million Acre-Feet

•

To estimate the annual recreation benefits from the flow regimes, a
computer model was designed that incorporates the flow value
functions for white-water boaters and anglers. The model allows an
analyst to specify average monthly flow levels and to designate
whether the flow is fluctuating or constant. In addition, the
analyst can choose to either apply 1985 use rates in the calculation
of benefits or to specify other monthly use rates that may be
appropriate for the analysis being conducted.
The calculations reported here are based on 1985 use rates. Thus,
recreation benefits were calculated by specifying the average daily
flows for each month of the year and whether the flows were constant
or fluctuating. The model then calculates surplus values per trip
from the flow value functions for each month, multiplies these values
by the number of trips for that month in 1985 for each group and SllillS
the monthly benefits to obtain an overall monthly benefit for all
groups combined. Aggregate annual recreation benefits are derived by
summing the monthly totals for each group individually, and all
groups combined. A diagram representing the process for calculating
total recreation benefits is presented in Figure 8-5.
Overall, Alternative 4 provides the largest aggregate recreation
benefits for all three groups of recreationists combined, a total of
about $9 million annually (Table 8-3). However, this alternative
produces the lowest level of recreational benefits, about $0.5
million annually, for anglers.

Figure 8-5
Calculation of Annual Recreation Benefits
for Alternative Flow Regimes

Specification of
Average Monthly Flow
Levels and Designation
of Months with
Fluctuations in Excess
of 10,000 els

-

Calculation of Surplus
Values Per Trip for
Anglers and
White-Water Boaters
From Flow
Value Functions

Specification of
Monthly Use Rates
for Anglers and
White-Water Boaters
(Number of Trips)

1--'
U1

co

I·

"'

...

~

..

Table 8-3.

..

Evaluation e>r Flow Regiaes (Annual Flow

Alternative
Annual tyow
Regimes-

= 8.25

White-Water Boaters
Commercial
Private
Passengers
Boaters
Annual
Annual
Rank
Benefits
Rank
Benefits

Million Acre-Feet)

Anlders
Annual
Rank
Benefits

All
Recreationists
Annual
Benefits
Rank

1

2

$4 '277 ,990

5

$ 681, 178

1

$734,798

2

$5,693,965

2

4

3,846,816

3

706,345

3

508,499

4

5,061,661

3

4

3,846,816

3

706,345

3

508,499

4

5 ,061'661

4

1

7,478,301

1

1,017,161

5

482,451

1

8,977,913

5

3

4,126,982

2

723,729

2

584,146

3

5,434,858

a/
- Flow regime 1
Flow regime 2
Flow regime 3
Flow regime 4
September-May
Flow regime 5

= Low to moderate constant average monthly flow releases
= Low to moderate average daily flows with extreme fluctuations
= Low to moderate average daily flow with moderate fluctuations
= High constant summer flows; low average daily flows, and extreme
= Low

'fr

fluctuations from

to moderate average daily flows with extreme fluctuations from April-October.
......
lJl

\D

160

Alternative 1 produces the highest benefits for anglers ($0.7 million
annually) and the second highest total overall benefits ($5.7 million
annually), but provides the lowest total benefits for private
white-water boaters. These results reveal how commercial white-water
boater preferences, as measured by surplus values, dominate the
overall ranking of alternatives.
Examination of the rankings in Table 8-3 reveals some important
effects of fluctuating flows. Recall that flow Alternatives 1 and 2
have similar average daily flows. However, Alternative 1 involves
constant flows while Alternative 2 involves extreme fluctuations in
daily flow levels. Recreational benefits for Alternative 2 are about
11 percent (approximately $0.6 million annually) below those for
Alternative 1. At higher average daily flow levels during the peak
boating season the effect of fluctuations would be larger.
It is also helpful to examine how each group ranked the
alternatives. Let us start with the white-water boaters, remembering
that Alternatives 2 and 3 are treated as being the same. Alternative
4 has the highest recreational benefits for white-water boaters,
including both commercial passengers and private boaters. This
result holds due to the constant flow of 25,000 cfs during the months
where the heaviest use by white-water boaters occurs. The rankings
of the other alternatives differ between these two groups primarily
because private boater surplus values for fluctuating flows are
substantially higher than those for constant flows which average less
than 10,000 cfs.
Alternatives 1 and 5 generate the highest annual recreation benefits
for anglers, because both have low constant flows around 10,000 cfs
during the months when most of the fishing occurs. Alternatives 2
and 3 have lower annual angler benefits due to the fluctuating flows
throughout the year. Alternative 4 produces the lowest total
benefits for anglers due to the low average flows with extreme daily
fluctuations during some of the good fishing months.
Alternative 4 produces the highest aggregate annual benefits when all
three groups of recreationists are combined, due to the dominance of
the commercial white-water boaters. The remaining four alternatives
produce similar levels of aggregate annual benefits for all
recreationists combined. The dominance of commercial white-water
boaters is also represented in the rank of alternatives for all
recreationists--the ranking for all recreationists is exactly the
same ranking as that for commercial passengers.

•

161

•
Three High-Water Flow Regimes
Trip values used to ev~luate the five 8.25 million acre-foot flow
regimes were derived from surveys of 1985 visitors. Monetary
benefits were calculated using estimated actual trips taken in 1985.
Thus, the annual benefits were calculated as if 1985 had involved
releases totaling 8.25 million acre-feet and this water had been
released during the year in the manner postulated in the flow regime
alternative being evaluated. Actually, not since 1982 have there
been hydrological conditions anywhere close to those postulated in
the five flow regimes. The 8.25 million acre-foot figure was used
because this is the minimum average annual release from Glen Canyon
Dam permitted under current water law. Now that Lake Powell has been
filled, many experts believe that substantially more than 8.25
million acre-feet will be available for release in most years until
well into the next century. The Upper Basin states are expected to
have surplus water for several decades yet, except during drought
periods. Furthermore, the Colorado River watershed produces greatly
varying runoff, depending on annual precipitation. Recent years
illustrate this well. During calendar year 1984, 20.8 million
acre-feet were released from Glen Canyon Dam. In calendar year 1985,
16.6 million acre-feet were released. Even when the Upper Basin
states are using their allocations fully, there will still be high
water years. Under current conditions, Bureau of Reclamation experte
predict that annual discharges from Glen Canyon Dam will exceed 8. 25
million acre-feet in about three years out of four.
In this section, we will explore the implications of high water for
recreation benefits. Evaluation of the 1985 release patterns is
particularly interesting since 1985 data were used for valuation and
participation. To assess the effects of extremely high water, 1984
will be evaluated as~well. A third flow regime will also be
developed which will be termed the "unconstrained optimal recreation
flow regime." That is, we shall ask what sort of flow pattern would
maximize recreational benefits given unlimited water and no other
objectives or constraints. This is not to suggest that the system
will be, or should ever be, operated solely for recreation. However,
it will give us a useful baseline figure for the maximum benefits
that could be earned if recreation were the only objective. Such a
standard of comparison provides a context for considering total
benefits from more realistic flow regimes.
In shifting from completely hypothetical flow regimes to the actual
years 1984 and 1985, several simplifications are necessary. We will
refer to the "1984 regime" and "1985 regime" to emphasize that the
dam releases being valued are simplified versions of what actually
happened. The valuation model holds average daily flows constant for
each entire month. It also defines each month as a fluctuating or

162

•
constant flow month using 10,000 cfs of variation per day as the
dividing line. To accommodate this simplified framework, the 1984
and 1985 average monthly flows will be assumed to be fixed average
daily flows for each month. To determine whether each month is a
fluctuating or constant flow month, daily maximum and daily minimum
flows were averaged over each entire month. If the average daily
maximum for a month exceeded the average daily minimum by more than
10,000 cfs, the month was defined as one with fluctuating flows;
otherwise it was defined as a constant flow month. Though resulting
values will not be as exact as those that would come from a more
complex model, they still should be good approximations of actual
benefits. Table 8-4 shows relevant data for the 1984 flow regime,
while Table 8-5 shows comparable figures for 1985. The total dam
releases equaling more than 20.8 million acre-feet in 1984 and
slightly less than 16.6 million acre-feet in 1985 were distributed as
shown in these tables. In the 24 month period, only April, October,
November, and December, 1985 qualified as fluctuating flow months.
All others were classified as constant flow months.
Table 8-6 shows the calculated benefits for the three recreation
groups separately and for all three groups combined under the two
flow regimes just described. In order to hold all effects except
flows constant, 1984 values are calculated using 1985 trips. This
makes the 1984 regime fully comparable to all other flow regimes in
this chapter. The results are quite interesting. Despite a
difference of more than 4 million acre-feet of water released in the
two flow regimes, the total benefits are quite close at just over $11
million. Furthermore, the total benefits of $11 million exceed total
benefits from all of the 8.25 million acre-foot regimes evaluated in
Table 8.3, where even the regime with the largest benefits
(Alternative 4) yielded only about $9 million. The reason for th~
favorable economic results for high water years is that they entail
fairly high constant flows in the May through September period when
so much white-water boating occurs.
Lower water would, of course, enhance fishing benefits. This is
apparent from comparing the 1984 or 1985 flow regimes where fishing
benefits are about $525,000 to Alternative 1, Table 8-3, where
fishing benefits are $735,000. At first glance, one wonders why the
1985 regime does not have significantly higher fishing benefits than
1984, given that average flows in October, November, and December are
much more favorable to fishing in 1985. Closer examination shows,
however, that any gains from lower average flows in these months are
more or less balanced by losses due to daily fluctuations in flow
levels.

'

\

Table 8-lJ.

...

Data tor Simplif'ied 198.la Flow Regime

Acre-Feet Per
Month

Average Flow
(CFS)

Average Maximum
(CFS)

Average Minimum
(CFS)

JANUARY

25,018

26,048

22,838

1,535,630

FEBRUARY

25,524

26,201

24 '091

1,465,638

MARCH

25,447

26,339

22,427

1,561,964

APRIL

26,825

27,702

24,738

1,593,416

MAY

40, 580

40,836

38' 174

2 ,490,854

JUNE

41,167

41,713

39 '728

2,445,347

JULY

35,387

36,678

34,725

2, 172,087

AUGUST

26,312

29' 166

23' 151

1,615,053

SEPTEMBER

25,042

25,817

23,358

1,487,545

OCTOBER

22,966

24' 151

21,253

1,409,696

NOVEMBER

25,635

26,165

24,435

1,522, 722

DECEMBER

25' 062

26' 127

22,918

1,53a~

Month

TOTAL

.

·~

20,838.313

I-'

O'I

w

Table 8-5.

J·

..;

Data ror Siaplif'ied 1985 Flow Begiae

j.

•

.,

Table 8-6.

Flow Regime

\~

Evaluation of 1984 and 1985 Flow Regimes

White-Water Boaters
Commercial
Private
Passengers
]3o(ltE:l!'~ _____
Anglers

1984

$9,578,038

$1,471,946

$525,591

1985

$9,436,994

$1,359,037

$524 '699

* Totals

do not equal the exact

SllD.

All
Re creationists

*
$11,320,727 *
~>11,575,576

of the components due to rounding.

I-'

0\
Ul

166

Finally, Table 8.7 looks at an unconstrained flow regime designed to
optimize recreational benefits. It assumes that water is available
when needed in any quantity at any time. Constant flows are optimal
in all months. The results indicate that 10,000 cfs would be optimal
in December, January, and February. This is to be expected, since
10,000 cfs is ideal for fishing and there is very little white-water
boating in these months. However, as soon as white-water boaters
begin to appear in modest numbers, their dominance becomes apparent.
The optimal flow for March, for example, is 26,400 cfs, despite the
fact that there were only 152 white-water trips in ~arch, 1985, while
there were 651 fishing trips. May, June, July, and August all have
optimal flows in excess of 32,000 cfs, reflecting the overriding
economic importance of commercial white-water boating.
Total benefits to all three groups combined are slightly less than
$12.4 million. This may be interpreted as an upper bound on
recreational benefits that the system could have produced. No matter
how the dam had been managed, the recreation benefits would not have
exceeded this figure.
It is also interesting to compare the maximum possible benefits of
$12.4 million with the benefits of $11.8 million and $11.3 million
for 1984 and 1985, respectively. Even though dam releases were
designed with little or no attention to recreation, recreational
benefits were surprisingly close to the optimum. Implicit in all of
this is the assumption that very high flows can be avoided. An
example of extremely adverse circu:nstances occurred in 1983 when the
reservoir filled completely and large amounts of excess water were
released through the spillways. At one point, the flow reached
112,360 cfs. Safety became a major concern and white-water boating
was halted for a time. Obviously, such floods are harmful to
recreation, and the extent of the damage would not be fully reflected
in a model such as ours that is based on averages and that is
calibrated up to only 45,000 cfs.

167

Table 8-7.

Month

The Unconstrained Optimal Recreation F1.ow
F1.ows Constant)

Average Daily Flows (cfs)

Regime~/

Acre-Feet Per Month

JANUARY

10,000

613,800

FEBRUARY

10,000

554,400

MARCH

26,400

1,620,432

APRIL

29,300

1,740,420

MAY

32,200

1,976,436

JUNE

32,800

1,948,320

JULY

32,600

2,000,988

AUGUST

32,300

1,982,574

SEPTEMBER

31,900

1,894,860

OCTOBER

29,400

1,804,572

NOVEMBER

24,500

1,455,300

DECEMBER

10,000

613,800
18,205,902

TOTAL

~/ Annual Benefits:

(All

Commercial White-Water Boating
Private White-Water Boating
Fishing

$10,197,556
1,590,312
571,264

TOTAL

$12,359,132

168

Summary
This chapter has combined values per trip reported by white-water
boaters and anglers, based on CV surveys of 1985 users, with the
number of trips taken by each group to evaluate various annual flow
regimes. Each regime portrays a simplified annual water release
pattern that could result from Glen Canyon Dam and Power Plant
operations, depending on hydrological conditions and operating
procedures. Five regimes that would each result in an annual release
of 8.25 million acre-feet were evaluated. Simplified regimes
patterned after 1984 (a 20.8 million acre-foot year) and 1985 (a 16.6
million acre-foot year) were also evaluated. Finally, an
unconstrained annual flow regime designed to optimize recreational
benefits was designed and evaluated.
With annual releases totalling 8.25 million acre-feet to work with,
white-water boaters and anglers compete for scarce water. From a
recreational perspective, the more water that can be allocated to
create high, constant flows in the primary white-water boating months
of May through September, the greater the recreational benefits.
Though not fully ideal from an economic viewpoint, this conclusion is
illustrated well by Alternative 4 (Tables 8-2 and 8-3) which would
have earned about $9 million under 1985 recreational conditions.
This would not leave sufficient water to provide preferred fishing
conditions during the rest of the year.
The best fishing regime (Alternative 1) would have earned only $5.7
million because there would not be enough water to provide good
white-water boating conditions. Full utilization of Glen Canyon
Power Plant for peak-power generation, as described in Alternative 2,
would have produced even lower recreational benefits ($5.1 million)
because of the adverse effects of daily fluctuations. Thus, when
modest amounts of water are available, dam operations can have a
major impact on recreation. Presumably this impact would be
accentuated in drought years when less than 8.25 million acre-feet
are available to be released.
While annual releases between 8.25 million acre-feet and 16.6 million
acre-feet were not explicitly analyzed, the recreational implications
of extra water are quite apparent. White-water boating and fishing
tend to occur at different times of the year. The more water there
is available, the more feasible it would be to accommodate both
groups through relatively high flows in the May through September
period and lower flows for the rest of the year.
Furthermore, extra water may make normal electricity generation and
recreation more compatible. Release of the extra water involves
periods of base-load power generation during the commercial

169

•
white-water boating season when there are large potential recreation
benefits to be earned from high constant flows. The implication of
extra water for recreation benefits are illustrated by the $11.6
million that would have been earned by the 1984 regime and the $11.3
million that would have been earned by the 1985 regime. These are
impressive benefit levels given that the maximum recreation benefits
under ideal and unconstrained conditions would be about $12.6
million.
High water, such as that found in the 1984 and 1985 regimes, does
create somewhat adverse conditions for fishing, but fishing benefits
still would have exceeded $0.5 million for both regimes. The flows
for the primary fishing months (September through April) were still
mostly in the 20,000 cfs to 30,000 cfs range. Though not ideal, the
"flatness" of the fishing flow value curves means that trips under
such flows are still fairly valuable, particularly if fluctuations
are avoided. High water certainly decreases fishing benefits, but
the effect was no more severe than, for example, adopting a peaking
regime in an 8.25 million acre-foot year.

170

.
CBAP'rER 9

COBQ.USIORS
The purpcse of this study was to assess the impacts of various
releases of water from Glen Canyon Dam on downstream recreation.
Minimal releases of 1,000 cfs are pcssible at any point in time
unless the reservoir is completely full. At the other extreme,
operating all eight turbines at Glen Canyon Power Plant releases as
much as 33,500 cfs. Furthermore, it is pcssible for flows to
fluctuate across this full range (1,000 to 33,500 cfs) on a daily
basis, if the power plant·were operated for maximum peak power
generation. Additional water can also be released through bypass
tubes and, if the reservoir is full, through spillways. Even after
allowing for technical and institutional constraints on dam
operation, many dam release strategies are feasible. The effects of
dam releases on downstream recreation may be very different depending
on which strategies are chosen in the future. The purpose of the
research reported here was to better understand and quantify these
effects. Some conclusions are drawn below.
White-Water Boating Conclusions
Conclusion 1: Glen Canyon Dam releases have substantial impacts on
white-water boating. The attribute survey of this group showed that
several important attributes of white-water trips are affected by
flows. Time at attraction sites and for layovers depends on the
speed of the current. The size and number of rapids are also
affected by dam releases.

Boaters, particularly those on commercial trips, enjoy fairly large
rapids that depend on substantial amounts of water in the river. At
relatively low or high flows, passengers--particularly those on
commercial oar trips--may have to walk around rapids, and this is
generally considered a negative attribute. High water may raise
concerns about safety in the minds of some boaters. The lack of
crowding is important to many boaters, and high water can al so
contribute to crowding at campsites and attraction sites. Litter is
a negative attribute that is reduced by occasional periods of very
high water. The attribute survey indicated a preference for medium
to high flows in the 16,000 cfs to 32,000 cfs range.
The survey of commercial trip guides and private trip leaders
indicated general agreement with the attribute survey results. For
example, guides and trip leaders agree that flows on the low and high
ends of the spectrum do increase the odds that passengers will have
to walk around rapids. Safety also becomes more of a concern at low
and high flows. Guides and trip leaders agree that low flows reduce
the time available for attraction sites and time in camp. High flows

171

•
create extra time for such activities. Guides and trip leaders also
believe that the number and size of campsites are limited at high
water. Overall, most guides and trip leaders have a preference for
medium to high water in the range between 16,000 and 30,000 cfs.
The contingent-valuation survey also supports the conclusion that the
white-water boating experience is affected by flows. For
constant-flow trips, where there were actual-trip data across a wide
range of flows, the average flow actually experienced was a
statistically significant and potent predictor of surplus values.
Surplus values are much larger for medium-high flows in the 25,000
cfs to 35,000 cfs range than for lower and higher flows. Fluctuating
flow values based on respondents' evaluations of scenarios also
support this relationship. Based on the economic analysis, the ideal
flow for commercial trips is about 33,000 ·cfs, while for private
trips it would be roughly 29,000 cfs.
Conclusion 2: Except at low average daily flows {less than 10,000
cfs), fluctuating daily nows are detrimental to white-water boating
when canpared to constant nows at the same average daily levels.

One of the primary attributes of a white-water boating trip is
experiencing the natural environment of Grand Canyon National Park.
Perceptible fluctuations in water (roughly speaking fluctuations of
10,000 cfs or more) make the canyon seem less natural to most
participants. Allowing for changes in water level makes camping and
mooring of boats for the night more difficult as well. Fluctuations
also increase the likelihood of arriving at rapids at disadvantageous
times when waiting for water level changes or walking around may be
necessary.
Guides and trip leaders in general agreed that lower fluctuations are
more desirable than higher fluctuations. Larger fluctuations tend to
reduce time available for stops at attraction sites and in camp,
increase the necessity of scouting rapids and checking boat moorings
during the night, and increase the difficulty of planning
itineraries. Guides find larger fluctuations more tolerable the
higher the average daily flow, but a large share of them would
consider the possible fluctuations between 1,000 cfs and 33,500 cfs
associated with full use of Glen Canyon Dam for on-peak power
generation "intolerable."

•

Not surprisingly, except for the low flows discussed below, when
trips with the same average daily flow were compared, those with
constant flows produce larger surplus values per trip than those with
fluctuating flows. For example, based on scenario values,
fluctuating flows around an average daily flow of 22,000 cfs would
produce a 27 percent lower surplus value for private trips and a 22
percent lower surplus value for commercial trips compared to constant
flow trips at that level.

172
At low flows, this relationship is reversed for commercial
passengers. Based on respondents' evaluation of scenarios,
willingness to pay is significantly larger for fluctuating flows
around an average daily flow of 5,000 cfs than for constant flows at
that level. Presumably, this reflects a desire to have higher water
for at least part of the day. It should be noted, however, that
there was no such difference for private trips. For the 5,000 cfs
scenarios, the private trip surplus values for constant and
fluctuating flows were statistically indistinguishable.
Conclusion 3: Annual white-water boating benefits could be enhanced
most by maintaining relatively high constant flows during the summer
months. About 67 percent of all white-water boating trips (both
commercial and private) occur during June, July, and August. High
stable flows in the range of 25,000 to 33,500 cfs would produce high
benefits per trip. For example, commercial benefits per trip at
33, 000 cfs are nearly $900, whereas ·they fall to only a bout $300 at
10,000 cfs. Thus, high flows in the summer could increase commercial
white-water boating benefits by as much as three times or more
compared to low flows. Though slightly less pronounced, the same
conclusion applies to private trip benefits. Further gains could be
achieved through high and stable flows in ~~Y and September when an
additional 25 percent of the trips occur.
Conclusion .IJ: Very high flows (in excess of .IJ0,000 cfs) will reduce
white-water benefits, particularly during June, July, and .August.
The flow value functions for both private and commercial trips
decline sharply after their respective maximum points. For example,
commercial benefits per trip at 45,000 cfs are almost $170 lower than
the maximum of nearly $900 per trip at 33,000 cfs. This represents a
decline of nearly 19 percent. The decline is even steeper for
private trips. From a maximum of $688 reached at 29,000 cfs, private
benefits decline to $376 per trip at 45,000 cfs, a decline of 45
percent (see Figure 5-1). Such losses would be largest in the summer
months of June, July and August, but this potential loss in benefits
is also a concern in May since May is a month when the bypass tubes
may be utilized along with full capacity operation of the power plant
in years of high upstream run-off. Full operation of the power plant
and bypass tubes involves flows of around 50, 000 cfs.
Conclusion 5: Loss of large numbers of camping beaches would have a
substantial adverse impact on white-water boating. Sandy beaches
provide the best camping sites in the canyon because they are
relatively flat and free of dense vegetation and rocks. Currently,
some stretches of the river have few beaches. This requires some
planning on the part of guides and trip leaders to assure passengers
a good place to spend the night in these areas. Time at attraction
sites may be cut short or some sites may have to be skipped
altogether. Parties may have to leave earlier in the morning or
spend more time rowing or motoring to reach good campsites in certain
stretches of the river.

•

173

..
Some concerns have been voiced about the effects of dam operations on
beaches. Glen Canyon Dam has cut off a source of sediment for beach
deposits and high flows may increase erosion rates for existing
beaches. Other research teams within the Glen Canyon Environmental
Studies are examining the processes which affect beaches and the
extent to which beaches are threatened under alternative flow
regimes. Their results, however, are not available to us at this
time. Our results do indicate that if dam operations result in
substantial reductions in the number or size of beaches in the
future, the damage to recreational benefits would be substantial. In
economic terms, trip values could well be reduced by one-third or
more.
Furthermore, an important difference exists between the short-term
effects of flows, as discussed in the previous conclusions, and the
potential effects of beach losses. With the possible exception of
litter cleanup from high flows, all of the effects of flows discussed
previously are felt only by those engaged in recreation at the time
the flows are occurring. Such effects are transitory in the sense
that different flows can produce different effects a week, a month,
or a year later. This year's dam operating procedures do not affect
next year's recreational benefits through these transitory effects.
If flows cause permanent destruction of beaches, however, the results
are not transitory, but irreversible. Barring new technologies to
economically move large quantities of sand into the system, flow
release patterns that damage beaches today would affect the
recreational experience into the indefinite future.
Glen Canyon Fishing Conclusions
Conclusion 6: Glen Canyon Dam releases have a substantial impact on
the value of the fishing experience downstream. The attribute survey

of anglers, along with expressed opinions cf guides and resource
managers, indicated that two attributes of fishing trips are
important and are sensitive to flows. First, the overriding goal of
Glen Canyon anglers is to catch fish, and the possibility of catching
trophy-sized fish is particularly important. Glen Canyon anglers may
enjoy the fresh air and scenery as well, but they are primarily
interested in catching fish. Second, flows ir..f'luence how easy it is
to handle a boat and the risk of damaging boats and/or motors.
It should be remembered at this point that our study was conducted
without any information regarding the effects of alternative flow
regimes on the long-term productivity of the fishery. Another team
of researchers under the GCES is attempting to identify the long-term
relationships between biological productivity and dam operations.
However, our study was conducted without the benefit of their
results. This study focused on the short-term impacts of flows.

174
"

Given some number and size distribution of fish in the river, flows
may still affect the ability of anglers to catch the fish. Of
course, anglers' views on the nature and extent of these effects are
based on their rather subjective perceptions rather than scientific
evidence, but this did not preclude the economic analysis. Many
econanic values are based on subjective perceptions.
While anglers vary in their opinions about what flows make for good
and poor fishing, the attribute survey did identify a tendency to
favor moderate flows in the neighborhood of 10,000 cfs. To some
extent,. this is a compromise. At low flows, fish are concentrated
and may be easier to find. However, at very low flows of 3,000 cfs
or less, it is difficult or impossible to get boats upstream. Also,
as flows fall farther below 10,000 cfs, there is an increased risk of
damage to boats and motors due to shallow water over unseen rocks and
gravel bars. At higher flows, fish become more dispersed and may be
difficult to locate. The current becomes more swift and this makes
fishing more difficult. At relatively high flows, say 25,000 cfs,
boats with low horsepower have trouble going upstream against the
current. Also, bank anglers begin to encounter difficulties as the
water floods low lying areas along the river bank. While higher
water may reduce the chances of damaging boats and motors on
submerged rocks, the chances of swamping boats increases as anchors
are dragged in the swift current. In the past, when flows reached
40,000 cfs, the National Park Service has required boats to have at
least 25 horsepower motors.
This tendency to favor moderate flows was also supported in the CV
survey results. Surplus value per trip doubles (from $60 to $126)
between the 3,000 cfs and the 10,000 cfs constant flow scenarios.
The value then declines by 25 percent (to $94) for 25,000 cfs. At
40,000 cfs, the value drops by another 33 percent (to $52) below the
maximum value. A similar, though less dramatic tendency exists for
fluctuating flows.
Conclusion 7: Except at low average daily flows, most anglers prefer
constant nows to nuctuating nows. Fluctuations have several

disadvantages that many anglers are concerned about. Large
fluctuations mean that anglers may have to operate part of the day at
low or high flows with all the previously mentioned disadvantages of
both. Furthermore, changing water levels add additional
difficulties. Falling water may make it difficult to get downstream
over rocks and gravel bars that were more submerged on the trip
upriver. Rising water increases the likelihood of swamping a boat
while anchored or while the bow is pulled up on shore along the
river. A few anglers favor fluctuating flows because they believe
that rising water may stimulate feeding of the fish. Nevertheless,
preferences, as expressed in both the attribute and CV surveys,
clearly indicate that a majority of anglers feel the disadvantages of

175

fluctuations outweigh the advantages. In monetary terms,
fluctuations can reduce surplus values by as much as 30 percent
compared to constant flows with the same daily average.
As in the case of white-water boating, this general conclusion about
the detrimental effects of fluctuations does not hold at very low
flows. For fishing, the surplus value per trip for 3,000 cfs with
fluctuations was not statistically different from the 3,000 cfs
constant flow scenario value. Perhaps for the average angler, the
disadvantages of fluctuations at such low average daily flows are
balanced by the advantages of having higher flows for at least part
of the day.
Conclusion 8: Annual angling benef'its could be enhanced •ost by
maintaining moderate to low constant nows in the range arolllld 10,000
cf's in the nons1m11er months. This conclusion is based on the fact

that fishing is highest during the nonsummer months of September
through ~.ay. During 1985, for example, 86 percent of the angler
trips occurred during the nonsummer months. Some modest fluctuations
would be consistent with this conclusion, but flows less than 5,000
cfs or greater than perhaps 20,000 cfs would significantly reduce
surplus values. Interestingly, with the exception of ~.ay and
September, the months when fishing participation is highest are the
months when white-water boating is lowest.
Conclusion 9: In managing releases from Glen Canyon Dam, ef'fects on
the productivity of' the downstream f'ishery could have large
recreation effects. As noted previously, our analysis was done

without the conclusions from GCES research on the relationships
between flows and fishery productivity. We are nevertheless able to
conclude that productivity is a major issue from an economic
standpoint. If dam management could help restore the trophy fishery
in Glen Canyon, the annual benefits from the fishery could easily
double or more, based on increased surplus value and increased
participation. If dam management contributed to a loss in
productivity, equally dramatic effects in the opposite direction
could be sustained. Both the history of the fishery and our CV
results testify to the economic volatility of this fishery.
Glen Canyon Day-Use Rafting Conclusions
Conclusion 10: Glen Canyon Dam releases do not appear to have
significant impacts on day-use rafting over a very broad range of
nows. As explained in Chapter 7, neither the attribute nor the CV

surveys succeeded
on the quality of
of departing from
becomes necessary

in identifying significant effects of stream flows
this activity. In particular, the possible effects
Lee's Ferry rather than the Dam, as generally
at flows greater than 29,500 cfs, were examined,

176

but no statistically significant impacts were identified. Only if
flows were so low or so high that trips had to be curtailed
altogether would day-use rafting be adversely affected by flows.
Then, the loss in surplus value would be about $26 per trip lost.
Conclusions From Combining White-Water Boating and Fishing Values
Conclusion 11: now regilles C011bining high constant flows in the
11<>nths of May through September with moderate or low nows during the
remainder of the year would be likely to produce the largest
recreational benefits. The unconstrained optimal flow regime

presented in Chapter 8 suggests that flows in excess of 30,000 cfs
are ideal from May through September. Since there is little or no
white-water boating in December, January, or February, and fishing
participation is high, constant flows of 10,000 cfs are ideal in
these months. The remaining months involve substantial compromising
between the two groups. Because of its potentially high value per
trip and the sensitivity of the trip values to flows, white-water
boating would tend to predominate if such compromises were made on
the basis of surplus values. Thus, the unconstrained optimal flow
regime dictates flows in the 20,000 cfs to 30,000 cfs range in ~arch,
April, October, and November.
To fully meet the requirements of the unconstrained optimal regime
over a 12-month period would require better than 18.2 million
acre-feet of water. Obviously, such a large amount will not be
available in most years. Even if other objectives such as water
storage and electricity generation could be ignored, compromises
between white-water boaters and anglers would still be necessary in
most years. Analysis of the 8.25 million acre-foot regimes indicated
that if such compromises were based on recreational surplus value$
alone, considerable emphasis would be placed on providing high,
constant flows in the summer months to benefit white-water boating
even though this would entail losses in fishing benefits. While a
model to calculate constrained optimum flows for recreation does not
yet exist (see discussion in the next section), we can use what we
know to speculate on what such a flow regime would look like when
water is too scarce to fully reach an unconstrained optimum. Flows
during December, January, and February would be below 10,000 cfs in
order to provide more water for the summer months. During the summer
months, flows would be constant at 25,000 cfs or more. In the
remaining months, flows would lie in between.
Conclusion 12: Trade-offs between white-water boating and fishing,
and, in a broader sense, between recreation and other objectives are
likely to be less severe the aore water is available. More water, up

to 18.2 million acre-feet per year if managed properly, makes it more
feasible to improve total benefits for white-water boaters and
anglers combined. Losses to anglers from higher water can be more

•

177

.
than counter balanced by gains to white-water boaters. Second, as we
learned from examining the 1984 and 1985 regimes, high water (less
than 40,000 cfs) means long period of high, constant flows during May
and June and possibly beyond. Historical dam operations during
periods of high runoff are thus conducive to a high level of
white-water boating benefits, yet adverse impacts on fishing are not
extreme.
Conclusion 13: In general, extreme flow levels will adversely affect
recreation. In this regard, there is no disagreement between
white-water boaters and anglers. Very low flows in the 1,000 to
5,000 cfs range are disagreeable to both groups as are flows in
excess of 40,000 cfs. Also, extreme daily changes, such as daily
fluctuations between 1,000 cfs and 33,500 cfs would have substantial
adverse effects on both.
Conclusion 111: Recreation benefits could be COJDpared to power
benefits, properly calculated, but recreation benefits reported here
are not directly C011parable to power revenues. Comparison of
recreation and power benefits is beyond the scope of the current
study, but a few comments and words of caution are warranted. There
will be a natural tendency to ask whether the recreation benefits as
calculated in Chapter 8 are larger or smaller than the power benefits
from Glen Canyon Dam. This is a legitimate question from an economic
point of view. It is al so a complex question that would require a
major research effort to answer. Glen Canyon Dam and Power Plant are
part of a much larger system and it would be necessary to understand
their economic role in that system before power benefits could be
estimated. Also, it would be important to focus on what economists
would call "changes at the margin." The issue would not be the total
benefits of power from Glen Canyon Dam compar·ed to the total benefits
of recreation. Rather, the analysis would have to focus on hew power
and recreation benefits change with alternative dam management
strategies. This whole analysis would have to be conducted using
standard benefit-cost concepts.
-

However, one conclusion about recreation and power benefit
comparisons is abundantly clear without further analysis. The
recreation benefits calculated in this study are not comparable to
power revenues in any economically meaningful sense. The rates which
the government charges for power from facilities like Glen Canyon Dam
are, by law, based to a considerable degree on the costs of the
project, not its benefits. The goal, then, is to obtain enough
revenues to cover the so-called separable costs of the power plant
and associated facilities. Thus, the cost of the power would not be
a good indicator of the benefits that are attributable to it.
Comparing power revenues to recreation benefits is like comparing
apples and oranges. The results of such a comparison would have no
relevance whatsoever for economically sound management of the dam.

178

Some Final Thoughts
Stepping back from specific conclusions and viewing this study in a
broader context indicates several avenues for future research. As in
any research endeavor, much remains undone. Before looking to the
future, however, it is worthwhile to review some of the noteworthy
features of the current research.
To begin with, the link between attribute surveys and CV surveys
utilized here is, so far as we know, a new approach to valuation
research. We at HBRS cannot claim credit for this innovation. It
was built into the study design prior to our involvement. Attribute
survey results provided a strong basis for the CV questions,
particularly those involving scenarios when the CV results were
obtained, and the expressions of preferences from the attribute
surveys provided valuable cross-checks for the relationships between
flows and trip values implied in the CV responses. As such, the
attribute surveys provided substantial support for the validity of
the CV results. Serious discrepancies between the attribute results
and the CV results would have indicated that the CV results might
have serious problems that required further attention. We expect to
see more studies in the future that exploit the complementarity
between attribute surveys and CV surveys.
In evaluating any CV study it is important to ask whether
state-of-the-art survey methodology and CV techniques were applied.
The surveys conducted in this study were carefully designed and
pretested by experienced survey researchers. Sampling strategies had
to be adapted to the characteristics of the recreational activities
being studied and the needs of the supporting agencies for relatively
quick results. Other constraints such as the prohibition on
interviewing in Grand Canyon National Park had to be honored. Within
these constraints, we have no misgivings regarding the sampling
procedures used in this study. The final versions of the mail
surveys were administered with sufficient follow-up contacts to
assure high response rates. rhe CV questions themselves were
designed and implemented based on the latest research on the subject,
with particular emphasis on simulated market contingent-valuation
comparisons. Data were carefully computerized and analyzed. In sum,
we believe that this study represents the current "state-of-the-art"
in valuation research.
Also noteworthy is the use of both actual trip values and scenario
values in this study. Most researchers in the field would probably
agree that values based on actual experiences are superior. We would
concur, and we used actual trip values for constant flow white-water
boating and day-use rafting trips. However, obtaining valuation data
for a sufficient range of conditions actually experienced is not
always feasible. Words in a scenario are probably not a perfect

179

•
substitute for an actual experience. However, based on comparison of
actual experience values and scenario values for both commercial and
private white-water boaters, they seemed to be an acceptable
substitute in this case.
Finally, the present study is noteworthy in the extent to which it
carries the research forward into the analysis of actual policy
options. Unfortunately, a great many CV studies stop with the value
per trip or per person, leaving the task of interpreting the policy
implications and recommendations to others. Thanks to the support
provided by the Bureau of R~clamation, we were able to value a number
of annual flow regimes in Chapter 8 and include relevant conclusions
from that analysis in this chapter. This is a significant step
toward a full and correct assessment of the implications of our
results for use in policy discussions.
One area in which future research is needed is the relationship
between participation, flows, and fishing success. Based on our·
scenarios, we were able to estimate how surplus value per trip
changes with flows, but were unable to deal satisfactorily with how
the number of fishing trips would change. The volatility of
participation in this fishery has been stressed repeatedly in this
report, as has the potential value of a rehabilitated trophy
fishery. However, full evaluation of the potential implications of a
changed fishery would require a good participation model. Some
investigation of how the values reported here have been affected by
the lures-only regulation, introduced in 1986, would also be
desirable. Participation was not an issue for white-water boating
because of the National Park Service's present restrictions on launch
schedules.
Returning for a moment to the apparent superiority of CV values based
on actual experience compared to those based on scenarios, it is
worth remembering that we were generally precluded from obtaining
experience-based values by dam release schedules during this study.
Only in the case of white-water boating constant flow values was
there sufficient variation in actual flows experienced to allow us to
use actual trip values. Future research should re-evaluate the
fishing and white-water boating experiences under actual conditions
of constant and fluctuating flows. Particular attention should be
given to the effects of fluctuations. The definition of fluctuations
used in this study, changes in flow of 10,000 cfs or more, is
obviously very crude. A 10,000 cfs fluctuation around an average
daily flow of 7,000 cfs is obviously very different from a 10,000 cfs
fluctuation around a daily average of 20,000 cfs, whether one is
fishing or rafting. Furthermore, a 10,000 cfs fluctuation at the dam
will not have the same impact on a white-water boater a hundred miles
down the river as it will just below Lee's Ferry. A useful first
attempt at analyzing fluctuating flow values has been made in this
study, but much more could be done.

180

..
More could be done on the modelling side as well. The unconstrained
optimal recreation regime discussed in Chapter 8 was determined
heuristically, the problem is clearly amenable to more formal
procedures. Constraints could easily be added to form a nonlinear
programming problem. One simple constraint would be the total
releases over a 12 month period. In this way an optimal release
scenario for any given quantity of water could be defined. Once
again the objective would not be to directly prescribe dam
operations. Instead, such a model would define a constrained optimal
recreation scenario for comparison with alternative plans designed to
meet a broader range .of objectives. If the optimum flow regime for a
given annual number of acre-feet would produce say, $11 million in
recreation benefits, while an alternative proposed for actual
implementation would have $10.5 million, then the adverse
recreational impact could be considered slight. If the alternative
has benefits of $6 million, then perhaps another alternative, less
harmful to recreation, could be sought.
Additional constraints could be added to such a model. Constraints
on dam storage capacity and the seasonality of runoff would allow the
model to more closely simulate dam operations.
The valuation model itself could be refined in many ways. Instead of
monthly data, a more sophisticated model might be developed to deal
with weekly, daily, or even hourly data. Hydrological models under
development elsewhere in the GCES might be adapted to trace
downstream river conditions resulting from dam operations. Another
part of the GCES experimented in a preliminary way with a simulation
model of white-water boating but more research would be needed before
a recreation simulation model and an economic evaluation model could
be linked.
Finally, looking beyond research on recreation alone, future research
could evaluate the trade-offs between electric power production at
Glen Canyon Dam and downstream recreation. As noted under Conclusion
14, such a comparison would require an adequate characterization of
the power generation and transmission network that absorbs Glen
Canyon power. The emphasis would be on power benefits, not power
revenues. The power valuation model and the recreation valuation
model would have to match up in terms of how water releases from Glen
Canyon Dam were characterized.
Enough has now been said to verify that a great deal remains undone.
Still, the research reported here does represent a solid step forward
in understanding how recreation between Glen Canyon Dam and Lake ~~ad
is affected by dam operations. To some extent, this research was
needed to establish the value of recreation compared to the more
institutionalized goals of water supply and power generation. Glen
Canyon Dam operations can have positive or negative effects on

181

recreation that are worth millions of dollars. Furthermore, our
research constitutes a new and powerful basis for integrating
recreation into the decision process. Few would advocate that
national environmental assets like the Grand Canyon be managed for
dollars alone, but dollars can be used to quantify some of the
effects of flows on the recreational environment. Such quantified
effects are rich in insights that could prove useful in future
decisions regarding Glen Canyon Dam operations •

•
•

182

LITERATURE CITED
Alfano, Geraldine, and Gerald Marwell. 1981. Experiments on the
Provision of Public Goods III: Non-divisibility and Free Riding in
'Real' Groups. Social Psychology Quarterly. 43: 300-309.
Anderson, Glen D., and Richard C. Bishop.
Problem. In: Natural Resource Economics.
Wingham, Massachusetts: Kluwer-Nijhoff.

.

1986. The Valuation
Daniel W. Bromley (ed.).

Bishop, Richard C., Thomas A. Heberlein, Michael P. Welsh, and Robert
M. Baumgartner. 1984. Does Contingent Valuation Work? Results of
the Sandhill Experiment. Invited paper, joint meetings of the
American Agricultural Economics Association and the Association of
Environmental and Resource Economists. Cornell University, August
5-8.
Bishop, Richard C., Thomas A. Heberlein, and ~.ary Jo Kealy. 1983.
Contingent Valuation of Environmental Assets: Comparisons With A
Simulated Market. Natural Resources Journal. 23(July): 619-633.
Bohm, Peter. 1972. Estimating Demand for Public Goods:
Experiment. European Economic Review. 3: 111-130.

An

Boyle, Kevin J., and Richard C. Bishop. 1984. A Comparison of
Contingent Valuation Techniques. Staff Paper No. 222. University of
Wisconsin-Madison, Department of Agricultural Economics.
Boyle, Kevin J., and Richard C. Bishop. Forthcoming. Welfare
Measurements Using Contingent Valuation: A Comparison of Techniques.
Subinitted to the American Journal of Agricultural Economics.
Brookshire, D.S., W.D. Schulze, M.A. Thayer and R.C. d'Arge. 1982.
Valuing Public Goods: A Comparison of Survey and Bedonie Approaches.
American Economic Review. 72(1) March: 165-177.
Brubaker, Earl. 1982. Sixty-Eight Percent Free Revelation and
Thirty-two Percent Free Ride? Demand Disclosures Under Varying
Conditions on Exclusion. In: Research in Experimental Economics.
V.L. Smith (ed.). 151-166.
Coursey, Don L., J.J. Hovis, and William D. Schulze. Forthcoming.
The Supposed Disparity Between Willingness to Accept and Willingness
to Pay Measures of Value. Quarterly Journal of Economics.
Crumbo, Kim. 1981. A River Runner's Guide to the History of the
Grand Canyon. Johnson Books, Boulder, Colorado.

•

183

..

Davis, R.K. 1964. The Value of Big Game Hunting in a Private
Forest. Transactions of the North American Wildlife and Natural
Resources Conference. 29: 393-403 •
Davis, R.K. 1963. Recreational Planning As An Economics Problem.
Natural Resources Journal. 3: 238-249.
Desvouges, William H., V.K. Smith, and Matthew P. McGivney. 1983. A
Comparison of Alternative Approaches for Estimating Recreation and
Related Benefits of Water Quality Improvements. Research Triangle
Institute. Report to U.S. Environmental Agency, EPA-230-05-83-001.
Washington, D.C.
Dickie, Mark, Dan Fisher, and Shelby Gerking. 1985. Market
Transaction and Hypothetical Demand Data: A Comparative Study.
Unpublished Manuscript, Department of Economics, ·university of
Wyoming.
Fisher, Ann, and Robert Raucher. 1984. Intrinsic Benefits of
Improved Water Quality: Conceptual and Empirical Perspectives. In:
Applied Microeconomics. Vol. 3. V. Kerry Smith and Ann P. Witte
(eds.). Greenwich, CT: JAI Press.
Freeman, A. Myrick. 1979. Hedonic Prices, Property Values and
Measuring Environmental Benefits: A Survey of the Issues.
Scandinavian Journal of Economics. 81(2): 154-173.
Hanneman, Michael W. 1984. Welfare Evaluations in Contingent
Valuation Experiments With Discrete Responses. American Journal of
Agricultural Economics. 66: 332-341.
Isaac, R., R. Walker, and S. Thomas. 1984. Divergent Evidence on
Free-Riding: An Experimental Examination of Possible Explanations.
Public Choice. 43: 113-149.
Janisch, Ed. 1985. Evaluation of Lee's Ferry Fishery and Future
Management. Report to Fisheries Branch, Arizona Game and Fish
Department. January.
Kim, Oliver and Mark Walker. 1984. The Free Rider Problem:
Experimental Evidence. Public Choice. 43: 3-24.
Knetsch, Jack L. and Robert K. Davis. 1966. Comparisons of Methods
for Recreation Evaluation. In: Water Research. Allen V. Kneese and
Stephen C. Smith (eds.). Baltimore, John Hopkins Press for Resources
for the Future.

184

Krutilla, John B., and Anthony C. Fisher. 1975. The Economics of
Natural Environments: Studies in the Valuation of Commodity and
Amenity Resources. Baltimore, Johns Hopkins University Press for
Resources for the Future.

..

Lavender, David. 1985. River Runners of the Grand Canyon. Grand
Canyon National History Association. Grand Canyon, Arizona.
Harwell, Gerald, and Ruth E. Ames. 1979. Experiments on the
Provision of Public Goods I Resources, Interest, Group Size and the
Free-Rider Problem. American Journal of Sociology. 85(6): 1335-1360.
Marwell, Gerald, and Ruth E. Ames. 1980. Experiments on the
Provision of Public Goods II Provision Points, Stakes, Experience and
the Free-Rider Problem. American Journal of Sociology. 85(4):
926-938.
McCollum, Daniel W. 1986. The TC Method:
Validity. Unpublished Ph.D. Dissertation.
University of Wisconsin-Madison.

Time, Specification, and
Department of Economics,

Mitchell, Robert C. and Richard T. Carson. 1981. An Experiment in
Determining Willingness to Pay for National Water Quality
Improvements. Resources for the Future, Draft Report to U.S.
Environmental Protection Agency. Washington, D.C.
Randall, Alan, and John R. Stoll. 1983. Existence Value in a Total
Valuation Framework. In: Managing Air Quality and Scenic Resources at
National Parks and Wilderness Areas. Robert D. Rowe and Lauraine G.
Chestnut (eds.). Boulder, Colorado: Westview Press.
Richards, ~erton T., D. Brent Wood, and David A. Caylor. 1985.
Sportfishing at Lees Ferry, Arizona: User Differences and Economic
Values. Final Report to Northern Arizona University Organized
Research Committee.
Ridgeway, James. 1984. River Rafting Companies Grow With the Flow.
In Business. May-June: 39-43.
Samuelson, P. 1954. The Pure Theory of Public Expenditure.
of Economics and Statistics. 36: 387-389.

Review

Scherr, B.A., and E.M. Babb. 1975. Pricing Public Goods: An
Experiment with Two Proposed Pricing Systems. Public Choice. 23:
35-48.

•

185

Schneider, Friedrich, and Werner W. Pommerehne. 1981. Free Riding
and Collective Action: An Experiment in Public Microeconomics.
Quarterly Journal of Economics. 97: 689-702 •
•

Schulze, William, Ralph d'Arge, and David Brookshire. 1981. Valuing
Environmental Commodities: Some Recent Experiments. Land Economics.
57(2): 151-172.
Sellar, Christine, John R. Stoll, and Jean-Paul Chavas. 1985.
Validation of Empirical ~~asures of Welfare Change: A Comparison of
Nonmarket Techniques. Land Economics. 61(May): 156-175.
Shelby, Bo. 1981. Research, Politics, and Resource Yanagement
Decisions: A Case Study of River Research in Grand Canyon. Leisure
Sciences. 4(3): 281-296.
Smith, V.L. 1977. The Principle of Unanimity and Voluntary Consent
in Social Choice. Journal of Political Economics. 85: 1125.
Smith, V. Kerry. 1983. Option Value: A Conceptual Overview.
Southern Economic Journal. 49(Jan): 654-668.
Smith, V. Kerry.
60: 292-296.
Stevens, Larry.
Red Lake Books:

1984.

A Bound for Option Value.

Land Economics.

1983. The Colorado River in Grand Canyon:
Flagstaff, Arizona.

A Guide.

Thayer, M.A. 1981. Contingent Valuation Techniques for Assessing
Environmental Impacts: Further Evidence. Journal of Environmental
Economics and Management. 81: 27-44.
Tideman, Nicholas T. 1983. An Experiment in the Demand Revealing
Process. Public Choice. 41: 387-401.
U.S. Water Resources Council. 1983. Economic and EnviroP.mental
Principles and Guidelines for Water and Related Land Resources
Implementation Studies. U.S. Government Printing Office.
Welsh, Michael P. 1986. Exploring the Accuracy of the Contingent
Valuation Method: Comparisons with Simulated ~~rkets. Unpublished
Ph.D. Dissertation. Department of Agricultural Economics, University
of Madison-Wisconsin.

'

186

GLOSSARY

Angler Day:

A visit to engage in fishing for one calendar day.

•

Attribute Survey: A survey used to identify the important
flow-sensitive characteristics (or attributes) of each recreational
experience.
Average Daily Flow: The average amount of water, measured in cubic
feet per second (cfs), released from Glen Canyon Dam over the course
of a 24-hour period.
Base-Load Generation: Operation of a power plant at constant levels
of output over periods of more than one day, in order to contribute to
the minimum constant amount of load of the power system.
Constant Flow: In this study, a constant flow occurs when dam
releases do not vary more than 10,000 cfs over a 24-hour period.
term is used in contrast to fluctuating flow, as defined below.

This

Contingent Valuation: This is a technique for valuation of nonmarket
goods, such as outdoor recreation. People are asked in surveys about
the value they would place on the good if a market or other means of
payment existed. All transactions are hypothetical.
Contingent-Valuation Survey: A survey used to quantify the effects of
different flow release patterns, as measured in dollar terms, on each
recreational experience.
CV:

Contingent valuation as defined above.

Dichotomous-Choice Questions: A technique of asking contingentvaluation questions where each respondent is asked whether she or he
would be willing to pay some specified offer amount (defined below),
and can answer either yes or no. Offer amounts are randomly assigned
and varied from respondent to respondent, so that data for a full
range of values can be obtained.
Expenditures: In recreation economics, expenditures are the amounts
of money people actually spent to take a recreational trip. In
recreation valuation, expenditures often take the place of market
value (defined below).
Flow Regime: A description by month of releases from Glen Canyon Dam
over an entire year, including average daily flows and whether such
flows are constant or fluctuating flows as defined elsewhere in this
glossary.

f

187

..
Flow Scenario: A description of the implications of a specified flow
for fishing or white-water boating used as part of some contingentval ua tion questions in this study.
Flow Value Function or Curve: A function stated mathematically or
plotted in a graph that expresses the relationship between surplus
value per trip (defined below) and average daily flow (defined above).
Fluctuating Flow: In this study, a fluctuating flow occurs when dam
releases vary more than 10,000 cfs over a 24-hour period. This term
is used in contrast to constant flows as defined above.
Hedonic Price Method: A technique for inferring values of nonmarket
goods from market values of related goods. For example, property
values in several areas with varying air quality might be analyzed to
estimate how much people would be willing to pay for cleaner air.
Hypothetical Bias: This is a catch-all category of potential
distortions in contingent-valuation results. The common thread
running through this category of bias is that distortions are
unintentional. That is, because people have never before been asked
to express their preferences for nonmarket goods in monetary terms and
because CV involves only hypothetical transactions, study participants
may not know their surplus values and may not give accurate estimates
of their values. This is in contrast to strategic bias, defined
below, where distortions are intentional.
Logit Equation: In contingent valuation, logit equations are used to
analyze dichotomous-choice (defined above) data. Logit equations
characterize the relationship between the offer amount (defined below)
and the probability that offer amount will be accepted. See Appendix
L, M, N, and 0 for more complete details about logit functions, how
they are estimated, and how values are derived from estimated logit
equations.
Market Value: In benefit-cost analysis of projects or policies
affecting market goods, market value is the quantity of the good
produced multiplied by market price per unit.
National Economic DeveloJlllent (NED) Benefits: Benefits evaluated from
the point of view of the nation as a whole (as opposed to regional
benefit, defined below). Benefits in this study are evaluated from
the NED perspective.

'

Offer Amount: In this study's dichotomous-choice questions (defined
above), respondents (i.e., white-water boaters, anglers, and day-use
rafters) were asked whether they would take a specified trip (an
actual trip or a trip described in a scenario) had their expenses been
larger by some specified amount. This specified amount is sometimes
referred to as the offer amount.

188

Peak-Load Generation: Operation of a power plant at levels which vary
greatly within a 24-hour period in order to help satisfy relatively
high power demands that typically occur for short periods.
Recreation Day:

A visit during one calendar day.

Regional Economic DeveloJlllent (RED) Benefits: Benefits evaluated from
the perspective of one region within the nation. These benefits were
also referred to as "local economic impacts" in this report. This
study adopted a National Economic DeveloJlllent (NED) perspective
(defined above) rather than RED perspective.
Strategic Bias: A term used to describe potential biases in
contingent-valuation results that would occur if study participants
intentionally state untrue responses in order to influence study
results in ways that they perceive to be advantageous to themselves.
For example, suppose that white-water boaters believed that our study
results would influence fees charged by the National Park Service.
Suppose, further, that they answered our CV questions in ways that
would lower our final results in order to minimize -this effect. This
would be a strategic bias in the value estimates. This term is used
in contrast to hypothetical bias (defined above) where distortions are
unintentional.
Surplus Value: In recreation economics, surplus value (also called
consumer's surplus) is the maximum amount a person is willing to pay
for access to the recreational resource over and above actual
expenditures (defined above).
Total Value: As used in recreation economics, total value equals
expenditures (as defined above) plus surplus value (as defined above).
Travel-Cost Method: A method of estimating the surplus value (defined
above) of a recreational activity using data on trip-related
expenditures (defined above) and number of trips taken.
Trip: In this study, one complete visit to Glen Canyon National
Recreation Area and/or Grand Canyon National Park to engage in
white-water boating, fishing, or day-use rafting. Except for day-use
rafting, trips typically involve more than one recreation day, as
defined above.
User Dav (also Visitor Day of Recreation): Twelve hours of recreation
of a given type. For example, two people fishing for six hours each
would constitute one user day.
f

Willingness to Pay: An approach for estimating values, where value is
defined as the maximum amount a consumer would be willing to pay for
the outputs rather than do without them.

JJ-1'

•

APPENDICES
GLEN CANYON DAM RELEASES AND
DOWNSTREAM RECREATION: AN
ANALYSIS OF USER PREFERENCES
AND ECONOMIC VALUES

Recreation
of the
Glen Canyon Environmental Studies

By
Richard C. Bishop
Kevin J. Boyle
Michael P. Welsh
Robert M. Baumgartner
Pamela R. Rathbun
HBRS

GCES Report No.
January, 1987

A- fl
LIST OF APPEHDICBS

APPENDIX A.

WHITE-WATER COMMERCIAL GUIDE SURVEY

APPENDIX B.

WHITE-WATER PRIVATE TRIP LEADER SURVEY

APPENDIX C.

WHITE-WATER BOATERS ATTRIBUTE SURVEY .

. c. 1

APPENDIX D.

WHITE-WATER COMMERCIAL BOATERS
CONTINGENT-VALUATION SURVEY .

• D. 1

WHITE-WATER PRIVATE BOATERS CONTINGENTVALUATION SURVEY

• E. 1

APPENDIX E.

.

• A. 1

•

• B.1

APPENDIX F.

GLEN CANYON ANGLERS ON-SITE ATTRIBUTE SURVEY

APPENDIX G.

GLEN CANYON ANGLERS ON-SITE PRE-CONTINGENTVALUATION SURVEY

• G. 1

APPENDIX H.

GLEN CANYON ANGLERS CONTINGENT-VALUATION SURVEY

• H.1

APPENDIX I.

DAY-USE RAFTERS ATTRIBUTE SURVEY

• I.1

APPENDIX J.

DAY-USE RAFTERS CONTINGENT-VALUATION SURVEY

• J. 1

APPENDIX K.

GLEN CANYON ANGLER CONTINGENT-VALUATION PRETEST
SURVEY AND RESULTS

. K. 1

TECHNICAL CONSIDERATIONS FOR DERIVING SURPLUS
VALUES FROM RESPONSES TO DICHOTOMOUS-CHOICE
VALUATION QUESTIONS

• L.1

ANALYSES OF RESPONDENTS' ANSWERS TO THE
WHITE-WATER BOATER CONTINGENT-VALUATION
QUESTIONS

• M. 1

ANALYSES OF RESPONDENTS' ANSWERS TO THE ANGLER
CONTINGENT-VALUATION QUESTIONS.

• N.1

ANALYSES OF RESPONDENTS' ANSWERS TO THE GLEN
CANYON DAY-USE RAFTER CONTINGENT-VALUATION
QUESTION

. 0.1

APPENDIX L.

APPENDIX M.

APPENDIX N.
APPENDIX O.

LITERATURE CITED
i

. • F.1

•

A. 1

•

APPENDIX A
VBITE-VATER <X>IH!RCIAL GUIDE SURVEY

l

A.2

•
There are many different alternatives for flow regimes from the
Glen Canyon Dam. The purpose of this study is to find out how
different flows affect river running in the Grand Canyon. As an
experienced boatman, you are the expert and have detailed knowledge of
the effects different flows have on river trips. Regardless of
whether you have taken 1 trip or 100 trips in the Grand Canyon, your
opinions are important. You can contribute your expertise to the
decision-making process by filling out this questionnaire.
We're going to ask you about the effect of different water levels
on the rapids, campsites, and attraction sites in the Grand Canyon, as
well as your preferences for alternative flow regimes. Please answer
the questions from your perspective as a commercial guide.

Please check below the kind of boat you most
trips: (PLEASE CHOOSE ORE TYPE OF BOAT)

o~ten

use for your

___ MOTOR RAFT----------> ___ 33 feet or larger
smaller than 33 feet
___ ROWING RAFT---------> ___ 20 feet or larger
___ less than 20 feet
___ DORY

Please answer all the remaining questions BASED OH THE ll.RD OF BOAT
YOU IHDICATED ABOVE.

A.3

..

RAPIDS

In this first section, we are interested in learning how different water
levels affect rapids.
1. Assuning you were to run an ENTIRE GR.AHD C.AHYON TRIP .AT .A CONST.ANT
FL<Jl, please specify the minimum and maximum water levels for
running rapids safely with passengers, the level that provides the
best ride, and the level you prefer, as a guide, for running
rapids.
(AllSVER IN CFS)

Minimum level for running
safely with i:assengers

cfs

Maximum level for running
safely with passengers

cfs

. Best ride for i:assengers

cfs

Water level you prefer
as a guide

cfs

2. Which rapids would be most problematic BEL<Jl your minimum safety
level'? (IF NONE, WRITE NONE)
1
2

3

3. Which rapids would

be

(IF NONE, WRITE NONE)
2

3

problematic ABOVE your maximum safety level'?

A.4
CAMPSITES
There may be several areas in the Canyon where camps become a problem
at certain water levels. For the items below, please identify the
water levels where camps become a problem.
1. Would certain flow levels cause problems with access to or use of
campsites for you?
1

NO---->Skip to next page, question 1

2

YES

PLEASE FILL IN EACH OF THE BLANKS BELOW IF APPLICABLE
2. Below

cfs, we would have problems getting to camp on time
because we would have to spend too much time traveling on the
river.

3. Above
cfs, important camps might be unavailable for use
because they are under water.

4. With daily fluctuations of more than

cfs, we would have
problems with hanging up boats, loading boats, or having to move
camp.

5. Which campsites would be most problematic at low water (9,000 cfs
or less) with the type of boat you most often use? (IF NONE, WRITE
NONE)
1 ~~~~~~~~~~~~~-

2 ~~~~~~~~~~~~~-

3 ~~~~~~~~~~~~~-

6. Which campsites would be most problematic at high water (32,000 cfs
or more) with the type of boat you most often use?
NONE)
1 ~~~~~~~~~~~~~-

2 ~~~~~~~~~~~~~-

3

~~~~~~~~~~~~~-

(IF NONE, WRITE

A.5
TIME FOR ATTRAC'l'IOHS

It may be that certain water levels allow more or less time for stops
at attraction sites. For the following items, please indicate which
flow levels, if any, would affect your stops at attraction sites.
1. Would certain flow levels cause problems with access to or use of
attractions for your trip?
1

NO---->Skip to next page, question 1

2

YES

PLEASE FILL IN EACH OF THE BLANKS BELOW IF APPLICABLE
2. Below
cfs there would not be enough time for stops at
attractions because of the need to "make up time"

3. From

to
cfs the amount of time for stops at
attractions would be "about right"

4. Above ----- cfs there would be extra time for stops at
attractions
5. Which attraction sites would be most problematic (difficult to pull
in, tie up, or load and unload passengers) at low water (9,000 cfs
or less)? (IF NONE, WRITE NONE)
1

2 _ _ _ _ _ _ _ _ _ _ __

3 -------------6. Which attraction sites would be most problematic (difficult to pull
in, tie up, or load and unload passengers) at high water (32,000 cfs
or above)? (IF NONE, WRITE NONE)
1 --------------

2 --------------

3 --------------

A.6

•

COllST.ART Fl.Oil LEVELS

1. How would you, as a commercial river guide using the boat you

usually pilot, evaluate each of the following water levels for a
commercial Grand Canyon river trip? Assume the water level would
be constant for the entire trip. (CIRCLE ONE NUMBER FOR EACH WATER
LEVEL)

iT

Very
Somewhat
Very
Somewhat
Flow Level Satisfactory Satisfactory Neutral Unsatisfactory Unsatisfactory
2,000 cfs

2

3

4

5

2

3

4

5

2

3

4

5

2

3

4

5

7,500 cfs

2

3

4

5

10,000 cfs

2

3

4

5

3,000 cfs

1

4,000 cfs
5,000 cfs

1

15,000 cfs

1

2

3

4

5

20,000 cfs

1

2

3

4

5

2

3

4

5

2

3

4

5

40,000 cfs

2

3

4

5

50,000 cfs

2

3

4

5

2

3

4

5

2

3

4

5

25' 000 cfs
30,000 cfs

60,000 cfs
80,000 or more

1

1

~

r

A.7

In the following questions, we would like to lmow about tolerable
daily fluctuations in flow.

1.

Do you feel that you can accurately describe the daily water level
fluctuations you have experienced in terms of cfs daily change?

NO----->Please skip to next page
2 YES

2.

If the flow is somewhere in the range from 5,000 to 9,000 cfs,
what is the largest daily change in flow that is tolerable?
_ _ _ _ _ cf s daily change

3.

If the flow is somewhere in the range from 9,000 to 16,000 cfs,
what is the largest daily change in flow that is tolerable?
_____ cfs daily change

4.

If the flow is somewhere in the range from 16,000 to 32,000 cfs,
what is the largest daily change in flow that is tolerable?
cf s daily change

5.

If the flow is somewhere above 32,000 cfs, what is the largest
daily change in flow that is tolerable?
- - - - - cfs daily change

l

A.8
ALTERHATIVE FLOW REGIMES
In this next section, we have listed 4 different scenarios for RELEASE
PATTERNS FROM THE GLEN CANYON DAM. For each one, please check the
disadvantages you perceive as a commercial river guide. After considering all four scenarios, please indicate which one you prefer most.
In these questions, we are interested only in factors related to your
ability to conduct trips, not impacts on the geology or biology of the
Canyon. These scenarios are for discussion only; they are not
currently being propcsed as actual operating procedures for the dam.
There would be no daily fluctuations, but flows would
change from one month to the next. Flows during May, June, July, and
August would be 10,000, 10,400, 12,750, and 14,400 cfs respectively,
with no daily fluctuations. The rest of the year, flows would range
from 8,300 cfs - 14,600 cfs, again with no daily fluctuations.
SCENARIO A:

Possible problems (check those which apply to this scenario during May
through August and from September through April)

May-Aug.

Sept.-Apr.
too much time on river
not enough time for attraction sites
no flexibility in running trip
unable to avoid other people at attractions or camps
difficulty with mooring boats
problems running rapids
difficult to plan timing for attraction sites or
camps
inadequate flow
not enough challenge/fun in rapids
lowest flows are too low
other (please

specify)~~~~~~~~~~~~~~-

Your OVERALL evaluation of this flow scenario (CIRCLE ONE):
completely
acceptable

somewhat
acceptable

neutral

somewhat
unacceptable

completely
unacceptable

r

A.9

•
SCENARIO B: Flows would be constant at 25,000 cfs during June, July,
and August. During the rest of the year, daily flows could range from
1,000 to 31,500 cfs.

Possible problems (check those which apply to this scenario during
June through August and from September through May)

June-Aug. Sept.-May
flow is too unpredictable
too much time on river
not enough time for attraction sites
no flexibility in running trip
unable to avoid other people at attractions or camps
difficulty with mooring boats
problems running rapids
difficult to plan timing for attraction sites or
camps
inadequate flow
problems with fluctuations at camps (hanging-up
boats, etc.)
not enough challenge/fun in rapids
lowest flows are too low
other (please

specify).~~~~~~~~~~~~~~-

Your OVERALL evaluation of this flow scenario (CIRCLE ONE):
completely
acceptable

somewhat
acceptable

neutral

somewhat
unacceptable

completely
unacceptable

A.10

•
Flows would vary by day, season, and month. During June,
July, and August, daily flows could range from 3,000 to 31,500 cfs
with a major peak at 3:00 p.m. During the rest of the year, daily
flows could range from 1,000 to 31,500 cfs.
SCEllARIO C:

Possible problems (check those which apply to this scenario during
June through August and from September through May)

June-Aug. Sept.-May
flow is too unpredictable
too much time on river
nGt enough time for attraction sites
no flexibility in running trip
unable to avoid other people at attractions or camps
difficulty with mooring boats
problems running rapids
difficult to plan timing for attraction sites or
camps
inadequate flow
problems with fluctuations at camps (hanging-up
boats, etc.)
not enough challenge/fun in rapids
lowest flows are too low
other (please specify)-"'---------------

Your OVERALL evaluation of this flow scenario (CIRCLE ONE):
completely
acceptable

somewhat
acceptable

neutral

somewhat
unacceptable

completely
unacceptable

A. 11

Flows would vary by day, season, and month. Throughout
the year, daily flows could range from 8,000 to 25,000 cfs. During
the summer there would be a major peak around 3:00 p.m.

SCERABIO D:

Possible problems {check those which apply to this scenario)

___ flow is too unpredictable
too much time on river
___ not enough time for attraction sites
___ no flexibility in running trip
unable to avoid other people at attractions or camps
___ difficulty with mooring boats
problems running rapids
___ difficult to plan timing for attraction sites or camps
___ inadequate flow
___ problems with fluctuations at camps {hanging-up boats, etc.)
___ not enough challenge/fun in rapids
___ lowest flows are too low
___ other {please specify)

Your OVER.ALL evaluation of this flow scenario {CIRCLE ONE):
completely
acceptable

somewhat
acceptable

neutral

somewhat
unacceptable

completely
unacceptable

OVER.ALL, how would you rank the four flow scenarios {A-D) presented

above from your perspective as a commercial river guide?
EACH SCENARIO)
Scenario
Letter
Most preferred scenario
Next most preferred scenario
Third most preferred scenario
Least preferred scenario

{PLEASE RANK

A. 12
GENERAL IMPACfS OB TRIP
1.

At what CONST.ART FLOW LEVELS in the Grand Canyon have you had to
do any of the following? Assume each water level is a constant
flow for the entire trip. FOR EACH ITEM, PLEASE CHECK ALL WATER
LEVELS THAT APPLY.

!

•

--------Constant Flow Level (cfs x 1000)-----0ver
1-5 5-10 10-15 15-20 20-25 25-35 35-50 ~
Stop at major rapids
such as Houserock, Hance,
Crystal or Lava to scout
Stop at any other rapids
to scout
Row or motor more than
usual to make up some
time
Row less than usual or
turn off motor because
you are ahead of schedule
Hurry to get out of
camp in the morning
Have passengers walk around
a rapid because the water
is too low
Have passengers walk around
a rapid because the water
is too high

I'

A.13

.
2. Assume that flow levels were varying from a low of 3,000 cfs to a high
of 25,000 cfs each day (or within a 24 hour period). UNDER THESE FLOW
CONDITIONS, which of the following would you be likely to do? (PLEASE
CIRCLE ONE NUMBER FOR EACH ITEM)
Very
Likely

Somewhat
Likely

Not at all
Likely

Stop to scout rapids more
often than usual

2

3

Select specific campsites because
they offer better protection
against water level changes

2

3

Camp above a major rapid to wait
for the water level to change
1

2

3

Change your itinerary to try to
reach certain points at a
good time

2

3

2

3

Spend more time at scheduled
attraction stops

2

3

Spend less time at scheduled
attraction stops

2

3

Spend less time in camp

2

3

Make camp earlier

2

3

Check on the boat mooring
during the evening to see if it
should be moved

1

A.14

EXTRA TIME Olf YOUR TRIP
1.

If your commercial trip was making good time and 6-8 fewer hours of
actual river time was needed to complete the trip in the scheduled
time, which of the following would you be most likely to do?
(CIRCLE ONE NUMBER FOR EACH ITEM)

Not at all
Likely

Somewhat
Likely

Very
Likely

Turn motor off and fl cat

1

2

3

Stay in camp later in the
morning

1

2

3

Make camp earlier in the
afternoon

1

2

3

2

3

Take longer lunches
Spend more time scouting
rapids

1

2

3

Spend more time at scheduled
attraction stops

1

2

3

Stop at additional attraction
sites

1

2

3

Spend a layover day at a
campsite

1

2

3

A.15
2. If your commercial trip was making good time and you decided to stop
at ADDITIONAL attraction sites (sites you don't usually visit), what
are the three additional attraction sites you would most likely
stop?
(FOR EACH ATTRACTION SITE, ALSO INDICATE WHAT YOU WOULD SUGGEST THAT
PASSENGERS DO. PLEASE TRY TO BE SPECIFIC--IF YOU WOULD SUGGEST A
HIKE, INDICATE WHERE (PLACE NAME) YOU WOULD RECOMMEND THAT THEY
HIKE).

A'l'TRAC?IOR SITE - - - - - - - - - - -

River Mile

What would you suggest that passengers do here?

ATTRACTION SITE - - - - - - - - - - -

River Mile

What would you suggest that passengers do here?

ATTRACTION SITE - - - - - - - - - -

River Mile

What would you suggest that passengers do here?

A.16

..
3. If your commercial trip was making good time and you decided to
spend an EXTRA layover day somewhere, what are the three campsites
at which you would be most likely to layover for a day.
(FOR EACH CAMPSITE, ALSO INDICATE WHAT YOU WOULD SUGGEST THAT
PASSENGERS DO. PLEASE TRY TO BE SPECIFIC--IF YOU WOULD SUGGEST A
DAY HIKE, INDICATE WHERE (PLACE NAME) YOU WOULD RECOMMEND THAT THEY
HIKE).
River Mile

What would you suggest that passengers do here on a layover day?

CAMPSITE - - - - - - - - - - -

River Mile

What would you suggest that passengers do here on a layover day?

CAMPSITE - - - - - - - - - - -

River Mile

What would you suggest that passengers do here on a layover day?

A.17
INFORHATIOB OH Fl.Oil RELEASES

1. In the pa.st, have you used information concerning scheduled Glen
Canyon Dam releases as a basis for making decisions about your raft
trips in the Grand Canyon (e.g. when to run particular rapids, when
to camp, where to moor boats, etc.)? (CIRCLE ONE NUMBER)
1

NO---->Skip to next page, question 5

2

YES

2.

What has been your

3.

How reliable have you found this information about schedule.d
releases to be? (CIRCLE ONE NUMBER)

H>ST

common means of obtaining this information?

1 Always accurate
2

Usually accurate

3 So-so

4.

4

Seldom accurate

5

Never accurate

How useful have you found this information about scheduled releases
to be? (CIRCLE ONE NUMBER)
1 I always based trip decisions on the information I
received about scheduled release patterns
2

Most of my trip decisions were based on the information I
received about scheduled release patterns

3 In making trip decisions I relied on my own observation
of river conditions as much as information I received
about scheduled release patterns
4

Most of my trip decisions were based on
of river conditions

5

All of my trip decisions were based on my own observation
of river conditions

my

own observation

A.18
5.

Prior to 1983, water was released from the Glen Canyon Dam in a
fairly regular pattern. This pattern often resulted in daily
fluctuations of 20,000 cfs. During the sunmer, the highest flow
releases occurred in the late afternoon or early evening, and
average weekend flow releases were less than average weekday flows.
Would a regular pattern of releases make it easier to conduct a
commercial white water raft trip in the Grand Canyon? (CIRCLE ONE
NUMBER)
1 DEFINITELY NOT

6.

2

PROBABLY NOT

3

PROBABLY YES

4

DEFINITELY YES

Do you have any suggestions on how to improve the ways that water
release patterns and general river information are communicated to
river runners?

A.19

•
RELATIORSHIP BE1'VEEll FLCll LEVELS AllD ACCIDERTS OR THE RIVER
1.

Which of the following factors, in your opinion, are the most
important causes of rafting accidents on the Colorado river?
(PLEASE RANK EACli FACTOR FROM 1 TO 7, RANKING THE MOST IMPORTANT
FACTOR AS 1)
EQUIPMENT FAILURE
_

VERY HIGH WATER (ABOVE 45, 000 CFS)

_VERY LOW WATER (BEl..OW 5,000 CFS)
DAILY FLUCTUATIONS IN WATER LEVEl..
BOATMAN INEXPERIENCE
BOATMAN ERROR
WEATHER

2.

Are there any other factors that should be added to the above list?

3.

Given your experience as a commercial guide, do you feel that
accidents on the Colorado river such as losing equipment, damaging
a boat, or passengers falling out of a boat are more likely to
happen under certain flow levels than at others?
1 NO, I think most water accidents happen INDEPENDENTLY
of flow level
2 YES, I think water accidents are more likely to happen
at the following flow levels (CIRCLE ALL THAT APPLY):
1 flows less than 5,000 cfs

1

2

5,000 - 8,999 cfs

3
4

9,000 - 15,999 cfs

5
6

16,000 - 31,999 cfs
32,000 - 45,000 cfs
above 45,000 cfs

A.20

4.

Do you feel that more SEVERE accidents (such as flipping a boat or
serious injuries to passengers) are more likely to happen under
certain flow levels than at others?
1 NO, I think the severity of the accidents is not related to
now level
2 YES, I think more severe accidents are likely to happen
at the following now levels (CIRCLE ALL THAT APPLY):
1 Flows less than 5,000 cfs
If so, where?
2

~~~~~-~-~~~~~~--

5,000 - 8,999 cfs
If so, where? ~---------------

3 9,000 - 15,999 cfs
If so, where?
4

16,000 - 31,999 cfs
If so, where?

5

32,000 - 45,000 cfs
If so, where?

6 Above 45,000 cfs
If so, where?

In this last section we would like to ask some questions about your
background which will help us compare your answers to those of other
guides.
1.

How old are you? ___years old

2.

How many years have you been a commercial guide in the Grand
canyon?
_ ___,,years

A.21

3.

For how many of these years was the majority of your time spent
guiding oar or motor trips?
~---

years on oar trips
____ years on motor trips

4.

About how many trips in Grand Canyon have you taken as a commercial
guide with each of the following types of boat? (IF NONE FOR A
PARTICULAR TYPE OF BOAT, WRITE IN 0)
trips on motorized rafts
___ trips on rowing rafts
trips on dorie.s

5.

What kind of commercial trips did you run in 1985?
THAT APPLY)

(CIRCLE ALL

1 MOTORIZED RAFT
2

RCYtlING RAFT

3

DORY

6.

In 1985, how many days did your trips usually take? ___ days

7.

At which flow levels have you run commercial Grand Canyon river
trips? (COMPLETE ALL THAT APPLY)
Lowest flow level:

cfs

Highest flow level: - - - - - - cfs
Largest daily change in flow level:

•

- - - - - cfs

If you would like a summary of the questionnaire results, please check
here:
TJIARI: YOU FOR YOUR HELPI

..

B. 1

APPENDIX B
WHITE-WATER PRIVATE TRIP LEADER SURVEY

•

B.2

There are many different alternatives for flow regimes from the
Glen Canyon Dam. The purpose of this study is to find out how
different flows affect river running in the Grand Canyon. As an
experienced boatman, you are the expert and have detailed knowledge of
the effects different flows have on river trips. Regardless of whether
you have taken 1 trip or 100 trips in the Grand Canyon, your opinions
are important. You can contribute your expertise to the
decision-making process by filling out this questionnaire.
We're going to ask you about the effect of different water levels
on the rapids, campsites, and attraction sites, as well as your
preferences for flow regimes in the Grand Canyon. Please answer the
questions from your perspective as an experienced boater and trip
leader.
Please check below the kind of boat you •ost of'ten use for your
trips: (PLEASE CHOOSE ONE TYPE OF BOAT)
___ MOTOR RAFT----------> ___ 33 feet or larger
___ smaller than 33 feet
RCMING RAFT---------> ___ 20 feet or larger
___ less than 20 feet

---

DORY

___ OTHER (please specify - - - - - - - - - - - - - - - - - - ·

Please answer all the remaining questions BASED OR THE IlRD OF BOAT YOU
IRDICATED ABOVE.

B.3
RAPIDS
In this first section, we are interested in learning how different
water levels affect rapids.
1. Assuming you were to run an ERTIRE GRAND C.ANYOR TRIP AT A CORSTART
FLOW, please specify the minimum and maximum water levels for
running rapids safely with a group, the level that provides the best
ride, and the level you prefer as a trip leader for running rapids.
(ARSWER Ill CFS)

Minimum level for running
safely with a group

cfs

Maximum level for running
safely with a group

cfs

Best ride

cfs

fo~

trip members

Water level you prefer
as a trip leader

cfs

2. Which rapids would be most problematic BELOW your minimum safety
level? (IF NONE, PLEASE WRITE NONE)
1
2

3

3. Which rapids would be problematic ABOVE your maximum safety level?
(IF NONE, PLEASE WRITE NONE)

•

2

3

B.4
CAMPSITES

There may be several areas in the Canyon where camps become a problem at
certain water levels. For the items below, please identify the water
levels where camps become a problem.
1. Would certain flow levels cause problems with access to or use of
campsites for you? (CIRCLE ONE)
NO---->Skip to next page, question 1
2

YES

PLEASE FILL IN EACH OF THE BLANKS BELOW IF APPLICABLE
2. Below

cfs, we would have problems getting to camp on time
because we would have to spend too much time traveling on the river.

3. Above

cfs, important camps might be unavailable for use
because they are under water.

4. With daily fluctuations of more than
cfs, we would have
problems with hanging up boats, loading boats, or having to move
camp.

5. Which campsites would be most problematic at low water (9,000 cfs or
less) with the type of boat you most often use? (IF NONE, WRITE
NONE)
1 _ _ _ _ _ _ _ _ _ _ _ _ __

2 --------------

3 --------------

6. Which campsites would be most problematic at high water (32,000 cfs
or more) with the type of boat you most often use? (IF NONE, WRITE
NONE)
1
2 _ _ _ _ _ _ _ _ _ _ __

3 --------------

B.5
TIME FOR ATTRACTIORS

It may be that certain water levels allow more or less time for stops
at attraction sites. For the following items, please indicate which
flow levels, if any, would affect your stops at attraction sites.
1. Would certain flow levels cause problems with access to or use of
attractions for your trip?
NO---->Skip to next page, question 1
2

YES

PLEASE FILL IN EACH OF THE BLANKS BELOW IF APPLICABLE
2. Below
cfs there would not be enough time for stops at
attractions because of the need to "make up time"

3. From

to
cfs the amount of time for stops at
attractions would be "about right"

4. Above
attractions

~~~~

cfs there would be extra time for stops at

5. Which attraction sites would be most problematic (difficult to pull
in, tie up, or load and unload trip members) at low water (9,000
cfs or less)? (IF NONE, WRITE NONE)
1 ~~~~~~~~~~~~~-

2 ~~~~~~~~~~~~~-

3 ~~~~~~~~~~~~~-

6. Which attraction sites would be most problematic (difficult to pull
in, tie up, or load and unload trip members) at high water (32,000
cfs or above)? (IF NONE, WRITE NONE)
1 ~~~~~~~~~~~~~-

2 ~~~~~~~~~~~~~-

3 ~~~~~~~~~~~~~-

B. 6

CORSTART FLCll LEVm.S

1 • How would you, as a private trip leader using the boat you usually
pilot, evaluate each of the following water levels for a private
Grand Canyon river trip? Assume the water level is constant for
the entire trip. (CIRCLE ONE NUMBER FOR EACH WATER LEVEL)

Very
Somewhat
Somewhat
Very
Flow Level Satisfactory Satisfactory Neutral Unsatisfactory Unsatisfactory
2,000 cfs

1

2

3

4

5

3,-000 cfs

1

2

3

4

5

4,000 cfs

1

2

3

4

5

2

3

4

5

5,000 cfs
7,500 cfs

1

2

3

4

5

10,000 cfs

1

2

3

4

5

15,000 cfs

1

2

3

4

5

20,000 cfs

1

2

3

4

5

25, 000 cfs

2

3

4

5

30,000 cfs

2

3

4

5

40,000 cfs

2

3

4

5

2

3

4

5

60,000 cfs

2

3

4

5

80,000 or more

2

3

4

5

50,000 cfs

1

.
"

In the following questions, we would like to know about tolerable
daily fluctuations in flow.

1.

Do you feel that you can accurately describe the daily water level
fluctuations you have experienced in terms of cf s daily change?
1 NO----->Please skip to next page
2

2.

YES

If the flow is somewhere in the range from 5,000 to 9,000 cfs,
what is the largest daily change in flow that is tolerable?
- - - - - cfs daily change

3.

If the flow is somewhere in the range from 9,000 to 16,000 cfs,
what is the largest daily change in flow that is tolerable?
- - - - - cfs daily change

4.

If the flow is somewhere in the range from 16,000 to 32,000 cfs,
what is the largest daily change in flow that is tolerable?
cf s daily change

5.

If the flow is somewhere above 32,000 cfs, what is the largest
daily change in flow that is tolerable?
cfs daily change

'

B.8
ALTERRATIVE FL<ll REGIMES

In this next section, we have listed 4 different scenarios for RELEASE
PATTERNS FROM THE GLEN CANYON DAM. For each one, please check the
disadvantages you perceive as a private trip leader. After considering
all four scenarios, please indicate which one you prefer most.
In these questions, we are interested only in factors related to your
ability to conduct trips, not impacts on the geology or biology of the
Canyon. These scenarios are for discussion only; they are not currently
being proposed as actual operating procedures for the dam.

There would be no daily fluctuations, but flows would
change from one month to the next. Flows during May, June, July, and
August would be 10,000, 10,400, 12,750, and 14,400 cfs respectively,
with no daily fluctuations. The rest of the year, nows would range
from 8,300 cfs - 14,600 cfs, again with no daily fluctuations.

SCENARIO A:

Possible problems (check those which apply to this scenario during May
through August and from September through April)

May-Aug.

Sept.-Apr.
too much time on river
not enough time for attraction sites
no flexibility in running trip
unable to avoid other people at attractions or camps
difficulty with mooring boats
problems running rapids
difficult to plan timing for attraction sites or
camps
inadequate flow
not enough challenge/fun in rapids
lowest flows are too low
other (please

specify)~~~~~~~~~~~~~~-

Your OVERALL evaluation of this flow scenario (CIRCLE ONE):
completely
acceptable

somewhat
acceptable

neutral

somewhat
unacceptable

completely
unacceptable

B.9
SCENARIO B: Flows would be constant at 25,000 cfs during June, July,
and August. During the rest of the year, daily flows could range from
1,000 to 31,500 cfs •

..

Possible problellS (check those which apply to this scenario during
June through August and from September through May)

June-Aug. Sept.-May
flow is too unpredictable
too much time on river
not enough time for attraction sites
no flexibility in running trip
unable to avoid other people at attractions or camps
difficulty with mooring boats
problems running rapids
difficult to plan timing for attraction sites or
camps
inadequate flow
problems with fluctuations at camps (hanging-up
boats, etc. )
not enough challenge/fun in rapids
lowest flows are too low
other (please specify)

~~~~~~~~~~~~~~-

Your OVERALL evaluation of this flow scenario (CIRCLE ONE):
completely
acceptable

'

somewhat
acceptable

neutral

somewhat
unacceptable

completely
unacceptable

B.10
Flows would vary by day, season, and month. During June,
July, and August, daily flows could range from 3,000 to 31,500 cfs
with a major peak at 3:00 p.m. During the rest of the year, daily
flows could range from 1 1 000 to 31,500 cfs.
SCBRARIO C:

.,

..
Possible problems (check those which apply to this scenario during
June through August and from September through May)

June-Aug. Sept.-May
flow is too unpredictable
too much time on river
not enough time for attraction sites
no flexibility in running trip
unable to avoid other people at attractions or camps
difficulty with mooring boats
problems running rapids
difficult to plan timing for attraction sites or
camps
inadequate flow
problems with fluctuations at camps (hanging-up
boats, etc.)
not enough challenge/fun in rapids
lowest flows are too low
other (please specify)_______________

Your OVERALL evaluation of this flow scenario (CIRCLE ONE):
completely
acceptable

sanewhat
acceptable

neutral

somewhat
unacceptable

completely
unacceptable

B. 11

•

SCEKARIO C: Flows would vary by day, season, and month.
During June,
July, and August, daily flows could range from 3,000 to 31,500 cfs
with a major peak at 3:00 p.m. During the rest of the year, daily
flows could range from 1,000 to 31,500 cfs.

Possible problellS (check those which apply to this scenario during

June through August and from September through May)
June-Aug. Sept.-May
flow is too unpredictable
too much time on river
not enough time for attraction sites
no flexibilfty in running trip
unable to avoid other people at attractions or camps
difficulty with mooring boats
problems running rapids
difficult to plan timing for attraction sites or
camps
inadequate flow
problems with fluctuations at camps (hanging-up
boats, etc.)
not enough challenge/fun in rapids
lowest flows are too low
other (please specify)

~~~~~~~~~~~~~~-

Your OVERALL evaluation of this flow scenario (CIRCLE ONE):
completely
acceptable

11

somewhat
acceptable

neutral

somewhat
unacceptable

completely
unacceptable

B.12
SCEllARIO D:

Flows would vary by day, season, and month. Throughout
the year, daily flows could range from 8,000 to 25,000 cfs. During
the s\.111mer there would be a major peak around 3:00 p.m.
Possible problems (check those which apply to this scenario)
~~-

flow is too unpredictable

~~-

too much time on river

~~-

not enough time for attraction sites

~~-

no flexibility in running trip

~~-

unable to avoid other people at attractions or camps

~~-

difficulty

~~-

problems running rapids

~~-difficult

~~-

wit~

mooring boats

to plan timing for attraction sites or camps

inadequate flow

~~-problems

with fluctuations at camps (hanging-up boats, etc.)

~~-

not enough challenge/fun in rapids

~~-

lowest flows are too low

~~-other

(please specify)~~~~~~~~~~~~~~~~-

Your OVERALL evaluation of this flow scenario (CIRCLE ONE):
completely
acceptable

scmewhat
acceptable

neutral

somewhat
unacceptable

completely
unacceptable

•

B.13

•
1. OVERALL, how would you rank these four flow scenarios (A-D) from
your perspective as a river guide? (PLEASE RANK EACH SCENARIO)

•

Scenario
Letter
Most preferred scenario
Next most preferred scenario
Third most preferred scenario
Least preferred scenario

2. Imagine that you have a private permit for next year and you could
choose one of the four water flow scenarios as well as the date for
the trip. What water flow scenario and what month would you
choose?
Scenario letter
Month

B. 14

GENERAL IMPACTS OB TRIP

1. As a private trip leader, have you experienced constant flow levels
during any of your Grand Canyon trips?
NO---->Skip to next page, Question 3
2

YES

2. At what <X>RSTART Ft.Oil LEYF.LS in the Grand Canyon have you had to do

any of the following? Assume each water level is a constant flow
for the entire trip. FOR EACH ITEM, PLEASE CHECK ALL WATER LEVELS
THAT APPLY.
--------Constant Flow Level (cfs x 1000)-----0ver
1-5 5-10 10-15 15-20 20-25 25-35 35-50 -2.Q
Stop at major rapids
such as Houserock, Hance,
Crystal or Lava to scout
Stop at any other rapids
to scout
Row or motor more than
usual to make up some
time
Row less than usual or
turn off motor because
you are ahead of schedule
Hurry to get out of
camp in the morning
Have trip members walk
around a rapid because
the water is too low
Have trip members walk
around a rapid because
the water is too high

..

---------------,.-------------------------~------------

B.15

•

3. As a private boater, have you experienced daily fluctuating flow
levels during any of your Grand Canyon trips?
NO---->Skip to next page, Question 1

...

2

YES

4. Assume that flow levels were varying from a low of 3,000 cfs to a
high of 25,000 cfs each day (or within a 24 hour period). UNDER
THESE Fl.CM CONDITIONS, which of the following would you be likely to
do? (PLEASE CIRCLE ONE NUMBER FOR EACH ITEM)
Very
Likely

Somewhat
Likely

1

2

3

Select specific campsites because
they offer better protection
against water level changes

2

3

Camp above a major rapid to wait
for the water level to change
1

2

3

Change your itinerary to try to
reach certain points at a
good time

1

2

3

Check on the boat mooring
during the evening to see if it
should be moved

1

2

3

Spend more time at scheduled
attraction stops

1

2

3

Spend less time at scheduled
attraction stops

1

2

3

2

3

2

3

Stop to scout rapids more
often than usual

,

Spend less time

in

Make camp earlier

camp
1

Not at all
Likely

B.16

EXTRA TIME OB YOUR TRIP
1.

If your trip was making good time and 6-8 fewer hours of actual
river time was needed to canplete the trip in the scheduled time,
which of the following would you be most likely to do? (CIRCLE ONE
NUMBER FOR EACH ITEM)
Not at all
Likely

Somewhat
Likely

Very
Likely

Stay in camp later in the
morn;i.ng

1

2

3

Make camp earlier in the
afternoon

1

2

3

Take longer lunches

1

2

3

Spend more time scouting
rapids

1

2

3

Spend more time at scheduled
attraction stops

1

2

3

Stop at additional attraction
sites

1

2

3

Spend a layover day at a
campsite

1

2

3

B.17

•

•

2.

If your trip was making good time and you decided to stop at
ADDITIONAL attraction sites (sites you don't usually visit), what
are the three additional attraction sites you would most likely
stop?
(FOR EACH ATTRACTION SITE, ALSO INDICATE WHAT YOU WOULD SUGGEST
THAT TRIP MEMBERS DO. PLEASE TRY TO BE SPECIFIC--IF YOU WOULD
SUGGEST A HIKE, INDICATE WHERE (PLACE NAME) YOU WOULD RECOMMEND
THAT THEY HIKE ) •

A'l'TRACfIOR SITE - - - - - - - - - -

River Mile

What would you suggest that trip members do here?

ATTRAC'l'ION SITE - - - - - - - - - -

River Mile

What would you suggest that trip members do here?

ATTRACfION SITE - - - - - - - - - - -

River Mile

What would you suggest that trip members do here?

11

B. 18

3. If your trip was making good time and you decided to spend an EXTRA
layover day somewhere, what are the three campsites at which you
would be most likely to layover for a day.
(FOR EACH CAMPSITE, ALSO INDICATE WHAT YOU WOULD SUGGEST THAT TRIP
MEMBERS DO. PLEASE TRY TO BE SPECIFIC--IF YOU WOULD SUGGEST A DAY
HIKE, INDICATE WHERE (PLACE NAME) YOU WOULD RECOMMEND THAT THEY
HIKE).
River Mile

What would you suggest that trip members do here on a layover
day?

CAMPSITE - - - - - - - - - - -

River Mile

What would you suggest that trip members do here on a layover
day?

CAMPSITE - - - - - - - - - - -

River Mile

What would you suggest that trip members do here on a layover
day?

•

B. 19

•
IHFORMATIOH ON FLOW RELEASF.S
1. In the pa.st, have you used information concerning scheduled Glen

Canyon Dam releases as a basis for making decisions about your raft
trips in the Grand Canyon(e.g. when to run particular rapids, when
to camp, where to moor boats, etc.)? (CIRCLE ONE NUMBER)
1 NO---->Sld.p to next page, question 5
2

YES

2.

What has been your

3.

How reliable have you found this information about scheduled
releases to be? (CIRCLE ONE NUMBER)

H>ST

common means of obtaining this information?

Always accurate
2

Usually accurate

3 So-so

4.

4

Seldom accurate

5

Never accurate

How useful have you found this information about scheduled releases
to be over? (CIRCLE ONE NUMBER)
1 I always based trip decisions on the information I
received about scheduled release patterns
2

Most of my trip decisions were based on the information I
received about scheduled release patterns

3

In making trip decisions I relied on my own observation
of river conditions as much as information I received
about scheduled release patterns

4

Most of my trip decisions were based on my own observation
of river conditions

5

All of my trip decisions were based on my own observation
of river conditions

B.20

5.

Prior to 1983, water was released from the Glen Canyon Dam in a
fairly regular pattern. This pattern often resulted in daily
fluctuations of 20,000 cfs. During the sunmer, the highest flow
releases occurred in the late afternoon or early evening, and
average weekend flow releases were less than average weekday flows.
Would a regular pattern of releases make it easier to conduct a
private white water raft trip in the Grand Canyon? (CIRCLE ONE
NUMBER)
1 DEFINITELY NOT

6.

2

PROBABLY NOT

3

PROBABLY YES

4

DEFINITELY YES

Do you have any suggestions on how to improve the ways that water
release patterns and general river information are communicated to
river runners?

•

B.21

•

RELATIONSHIP BE"l'WEEH FLOW LEVELS AID> ACCIDEBTS OH THE RIVER
1.

Which of the following factors, in your opinion, are the most
important causes of rafting accidents on the Colorado river?
(PLEASE RANK EACH FACTOR FROM 1 TO 7, RANKING THE MOST IMPORTANT
FACTOR AS 1)
EQUIPMENT FAILURE
_VERY HIGH WATER (ABOVE 45,000 CFS)
_VERY LOW WATER (BELOW 5,000 CFS)
DAILY FLUCTUATIONS IN WATER LEVEL
BOATMAN INEXPERIENCE
BOATMAN ERROR
WEATHER

2.

Are there any other factors that should be added to the above list?

3.

Given your experience as a boater, do you feel that accidents on
the Colorado river, such as losing equipment, damaging a boat, or
people falling out of a boat are more likely to happen under
certain flow levels than at others?
NO, I think most water accidents happen INDEPENDENTLY
of flow level
2 YES, I think water accidents are more likely to happen
at the following flow levels (CIRCLE ALL THAT APPLY):
1 Flows less than 5,000 cfs

,.

2

5,000 - 8,999 cfs

3
4

9,000 - 15,999 cfs
16,000 - 31,999 cfs

5 32,000 - 45,000 cfs
6 Above 45,000 cfs

B.22

4.

Do you feel that more SEVERE accidents (such as flipping a boat or
serious injuries to trip members) are more likely to happen under
certain flow levels than at others?

.

1 NO, I think the severity of the accidents is not related to
flow level
2

YES, I think more severe accidents are likely to happen
at the following flow levels (CIRCLE ALL THAT APPLY):
1 Flows less than 5,000 cfs
If so, where?
2

5,000 - 8,999 cfs
If so, where?

3 9,000 - 15,999 cfs
If so, where?
4

16,000 - 31,999 cfs
If so, where?

5

32,000 - 45,000 cfs
If so, where?

6 Above 45,000 cfs
If so, where?

In this final section we would like to ask some questions about your
background which will help us compare your answers to those of other
guides.
1.

How old are you? ___ years old

2.

How many years of white water boating trip leader experience have
you had?
_ ___.,years

ir

B.23

3.

About how many Grand Canyon river trips have you taken as a
passenger or trip leader with each of the following types of boat?
(IF NONE FOR A PARTICULAR TYPE OF BOAT, WRITE IN 0)
trips on motorized rafts
___ trips on rowing rafts
___ trips on dories

4.

At which flow levels have you run Grand Canyon river trips?
(COMPLETE ALL THAT APPLY)
Lowest flow level:
Highest flow level:

- - - - - - cfs
cfs

Largest daily change in flow level:

cfs

If you would like a summary of the questionnaire results, please check
here: _ __

TILUIK YOU FOR muR HELP!

c. 1

.APPENDIX C

WHITE-WATER BOATERS ATTRIBUTE SURVEY

C.2

In this first section, we are interested in finding out about your white
water raft trip in the Grand Canyon and how much you enjoyed it.
1•

Overall, how would you rate your raft trip?

(CIRCLE ONE NUMBER)

POOR
2

FAIR, it just didn't work out very well

3
4

GOOD, but a number of things could have been different

5

EXCELLENT, only minor problems

6

PERFECT

VERY GOOD, but could have been better

2.

What things would contribute most to an excellent or perfect raft
trip in the Grand Canyon for you?

3.

What things would contribute most to a poor raft trip in the
Grand Canyon for you?

4.

When you first decided to take a Grand Canyon trip, what was the
ORE thing you looked forward to most?

,

C.3

5.

Where did you put-in (start trip)?

(CIRCLE ONE NUMBER)

LEE'S FERRY

"
6.

2

PHANTOM RAN CH

3

OTHER (specify

Where did you take-out (end trip)?

(CIRCLE ONE NUMBER)

PHANTOM RANCH

2

WHITMJRE WASH

3

DIAMOND CREEK

4

LAKE MEAD

5

OTHER (specify

7.

How long was your trip?

8.

Including yourself, about how many people were there on this raft
trip? (PLEASE INCLUDE THE GUIDE/TRIP LEADER AND ALL OF THE BOATS
IN YOUR GROUP)

----DAYS

_ _ _ _PEOPLE

9.

What type of boat were you on?

(CIRCLE ONE NUMBER)

MOTOR POWERED RAFT

.. 7

2

OAR POWERED RAFT

3

COMBINATION MOTOR/OAR RAFT

4

DORY

5

KAYAK/CANOE

6

PADDLE RAFT

7

OTHER (specify

C.4
10. Was your Grand Canyon raft trip:

(CHOOSE ONE)

RUN BY A COMMERCIAL OUTFITTER
2

A PRIVATE TRIP---->Were you primarily responsible for
operating a boat on this trip?
YES
2

NO

11. Including this trip, how many times have you rafted the Colorado
River below Lee's Ferry?
_____TIMES

12. Besides your Grand Canyon trip(s), how many white water raft or
kayak trips have you taken at other locations? (CIRCLE ONE
NUMBER)
1

NONE

2

1-2

3

3-5
6-10
11-20
MORE THAN 20

4

5
6

13. If you had the opportunity, would you take a Grand Canyon white
water raft trip again? (CIRCLE ONE NUMBER)
DEFINITELY NOT
2

PROBABLY NOT

3

PROBABLY YES

4

DEFINITELY YES

'·

C.5
14. River trips through the Grand Canyon have a number of features.
People differ in what they feel is important for them
personally. In this next section, we list a number of features
of a Grand Canyon river trip. Please indicate how important each
feature was for you on your trip. (CIRCLE ONE NUMBER FOR EACH
ITEM)
Not at all Somewhat Very
Didn't
Important Important Important Experience
Observing flora, fauna, and
geology
Being in a natural setting

1

Being on the Colorado River

2

3

0

2

3

0

2

3

0

Being with family/friends

1

2

3

0

Relaxing; getting away from
it all

1

2

3

0

Camping on sandy beaches

2

3

0

Large rapids

2

3

0

Stopping at side canyons or
creeks

2

3

0

Learning about the history of
the Grand Canyon

2

3

0

Photographing the Grand Canyon

2

3

0

2

3

0

Hiking in the side canyons

2

3

0

Floating on quiet stretches of
the river
1

2

3

0

Seeing wildlife

2

3

0

Interacting with my guide or
trip leader

2

3

0

Visiting archeological sites

2

3

0

Seeing few other people while
floating

1

C.6
Not at all Somewhat Very
Didn't
Important Important Important Experience
Sense of accomplishment of
making it through the trip

1

2

3

0

Feeling safe

1

2

3

0

Confidence in my guide or
trip leader

1

2

3

0

2

3

0

2

3

0

Interacting with others
on my trip

2

3

0

Seeing few other people at beaches
or attraction sites
1

2

3

0

Fishing in the Grand Canyon

2

3

0

Good weather
Good food

1

Did we miss anything else that was

important?~~~~~~~~~~

15. On average, how crowded did you feel the river was while you were
floating? (Circle the number on the scale best representing your
feelings.)
2

not at all
crowded

3

4

slightly
crowded

5

6

7

moderately
crowded

8

extremely
crowded

16. On your trip, did you feel you had enough time to hike the side
canyons and see other attractions? (CIRCLE ONE NUMBER)
1 YES, THERE WAS ENOUGH TIME FOR HIKING
2

NO, THERE WAS NOT ENOUGH TIME FOR HIKING

3

THERE WAS TOO MUCH TIME FOR HIKING

9

..

c.1

Rapids are an important part of the Grand Canyon trip for some
people. In this next section, we would like to get your expectations
and feelings about the rapids.

1.

What role did rapids play in your decision to take this trip?
(CIRCLE ONE NUMBER)
1 RAPIDS WERE THE M:>ST IMPORTANT REASON FOR TAKING THE
TRIP
2

RAPIDS WERE ONE OF THE TWO OR THREE M:>ST IMPORTANT
REASONS FOR TAKING THE TRIP

3 RAPIDS WERE ONLY ONE OF MANY IMPORTANT REASONS FOR
TAKING THE TRIP
4

2.

RAPIDS WERE NOT AN IMPORTANT REASON FOR TAKING THE TRIP

Would you say the rapids you encountered on your trip were:
(CIRCLE ONE NUMBER)
1 SMALLER THAN YOU EXPECTED
2 BIGGER THAN YOU EXPECTED

3.

3

ABOUT WHAT YOU EXPECTED----->Skip to Question 4

4

HAD NO EXPECTATIONS--------->Skip to Question 4

If the rapids were smaller or bigger than you expected, how did
you feel about it? (CIRCLE ONE NUMBER)
LIKED IT
2 DIDN'T MAKE ANY DIFFERENCE
3

DIDN'T LIKE IT

C.8

4.

On this trip, did you have to wait above any rapid before running
it?

•

NO
2 YES---->Did you have to wait for:
APPLY)

(CIRCLE ALL THAT

OTHER BOATS TO GO THROUGH

5.

WATER LEVEL

3

OTHER---------------

TO

CHANGE

Did you have to walk around any rapids?
1

2

6.

2

NO
YES---->Which rapids?_ _ _ _ _ _ _ _ _ _ _ _ _ _ _~

In general, which type of rapid did you enjoy most on this trip:
(CHOOSE ONE)
BIG RAPIDS

7.

2

MEDIUM RAPIDS

3

SMALL RAPIDS

4

LIKED ALL TYPES OF RAPIDS EQUALLY

5

DON'T LIKE RAPIDS AT ALL---->Skip to Question 8

What is the ONE thing you liked most about the rapids on this
Grand Canyon trip? (PLEASE TRY TO BE SPECIFIC)

•

..

c.9
8.

For each of the following rapids, please indicate how you felt
about that particular rapid. If you didn't run it or don't
remember it, circle the number 11 011 • (SEE THE ENCLOSED MAP OF
THESE RAPIDS)

Name of Rapid

Don't
Somewhat Strongly Remember or
Strongly Somewhat
Like
Didn't Run
Disliked Disliked Neutral Liked

House Rock Rapid

1

Hance Rapid

2

3

4

5

0

2

3

4

5

0

Horn Creek Rapid

1

2

3

4

5

0

Hermit Rapid

1

2

3

4

5

0

Crystal Rapid

1

2

3

4

5

0

2

3

4

5

0

Lava Falls Rapid

9.

Below are a number of characteristics of rapids which you may or
may not have experienced on this trip. For each characteristic,
please indicate how it affects your enjoyment of a rapid.
(CIRCLE ONE NUMBER FOR EACH CHARACTERISTIC)
Greatly
Somewhat
Somewhat Greatly
Decreases Decreases Doesn't Increases Increases
Enjoyment Enjoyment ~.atter Enjoyment Enjoyment

•

Roller coaster ride
(standing waves)

2

3

4

5

Rapid with large waves

2

3

4

5

Rapid with small waves

2

3

4

5

Rocks sti eking out of
water

2

3

4

5

Getting hit by water

2

3

4

5

Force of the water
that hits you

2

3

4

5

C.10
Greatly
Somewhat
Somewhat Greatly
Decreases Decreases Doesn't Increases Increases
Enjoyment Enjoyment Matter Enjoyment Enjoyment
Concern about damage
to personal equipment

1

2

3

4

5

1

2

3

4

5

Fear of tipping over

2

3

4

5

Fear of falling out
of boat and being in
the water for a long
time

2

3

4

5

Hanging onto the boat to
avoid being tossed out

Large number of rapids

1

2

3

4

5

Long rapid

1

2

3

4

5

Short fl ca ting time
between rapids

2

3

4

5

Waiting at a rapid for
other trips to run it

2

3

4

5

2

3

4

5

Having to walk around a
rapid

2

3

4

5

Learning how to "read"
rapids from the guide
or trip leader

2

3

4

5

2

3

4

5

Having to avoid tricky
eddies and holes

Paddling/rowing through
the rapids
Other

1

1

.•

c. 11

•
Besides rapids, the water level on the river may also affect a
person's trip. In this next section, we are interested in your
perceptions of the water level during your trip.

1.

When you were planning your trip, did you know before you left
home what the expected water level was for the dates of your
trip?
1

NO

2

YES--->Did this information about the expected water
level have any influence on your decision WHEN
to take this trip?
1

NO

2 YES (please explain- - - - - - -

2.

Overall, did you expect the water level on this trip to be:
LOWER THAN IT WAS
2 ABOUT THE SAME AS IT WAS
3

4

3.

HIGHER THAN IT WAS
DIDN'T KNOW WHAT TO EXPECT

If you had the choice, would you have preferred the overall water
level to be: (CIRCLE ONE NUMBER)
1 LOWER
2

ABOUT THE SAME

HIGHER
4 DON'T KNOW OR DOESN'T MATTER
3

4.

Overall, was the speed of the water (current):
TOO SLOW
2 TOO FAST
3

ABOUT RIGHT

4

DON'T KNOW

c. 12
5.

Did you notice the water level change during your trip?
1

NO

2 YES---->How often did you notice it changing?
(CIRCLE ONE)
EVERY DAY
2

ALMOST EVERY DAY

3 ONLY ON A FEW DAYS
---->What made you aware of the water level change?

6.

Do you think that daily fluctuations in the water level would
make you feel more or less like you were in a natural setting?
(CIRCLE ONE NUMBER)
MUCH t-DRE LIKE A NATURAL SETTING
2 SOMEWHAT MORE LIKE A NATURAL SETTING
3

WOULDN'T HAVE ANY EFFECT

4 SOMEWHAT LESS LIKE A NATURAL SETTING

1.

5

MUCH LESS LIKE A NATURAL SETTING

6

DON'T KNOW

Ideally, how often would you prefer that the water level change
during a trip? (CIRCLE ONE NUMBER)
1

EVERY DAY

2

ALMOST EVERY DAY

3
4

ONLY ON A FEW DAYS
NEVER CHANGE

5

DON'T KNCM

6 DON'T CARE ABOUT WATER LEVEL CHANGE

C.13

•

,

In addition to your general feelings about your Grand Canyon trip, we
would like to know how you felt about about some specific rapids on
this trip.

1.

Did you encounter Hance Rapid on your Grand Canyon raft trip?
1

NO---->Skip to Question 2

2

DON'T REMEMBER---->Skip to Question 2

3

YES--->Please indicate whether each of the following
statements was true when you were at Hance Rapid.
(PLEASE CIRCLE ONE NUMBER FOR EACH STATEMENT)

Had to wait for water to rise
before running this rapid

Tus

No

Don't
Remember

1

2

0

2

0

Had to wait for water level
to drop before running this rapid
Water level was low when we
ran this rapid

1

2

0

Water level was medium when
we ran this rapid

1

2

0

2

0

2

0

We stopped to scout this rapid
before running it

2

0

It was difficult to avoid
obstacles on this rapid

2

0

2

0

Water level was high when
we ran this rapid
We had large waves when running
this rapid

We had to walk around this rapid

1

1

[Hance Rapid is between mile 76 and mile 77, about 11 miles above
Phantom Ranch]

C.14

2.

Did you encounter Crystal Rapid on your Grand Canyon raft trip?
1

NO---->Skip to Question 3

2

DON'T REMEMBER---->Skip to Question 3

3

YES--->Please indicate whether each of the following
statements was true when you were at Crystal Rapid
(PLEASE CIRCLE ONE NUMBER FOR EACH STATEMENT)
Don't
Yes
No
Remember

Had to wait for water to rise
before running this rapid
Had to wait for water level
to drop before running this rapid

1

Water level was low when we
ran this rapid

2

0

2

0

2

0

Water level was medium when
we ran this rapid

1

2

0

Water level was high when
we ran this rapid

1

2

0

We had large waves when running
this rapid

1

2

0

We stopped to scout this rapid
before running it

2

0

It was difficult to avoid
obstacles on this rapid

2

0

We had to walk around this rapid

2

0

[Crystal Rapid is between mile 98 and mile 99 on the river, about
10 miles below Phantom Ranch.]

C.15

3.

Did you encounter Lava Falls Rapid on your Grand Canyon raft
trip?
NO---->Skip to Question 4
2

DON'T REMEMBER---->Skip to Question 4

3

YES--->Please indicate whether each of the following
statements was true when you were at Lava Falls
Rapid. (PLEASE CIRCLE ONE NUMBER FOR EACH
STATEMENT)
No

Don't
Remember

Had to wait for water to rise
before running this rapid

2

0

Had to wait for water level
to drop before running this rapid

2

0

Water level was low when we
ran this rapid

2

0

Water level was medium when
we ran this rapid

2

0

Water level was high when
we ran this rapid1

2

0

2

0

We stopped to scout this rapid
before running it

2

0

It was difficult to avoid
obstacles on this rapid

2

0

We had to walk around this rapid

2

0

Yes

We had large waves when running
this rapid

1

[Lava Falls Rapid is between mile 179 and mile 180 on the river,
about 20 miles below Havasu Creek.]

c. 16
4.

Hance, Hermit, Crystal, and Lava Falls are four rapids which many
people remember after a Grand Canyon trip. These rapids are
somewhat different at different water levels:
at low water:
(9000 CFS or less)

Waves are smaller, water is slower, and
some rocks are exposed. Trips sometimes
have to wait for several hours for the
water level to rise before running the
rapid. The guide or trip leader often
has to do some tricky maneuvering.

at medium water:
(9,000-16,000 CFS)

Waves are somewhat larger and the water
is faster. The guide or trip leader does
less maneuvering, but there still may be
some rocks or obstacles to avoid.

at medium high water:
(16,000-32,000 CFS)

The waves are very large and the water is
fast. The guide or trip leader often
stops to scout the rapid, and may route
away from the biggest waves to reduce the
risk of flipping the raft.

at high water:
(above 32,000 CFS)

Waves are extra large and the probability
of flipping oar and paddle rafts is
greater. The guide or trip leader
usually stops to scout the rapid, and
passengers on oar powered trips often
walk around the rapid.

If you had your choice of low, medium low, medium high, or high
water, which would you prefer for running these four rapids?
(CHOOSE ONE )
LOW WATER
2

MEDIUM WATER

3

MEDIUM HIGH WATER

4 HIGH WATER
5

MAKES NO DIFFERENCE TO ME

f

c.11
5.

•

As the water level changes, the number and size of rapids are
affected. At higher water, there are fewer, but larger rapids.
The opposite is true at lower water. If you had to, which of the
following situations would you choose on the river? (CHOOSE ONE
ONLY)
THE SAME SIZE AND NUMBER I FOUND ON MY TRIP (same water
level)
2 FEWER BUT BIGGER RAPIDS

(higher water)

3

MORE BUT SMALLER RAPIDS

(lower water)

4

MAKES NO DIFFERENCE TO ME AS LONG AS THERE ARE RAPIDS

In this section, we are interested in learning about your guide or
trip leader. [ROTE: Ir you were the guide or trip leader on this
trip, please answer questions 1-3 as you remember doing these
things.]

1.

Which of the following did you hear the guide or trip leader(s)
say they were concerned about? (CIRCLE ALL THAT APPLY)
WATER WAS TOO HIGH
2 WATER WAS TOO LOW
3

RAPIDS WERE TOO BIG

4

LACK OF RAPIDS

5

TOO MANY OBSTACLES IN THE RAPIDS

6 FLUCTUATING WATER MADE MOORING DIFFICULT
7 WATER LEVELS FLUCTUATED TOO MJCH
8 WATER WAS TOO FAST
9

WATER WAS TOO SLOW

10 TOO MUCH TIME SPENT ROWING OR WITH M'.JTOR ON
11 LARGE STANDING WAVES
12 DANGEROUS HOLES OR EDDIES IN THE RAPIDS
13 NONE OF THE ABOVE

c. 18
2.

3.

Did your guide or trip leader ever tell you to move your tent or
gear higher on the beach to avoid water level changes?
1

YES

2

NO

•

During your trip, how often do you recall that your guide or trip
leader(s) did the following things? (CIRCLE ONE NUMBER FOR EACH
ITEM)
Never
Did
Took an easier route to avoid
rocks or a big hole

Sometimes
Did

Often Didn't
Did Notice

2

3

0

Waited above a rapid for water
level to change

2

3

0

Stopped at a rapid to scout it

2

3

0

1

Waited above a rapid for another
party to run it

1

2

3

0

Rowed or motored more than usual
to make up some time

1

2

3

0

Rowed or motored less than usual
because you were ahead of schedule

1

2

3

0

Hurried to get out of camp
in the morning

1

2

3

0

2

3

0

2

3

0

Selected certain campsites because
the water level might change

2

3

0

Had boat(s) float away

2

3

0

Slept on the boat so mooring could
be changed during the night
Moved boat mooring during the night

1

C.19
Sometimes
Did

Never
Did

#

Of ten Didn't
Did Notice

Had to drag boat(s) off
the beach be ca use water 1 ev el
changed during the night

1

2

3

0

Had you walk around a rapid
because the water was too
high or too low

1

2

3

0

Other

We are also interested in learning about the campsites, vegetation,
beaches, and wildlife you encountered on your raft trip.

1.

How important were each of the following campsite characteristics
to you on your river trip in the Grand Canyon? (CIRCLE ONE
NUMBER FOR EACH CHARACTERISTIC)
Didn't
Not at all Somewhat Very
Important Important Important Experience
Natural appearance

2

3

0

2

3

0

Space between sleeping areas

2

3

0

Sandy area to sleep on

2

3

0

2

3

0

Privacy (change clothes, etc. )

2

3

0

Vegetation at beaches

2

3

0

Large campsite

2

3

0

Flat area to sleep on

A flat beach by the river

•

1

1

C.20

Not at all Somewhat Very
Didn't
Important Important Important Experience
Space for recreation

1

2

3

0

Large boulders/rocks

1

2

3

0

Convenient kitchen setups

2

3

0

Shade at camping areas

2

3

0

Scenic view of the Canyon

2

3

0

Cl€an, unlittered campsites

1

2

3

0

Place to dock boat(s)

1

2

3

0

Eddies or calm backwater

1

2

3

0

2

3

0

2

3

0

Side canyons for hiking

2

3

0

Clear water in side canyons

2

3

0

Nearness to river

2

3

0

2

3

0

Nearness to quiet water

2

3

0

Nearby archeological sites

2

3

0

Isolation from other groups

2

3

0

Presence of wildlife

2

3

0

Lack of dead vegetation
Few flies or mosquitoes

Nearness to rapids

2.

1

1

Thinking back on your Grand Canyon raft trip, what was the ONE
most important characteristic of a campsite for you personally?
TRY TO BE SPECIFIC.

C.21

3.

Did you ever feel crowded at any of the campsites?
1 NO
2 YES----->Why? (CIRCLE ALL THAT APPLY)
1 OUR GROUP WAS TOO LARGE FOR THE CAMPSITE
2 OTHER GROUPS WERE USING THE CAMPSITE
OTHER GROUPS WERE VISIBLE FROM OUR CAMPSITE
4 OTHER (

3

~~~~~~~~~~~~~~~~-

In this last section we would like to ask you some questions about
your background which will help us compare your answers to those of
other people. We would stress that all of your answers are strictly
confidential.
1.

How old are you?

2.

Are you:

old

MALE
.2

3.

--~years

FEMALE

How many years of school have you completed? (CHOOSE ONE)
2

3

~~some

4

5

6

7

8

9

college or technical school

_ _B.A. or equivalent
_ _M.A. or equivalent
_ _Advanced degree (M.D. Ph.D., etc.)

10

11

12

C.22
4.

Please circle the response that comes closest to your total
income before taxes. If ~ou are a student and unmarried,
please give your parents' income.

famil~

Less than $10,000

7

$50,000 to $59,999

2

$10,000 to $17,499

8

$60,000 to $69,999

3

$17,500 to $24,999

9

$70,000 to $79,999

4

$25,000 to $32,499

10

$80,000 to $89,999

5

$32,500 to $39,999

11

$90,000 to $99,999

6

$40,000 to $49,999

12

$100,000 or more

•

•

TB.ARK YOU FOR YOUR H:EI.P !

___ Check here i f you would like a summary of the results of this
survey.

'
Please return this survey in the enclosed envelope to:
HBRS

4513 Vernon Boulevard
Madison, WI 53705

D. 1

•
APPENDIX D
VHITE-VATER COMMERCIAL BOATERS COKTIRGENT-VALUATIOB SURVEY

D.2

•

This questionnaire refers to the white water trip you took in the
Grand Canyon that started on
refer to this trip when answering the questions in this survey.

Please
It is

important that this survey be completed by the person to whom it was
sent.

D.3
page 1
In this first section, we are interested in finding out about your
white water trip in the Grand Canyon and how much you enjoyed it.

A1. overall, how would you rate your white water trip? (CIRCLE ONE NUMBER)
1 POOR
2 FAIR, it just didn't work out very well
4

5 EXCELLENT, only minor problems
6

PERFECT

A2. Where did you put-in (start trip)?

(CIRCLE ONE NUMBER)

1 LEE'S FERRY
2

PHANTOM RANCH

3

OTHER ( s p e c i f y - - - - - - - - - - -

A3. Where did you take-out (end trip)?

(CIRCLE ONE NUMBER)

1 PHANTOM RANCH
2 WHITMORE WASH
3 DIAIDND CREEK
4 LAKE MEAD

5

OTHER {specify

A4. How long was your trip? _ _ _ _ _ DAYS
A5. What type of boat were you on? (CIRCLE ONE NUMBER)
1 IDTOR POWERED RAFT
2 OAR POWERED RAFT

•

3
4

COMBINATION MJTOR/OAR RAFT
DORY

5 KAYAK
6 PADDLE RAFT
7

OTHER (specify

D.4
page 2
A6. Was your Grand Canyon white water trip:

(CIRCLE ONE NUMBER)

1

RUN BY A COMMERCIAL OUTFITTER

2

A PRIVATE TRIP-->Were you primarily responsible for operating
a boat on this trip?
1

YES

2

NO

A7. How many times have you taken a white water trip on the Colorado
River below Lee's Ferry, including this trip?
-----TIMES

A8. If you had the opportunity, would you take a Grand Canyon white
water trip again? (CIRCLE ONE NUMBER)
1 DEFINITELY NOT
2

PROBABLY NOT

3

PROBABLY YES

4

DEFINITELY YES

A9. On average, how crowded did you feel the river was while you were
floating? (Circle the number on the scale best representing your
feelings.)
1

2

not at all
crowded

3

slightly
crowded

4

5

6

moderately
crowded

7

8

9

extremely
crowded

i

D.5
page 3
Rapids are an important part of the Grand Canyon trip for many people.
In this next section, we would like to get your expectations and
feelings about the rapids.

A10. What role did rapids play in your decision to take this trip?
(CIRCLE ONE NUMBER)
1 RAPIDS WERE THE M)ST IMPORTANT REASON FOR TAKING THE TRIP
2 RAPIDS WERE ONE OF THE TWO OR THREE MOST IMPORTANT REASONS FOR
TAKING THE TRIP
3 RAPIDS WERE ONLY ONE OF MANY IMPORTANT REASONS FOR TAKING THE
TRIP
4 RAPIDS WERE NOT AN IMPORTANT REASON FOR TAKING THE TRIP

A11. Did you have to walk around any rapids?
1 NO
2 YES---->Which rapids?_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

A12. In general, which type of rapid did you enjoy most on this trip:
(CHOOSE ONE)
BIG RAPIDS
2 MEDIUM RAPIDS
3 SMALL RAPIDS
4 LIKED ALL TYPES OF RAPIDS EQUALLY

5 DON'T LIKE RAPIDS

•

D.6
page 4
Besides rapids, the water level on the river may also affect a
person's trip. In this next section, we are interested in your
feelings about the water level during your trip.

•

A13. If you had the choice, would you have preferred the overall water
level to be: (CIRCLE ONE NUMBER)
1 LOWER
ABOUT THE SAME
3 HIGHER
4 DON'T KNOW OR DOESN'T MATTER
2

A14.

Did you notice whether the water level changed during your trip?
1 NO
2

YES---->How often did you notice it changing?
1 EVERY DAY
2

(CIRCLE ONE)

ALMOST EVERY DAY

3 ONLY ON A FEW DAYS
---->What made you aware of the water level change?

A15. Do you think that daily fluctuations in the water level would make
you feel more or less like you were in a natural setting? (CIRCLE
ONE NUMBER)
1 MUCH MlRE LIKE A NATURAL SETTING
2 SOMEWHAT K:>RE LIKE A NATURAL SETTING
3 NATURAL SETTING REGARDLESS OF FLUCTUATIONS
4 SOMEWHAT LESS LIKE A NATURAL SETTING
5 MJ CH LESS LIKE A NATURAL SETTING
6 DON'T KNOW

•

D.7
page 5

•

A16. If you had a choice, would you have preferred a trip with daily
fluctuations in the water level or one with a constant water level?
(CIRCLE ONE NUMBER)
1 I WOULD PREFER A TRIP WITH DAILY FLUCTUATIONS
2 I WOULD PREFER A TRIP WITH CONSTANT WATER LEVELS
3 MAKES NO DIFFERENCE TO ME

A17. On your trip, did you feel you had enough time to hike the side
canyons and see other attractions? (CIRCLE ONE NUMBER)
1 YES, THERE WAS ENOUGH TIME FOR HIKING
2 NO, THERE WAS NOT ENOUGH TIME FOR HIKING
3 THERE WAS TOO MUCH TIME FOR HIKING

A18. Did you ever have to share the beach where you were camping with
other groups during your trip? (CIRCLE ONE NUMBER)
1 NO
2 YES---->How many nights did this happen?
ONE NIGHT
2 TWO NIGHTS
3 THREE NIGHTS
4 FOUR OR MJRE NIGHTS

&

(CIRCLE ONE NUMBER)

D.8
page 6
A19. Could you see the camps of other groups from any of your campsites
during your last trip? (CIRCLE ONE NUMBER)
1 NO
2 YES---->Were these groups sharing the beach with your group
or did they have a separate beach? (CIRCLE ONE)
1 WE SHARED THE BEACH
2 THEY WERE ON A SEPARATE BEACH

A20.

If you had a choice, would you prefer a campsite: (CIRCLE ONE)
1 ON THE SAME BEACH AS ANOTHER PARTY
2 WHERE YOU MIGHT BE ABLE TO SEE OR HEAR ANOTHER PARTY
3 OUT OF SIGHT AND HEARING OF OTHERS

In this next section we would like to find out how you traveled to the
Grand Canyon and what types of items you purchased for your white
water trip. This information will help us to compare your responses
with those of other people.

A21. How would you best describe you reason(s) for taking your Grand
Canyon white water boat trip? (CIRCLE ONE NUMBER)
1 THE WHITE WATER BOAT TRIP WAS THE ONLY REASON FOR MAKING
THE TRIP
2 THE WHITE WATER BOAT TRIP WAS THE K)ST IMPORTANT REASON FOR
MAKING THE TRIP
3

4

THE WHITE WATER BOAT TRIP WAS ONE OF SEVERAL EQUALLY
IMPORTANT REASONS FOR TAKING THE TRIP
THE WHITE WATER BOAT TRIP WAS NOT AN IMPORTANT REASON FOR
MAKING THE TRIP

•
•

D.9
page 7

•

A22. Was any part of your trip to the Grand Canyon by airplane?
(CIRCLE ONE NUMBER)
YES---->How much time did it take to fly ~ way?
_ _ TOTAL HOURS OF FLYING
2

NO

A23. Did you drive at least part of the way to the Grand Canyon for
your white water trip?
YES---->How much time did you spend driving ~ way?
_ _ TOTAL HOURS OF DRIVING
- - DAY(S) DRIVING
2 NO----->Skip to question A26, next page

A24. What type of vehicle did you use to get to the Grand Canyon?
(CIRCLE ONE NUMBER)
FULL SIZED AUTOMOBILE
2 INTERMEDIATE SIZED AUTOMOBILE
3 COMPAC'T AUTOMOBILE
4 SMALL TRUCK (Toyota, Chevy S10, Bronco II, etc.)
5 R.V., FULL SIZE TRUCK, VAN
6 OTHER, (please s p e c i f y ) - - - - - - - - - - - - - - A25. How many people travelled with you (in the same vehicle) to the
Grand Canyon?
MYSELF AND _ _ OTHER PEOPLE

D.10

.,
page 8
A26. Please estimate how much your trip cost (COSTS FOR YOU
INDIVIDUALLY, NOT OTHERS FOR WHOM YOU MIGHT HAVE PAID). Include
only money spent on items specifically for this trip. If a certain
item was not purchased for this trip, please put $0.
Payment to Rafting Company

$_ _ _ __

Airfare

$

Car Rental

$- - - - -

Gas and Oil for vehicle

$- - - - -

Food and Beverages (before and after
white water trip)

•

-----

$- - - - -

Lodging, Camping (before and after
white water trip)
Personal gear (suntan lotion, sun
glasses, film for camera)

$_ _ _ __

Other (please specify) - - - - - - $_ _ _ __

TOTAL AK>U:HT TRIP <X>ST (Please add all

payments and fill in the total on this
line)

$

-----

A27. Would you still have gone on the Grand Canyon white water trip if
your costs had been $
aore than the total you just calculated
in Question A26? (CIRCLE ONE NUMBER)
1 YES, I WOULD PAY THIS AM)UNT TO TAKE THE TRIP
2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

'

D. 11

PLEASE READ CAREFULLY
Many factors influence the quality of Grand Canyon white water
trips. For example, in a recent survey white water boaters told us that
things like good weather, good social interaction, good guides and trip
leaders, the number of layovers at attraction sites, running rapids,
good food, and many other things would contribute to an excellent or
perfect trip. The present survey, however, is focusing specifically on
those aspects of the trip that are affected by water flows in the
Colorado River.
Water flows in a river like the Colorado are often measured in cubic
feet per second (cfs) passing a given point. For our study, water flows
are being described in terms of four categories: low flows (5,000 cfs),
moderate flows (13,000 cfs), moderately high flows (22,000 cfs), and
high flows (40,000 cfs) as measured by releases at Glen Canyon Dam, the
last dam above the Grand Canyon. These flow levels are only a few of
the many alternative flows that are possible given legal restrictions on
releases from Glen Canyon Dam and they are being used here to find out
about your preference for various Colorado river flows through the Grand
Canyon.
The amount of water being released from Glen Canyon Dam can also
vary from time to time within any one day. These daily fluctuations,
when they occur, typically follow a regular pattern. Flow releases from
the Dam increase during the morning to provide high water during the
afternoon, and decrease in the late afternoon.and evening, resulting in
low water at night and in the early morning hours.

'

In the case descriptions that follow, we will describe the effects
of each of these types of flow patterns. For each type of flow we would
like you to tell us how it would affect the quality of a Grand Canyon
white water trip for you. A previous study of boating in the Grand
Canyon shows that white water boaters tend to give a high rating to
their trip regardless of the flow they actually experienced. However,
most boaters were able to indicate a preference for one type of flow
over others. Information from this previous survey is presented as an
aid in your evaluation of different river conditions and represents the
general opinion of boaters in our previous study. Your opinion about
water levels, however, may be different. For each type of condition, we
would like you to tell us how the river flow would affect the quality of
your white water trip.
Your white water trip in the Grand Canyon started on ~~~~­
Records show that during your trip the average water level was about
cfs, with daily changes ranging from an average daily low of
cfs to an average daily high flow of
cf s.

D.12
CASE 1

•

At a constant flow of 5,000 cfs, the speed of the river is relatively
slow, reducing time for side canyon visits and other attractions.
Boaters must break camp early to stay on schedule. Although rapids are
present at this low water level, the waves are smaller and do not produce
the big "roller coaster" ride created by higher flows. Due to exposed
rocks, some rapids may be so difficult that it is likely passengers would
have to walk around them. However, camping opportunities are abundant
with many large sandy beaches exposed.
B1. Do you think a Grand Canyon white water trip under the conditions

described for Case 1 above would be better or worse than your last
Grand Canyon white water trip? (CIRCLE ONE NUMBER)
MUCH BETTER
2 SOMEWHAT BETTER
3

ABOUT THE SAME

4

SOMEWHAT WORSE

5

MUCH WORSE

We would now like you to imagine that you are presently deciding whether
or not to go on a Grand Canyon white water trip. Imagine that the trip
would be the same as your last trip (e.g., the same people, same food,
etc.) with two exceptions:
The water level would be constant at 5,000 cfs (see Case 1 above)
AND

Your individual costs for the trip increased by $---- (over the
total cost you calculated on page 8, question A26)

B2. Would you go on this trip?

(CIRCLE ONE NUMBER)

YES, I WOULD PAY THIS AMOUNT TO TAKE THE TRIP
2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

•

D.13
CASE 2

With flows fluctuating daily from 1,000 to 17,000 cfs, around an
average daily flow of 5,000 cfs, most people are aware of changes in the
water level. Trip speed is relatively slow, reducing time for side
canyon visits, and boaters must break camp early to stay on schedule.
Large sandy beaches are generally abundant, but boatmen must take ca.re
selecting mooring sites. Occasionally, due to low water in the morning,
gear will have to be carried a long ways (perhaps across slippery rocks)
to be loaded on the boats. Boatmen may have to wait above certain rapids
for the water to rise, or hurry to get to a rapid before the water
falls. Due to exposed rocks, some rapids may be so difficult that it is
likely passengers would have to walk around them. At other rapids,
however, higher flows may produce large waves and a bigger "roller
coaster" ride than at a low constant flow.

B3. If you had to choose, which would you prefer: low water with small
or no fluctuations or low water with large daily fluctuations?
(CIRCLE ONE NUMBER)
1 LOW WATER WITH SMALL OR NO FLUCTUATIONS
2 LOW WATER WITH LARGE DAILY FLUCTUATIONS
3 MAKES NO DIFFERENCE TO ME

Now imagine that you are deciding whether or not to go on a Grand Canyon
white water trip. Imagine that the trip would be the same as your last
trip (e.g., the people, food, etc.) with two exceptions:
There would be large daily fluctuations from a low flow of 1,000 cfs
to a high flow of 17,000 cfs around an average of 5,000 cfs (see
description for Case 2 above)
AND

Your individual costs for the trip increased by $
tot al cost you calculated on page 8, question A26)

~~~-

•

B4. Would you go on this trip?

(CIRCLE ONE NUMBER)

1 YES, I WOULD PAY THIS AM:>UNT TO TAKE THE TRIP
2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

(over the

D.14
CASE 3

At moderate water levels (around 13,000 cfs), the pace of the river
is slightly faster than at low flows, leaving a little more time for
hiking in side canyons and stops at attractions. Most boating groups
will not have a problem staying on schedule. Rapids tend to have larger
waves and provide a little more of a "roller coaster" ride than at low
water. Passengers may have to walk around only a few rapids. Campsites
are still large and plentiful.
C1. Do you think a Grand Canyon white water trip under the conditions
described for Case 3 above would be better or worse than your last
Grand Canyon white water trip? (CIRCLE ONE NUMBER)

1 MU CH BETTER
2 SOMEWHAT BETTER
3

ABOUT THE SAME

4

SOMEWHAT WORSE

5

MUCH WORSE

We would now like you to imagine that you are presently deciding whether
or not to go on a Grand Canyon white water trip. Imagine that the trip
would be the same as your trip (e.g., the same people, same food, etc.)
with two exceptions:
The water level would be constant at 13,000 cfs (see description for
Case 3 above)
ABD

Your individual costs for the trip increased by $_ _ __ (over the
total cost you calculated on page 8, question A26)
C2. Would you go on this trip?

(CIRCLE ONE NUMBER)

1 YES, I WOULD PAY THIS AMJUNT TO TAKE THE TRIP
2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

•

D.15

•
CASE 4

•

At moderately high water levels (around 22,000 cfs), the pace of the
river is faster than at lower flows, leaving more time for side canyons
and stops at attractions. Boating groups do not have a problem staying
on schedule. Rapids have larger waves and provide a bigger "roller
coaster" ride than at moderate water. Only a few passengers choose to
walk around some of the bigger rapids for their safety. Some potential
campsites are under water in some areas of the canyon, but generally
campsites are plentiful al though a bit smaller in size.
D1. Do you think a Grand Canyon white water trip under these

conditions (Case 4 above) would be better or worse than your
last Grand Canyon white water trip? (CIRCLE ONE NUMBER)
MUCH BETTER
2

SOMEWHAT BETTER

3 ABOUT THE SAME
4 SOMEWHAT WORSE
5 MJCH WORSE

We would now like you to imagine that you are presently deciding whether
or not to go on a Grand Canyon white water trip. Imagine that the trip
would be the same as your last trip (e.g., the people, food, etc.) with
two exceptions:
The water level would be constant at 22,000 cfs (see description for
Case 4 above)
.AND

Your individual costs for the trip increased by $~~~­ (over the
total cost you calculated on page 8, question A26)

D2. Would you go on this trip?

(CIRCLE ONE NUMBER)

1 YES, I WOULD PAY THIS

A~DUNT

TO TAKE THE TRIP

2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

•

D.16
CASE 5

With large daily fluctuations from 10,000 cfs - 31,500 cfs,
around an average daily flow of 22,000 cfs, most people are aware of
water level changes. The boatmen will have to take more care in
selecting mooring and camping sites. Due to low water levels in the
morning, gear may have to be carried (perhaps across rocky areas) to
be loaded on the boats. Boatmen may decide to wait above certain
rapids for the water level to rise or may have to hurry to get to a
certain rapid before the water level falls. In addition, some rapids
may be difficult due to exposed rocks at low water levels and other
rapids might be quite large at high water levels, and it is likely
that passengers may have to walk around a few rapids. When the water
is high or rising, however, the standing waves in some of the major
rapids become larger, resulting in a bigger "roller coaster" ride.
D3. If you had to choose, which would you prefer: moderately high
water with small or no fluctuations or moderately high water with
large daily fluctuations? (CIRCLE ONE NUMBER)
1 MJDERATELY HIGH WATER WITH SMALL OR NO FLUCTUATIONS
2

MJDERATELY HIGH WATER WITH LARGE DAILY FLUCTUATIONS

3 MAKES NO DI FFE REN CE TO ME

Now imagine that you are deciding whether or not to go on a Grand
Canyon white water trip. Imagine that the trip would be the same as
your last trip (e.g., the people, food, etc.) with two exceptions:
There would be large daily fluctuations from a low flow of
10,000 cfs to a high flow of 31,500 cfs around an average of
22,000 cfs (see description for Case 5 above)
AND

Your individual costs for the trip increased by $---- (over
the total cost you calculated on page 8, question A26)
D4. Would you go on this trip?

(CIRCLE ONE NUMBER)

1 YES, I WOULD PAY THIS AMJUNT TO TAKE THE TRIP
2

NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

•
•

D. 17

"

•

CASE 6
At high water levels (around 40,000 cfs), the current is fast.
Trips are able to stop at additional side ·canyons and spend additional
time at attraction sites. Fewer rapids are present, as sane of the
smaller rapids are "washed out." In other rapids, however, the waves
are very large and some passengers, especially those on oar powered
trips, face an increased likelihood of having to walk around one or
more of the major rapids for their safety. Campsites become more
scarce as sandbars and shore areas are flooded, and campsites are much
smaller. In some areas of the Canyon, there is an increased chance of
camping with or near other groups.
E1. Do you think a Grand Canyon white water trip under the conditions

described above for Case 6 would be better or worse than your last
Grand Canyon white water trip?
MUCH BETTER
2 SOMEWHAT BETTER
3

ABOUT THE SAME

4

SOMEWHAT WORSE

5

MUCH WORSE

We would now like you to imagine that you are presently deciding whether
or not to go on a Grand Canyon white water trip. Imagine that the trip
would be the same as your last trip (e.g., the people, food, etc.) with
two exceptions:
The water level would be constant at 40,000 cfs (see Case 6 above)

Your individual costs for the trip increased by $---- (over the
total cost you calculated on page 8, question A26)

•

E2. Would you go on this trip?

(CIRCLE ONE NUMBER)

YES, I WOULD PAY THIS AM:lUNT TO TAKE THE TRIP

•

2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

D.18
CASE 7

There are indications that certain types of flow patterns in the
long run may reduce the number of sandy beaches in the Grand Canyon.
At present, the area between Hance Rapids and Havasu has fewer beaches
than other parts of the canyon. Trip leaders must plan schedules very
closely to ensure a good campsite in this area. As beaches disappear,
this careful planning would have to be extended to other parts of the
canyon.
This planning might mean missing some attraction sites to get to
camp early or longer stops at some attraction sites. Fewer beaches
would increase the likelihood of camping near other i:arties and
perhaps sharing a beach with other parties. Some camps might have to
be made in areas without any sand.

Fl. If the n1.111ber of beaches in the Grand Canyon were substantially

reduced, the effects described above would become much more
likely. We would like you to imagine that you are presently
deciding whether or not to go on a Grand Canyon white water
trip. All of the details of this trip would be the same as your
last trip with two exceptions:
The number of beaches was substantially reduced (see Case 7
above)

AHD

____ (over

Your individual costs for the trip increased by $
the total cost you calculated on page 8, question A26)
,

F2. Would you go on this trip?

(CIRCLE ONE NUMBER)

YES, I WOULD PAY THIS AMOUNT TO TAKE THE TRIP
2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE· THE TRIP

•

D.19

•
In the previous case descriptions there were a number of different
questions where we asked whether you would still take the white water
trip if your expenses increased by a certain dollar amount.
1t

To help us better understand your responses, we would like to know the
extent to which various factors affected your anSoJers. Please tell us
whether the following statements were true or not for you when answering
those questions. (CIRCLE ONE RESPONSE FOR EACH STATEMENT)
Definitely Probably Probably Definitely
True
True
False
False
G1. My main concern was that the

Park Service might start charging
a fee for private Grand Canyon
white water trips.

1

2

3

4

G2. My main concern was that rafting
companies might increase the price
of commercial Grand Canyon white
water trips.

1

2

3

4

G3. My responses reflected my best
guess as to whether I would pay
the specified amounts for the
Grand Canyon white water trip.

1

2

3

4

2

3

4

2

3

4

2

3

4

G4. My responses reflected the fact
that I feel the Grand Canyon white
water boating experience should be
maintained for all people to enjoy.
G5. I just don't know how much more I

would really pay for a Grand Canyon
white water trip regardless of the
conditions.

G6. I just don't want to have to pay
more for a Grand Canyon white water
trip regardless of what the
conditions would be.
1
G7. Other, please describe:

D.20

In this final section, we would like to ask some questions about your
background which will help us compare your answers with those of other
people. We stress that all of your answers are strictly confidential.

•

B1. How old are you? ___ YEARS OLD
B2. Are you:

MALE
2

FEMALE

B3. How many years of school have you completed? (CIRCLE OR CHECK THE
HIGHEST YEAR OR LEVEL)
1

2

3

4

5

6

7

8

9

10

11

12

_ _;Some college

_ _ M.A., M. S.

_ _B. A. or equivalent

_ _Advanced degree (J.D., M.D.,
Ph.D)

B4. Please circle the response that comes closest to your total family
income before taxes.

(CIRCLE ONE NUMBER)

1 Less than $10,000
2 $10,000 to $14,999
3 $15,000 to $19,999
4 $20,000 to $24,999
5 $25,000 to $29,999
6 $30,000 to $34,999
7 $35,000 to $39 ,999
8 $40,000 to $44,999

9
10
11
12
13
14
15

$45,000 to $49,999
$50,000 to $59,999
$60,000 to $69,999
$70,000 to $79,999
$80,000 to $89,999
$90,000 to $99,999
$100,000 or more

.i

D.21
1

B5. With reference to your primary occupation, are you currently
(CIRCLE ONE NUMBER):

1

EMPLOYED FULL-TIME

2

EMPLOYED PART-TIME

3 FULL-TIME HOMEMAKER
4 TEMPORARILY UNEMPLOYED
5

NOT EMPLOYED, NOT LOOKING FOR WORK

6

RETIRED, NOT WORKING

7

RETIRED, WORKING PART-TIME

_ _ Check here if you would like a copy of results

THAR YOU FOR YOUR HELP!

Please return this survey in the enclosed envelope to:
HBRS

4513 Vernon Boulevard
Madison, WI 53705

E.1

APPENDIX E

t
WHITE-WATER PRIVATE BOATERS CORTIBGENT-VALUATIOH SURVEY

,.

..

E.2

This questionnaire refers to the white water trip you took in the
Grand Canyon that started on
refer to this trip when answering the questions in this survey.

Please
It is

important that this survey be completed by the person to whom it was
sent.

..

E.3
page 1

t

In this first section, we are interested in finding out about your
white water trip in the Grand Canyon and how much you enjoyed it.
A1. Overall, how would you rate your white water trip? (CIRCLE ONE NUMBER)
1
2
3
4

POOR
FAIR, it just didn't work out ver-y well
GOOD, but a number of things could have been different
VERY GOOD, but could have been better .

5 EXCELLENT, only minor problems
6 PERFECT
A2. Where did you put-in (start trip)?
1 LEE'S FERRY
2 PHANTOM RANCH
3

OTHER ( s p e c i f y - - - - - - - - - - -

A3. Where did you take-out (end trip)?
1 PHANTOM RANCH
2 WHITMORE WASH
3 DIAMOND CREEK
4 LAKE MEAD
5

(CIRCLE ONE NUMBER)

(CIRCLE ONE

NU~ffiER)

OTHER (specify

A4. How long was your trip? _ _ _ _ _ DAYS
A5. What type of boat were you on? (CIRCLE ONE NUMBER)
MOTOR POWERED RAFT
2 OAR POWERED RAFT
3 COMBINATION MOTOR/OAR RAFT
4

DORY
5 KAYAK
6 PADDLE RAFT
7 OTHER (specify

E.4
page 2

A6. Was your Grand Canyon white water trip:
1

RUN BY A OOMMERCIAL OUTFITTER

2

A

(CIRCLE ONE NUMBER)

I

PRIVATE TRIP-->Were you primarily responsible for operating
a boat on this trip?
YES
2

NO

A7. How many times have you taken a white water trip on the Colorado
River below Lee's Ferry, including this trip?
-----TIMES

A8. If you had the opportunity, would you take a Grand Canyon white
water trip again? (CIRCLE ONE NUMBER)
1 DEFINITELY NOT
2

PROBABLY NOT

3 PROBABLY YES
4 DEFINITELY YES

A9. On average, how crowded did you feel the river was while you were
floating? (Circle the number on the scale best representing your
feelings.)
1

2

not at all
crowded

3

slightly
crowded

4

5

6

moderately
crowded

7

8

9

extremely
crowded

"

.

E.5
page 3

l

Rapids are an important part of the Grand Canyon trip for many people.
In this next section, we would like to get your expectations and
feelings about the rapids.

A10. What role did rapids play in your decision to take this trip?
(CIRCLE ONE NUMBER)
1 RAPIDS WERE THE M)ST IMPORTANT REASON FOR TAKING THE TRIP
2 RAPIDS WERE ONE OF THE TWO OR THREE MOST IMPORTANT REASONS FOR
TAKING THE TRIP
3 RAPIDS WERE ONLY ONE OF MANY IMPORTANT REASONS FOR TAKING THE
TRIP
4 RAPIDS WERE NOT AN IMPORTANT REASON FOR TAKING THE TRIP

A11. Did you have to walk around any rapids?
1 NO
2 YES---->Which rapids?_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

A12. In general, which type of rapid did you enjoy most on this trip:
(CHOOSE ONE )
1 BIG RAPIDS
2 MEDIUM RAPIDS
3 SMALL RAPIDS
4 LIKED ALL TYPES OF RAPIDS EQUALLY
5 DON'T LIKE RAPIDS

.

E.6

•
page 4
Besides rapids, the water level on the river may also affect a person's
trip. In this next section, we are interested in your feelings about the
water level during your trip.

J

A13. If you had the choice, would you have preferred the overall water
1 evel to be:
(CIRCLE ONE NUMBER)
1

2

LOWER
ABOUT THE SAME

3 HIGHER
4 DON'T KNOW OR DOESN'T MATTER

A14.

Did you notice whether the water level changed during your trip?
NO
2 YES---->How often did you notice it changing? (CIRCLE ONE)
EVERY DAY
2 ALMJST EVERY DAY
3 ONLY ON A FEW DAYS
---->What made you aware of the water level change?
1

A15. Do you think that daily fluctuations in the water level would make
you feel more or less like you were in a natural setting? (CIRCLE
ONE NUMBER)
1 MU CH M) RE LIKE A NATURAL SETTING
2 SOMEWHAT MORE LIKE A NATURAL SETTING
3 NATURAL SETTING REGARDLESS OF FLUCTUATIONS
4 SOMEWHAT LESS LIKE A NATURAL SETTING
5 MUCH LESS LIKE A NATURAL SETTING
6 DON'T KNOW

•

E.7

..

page 5

I

A16. If you had a choice, would you have preferred a trip with daily
fluctuations in the water level or one with a constant water level?
(CIRCLE ONE NUMBER)
1 I WOULD PREFER A TRIP WITH DAILY FLUCTUATIONS
2 I WOULD PREFER A TRIP WITH CONSTANT WATER LEVELS
3 MAKES NO DIFFERENCE TO ME

A17. On your trip, did you feel you had enough time to hike the side
canyons and see other attractions? (CIRCLE ONE NUMBER)
1 YES, THERE WAS ENOUGH TIME FDR HIKING
2 NO, THERE WAS NOT ENOUGH TIME FDR HIKING
3 THERE WAS TOO MUCH TIME FOR HIKING
:·

\

A18. Did you ever have to share the beach where you were camping with
other groups during your trip? (CIRCLE ONE NUMBER)
NO
2 YES---->How many nights did this happen? (CIRCLE ONE NUMBER)
1 ONE NIGHT
2 TWO NIGHTS
3 THREE NIGHTS
4 FOUR OR M:>RE NIGHTS

E.8

..
page 6
A19. Could you see the camps of other groups from any of your campsites
during your last trip? (CIRCLE ONE NUMBER)
1 NO
2 YES---->Were these groups sharing the beach with your group
or did they have a separate beach? (CIRCLE ONE)
1 WE SHARED THE BEACH
2 THEY WERE ON A SEPARATE BEACH

A20.

If you had a choice, would you prefer a campsite: (CIRCLE ONE)
1 ON THE SAME BEACH AS ANOTHER PARTY
2 WHERE YOU MIGHT BE ABLE TO SEE OR HEAR ANOTHER PARTY
3 OUT OF SIGHT AND HEARING OF OTHERS

In this next section we would like to find out how you traveled to the
Grand canyon and what types of items you purchased for your white
water trip. This information will help us to compare your responses
with those of other people.

A21. How would you best describe you reason(s) for taking your Grand
Canyon white water boat trip? (CIRCLE ONE NUMBER)
1 THE WHITE WATER BOAT TRIP WAS THE ONLY REASON FOR l-iAKING
THE TRIP
2 THE WHITE WATER BOAT TRIP WAS THE MJST IMPORTANT REASON FOR
MAKING THE TRIP
3 THE WHITE WATER BOAT TRIP WAS ONE OF SEVERAL EQUALLY
IMPORTANT REASONS FOR TAKING THE TRIP
4 THE WHITE WATER BOAT TRIP WAS NOT AN IMPORTANT REASON FOR
MAKING THE TRIP

E.9
page 7

l

A22. Was any part of your trip to the Grand Canyon by airplane?
(CIRCLE ONE NUMBER)
1 YES----> How much time did it take to fly one way?
--- TOTAL HOURS OF FLYING
2

NO

A23. Did you drive at least part of the way to the Grand Canyon for
your white water trip?
1 YES---->How much time did you spend driving one way?
_ _ DAY ( S) DRIVING
- - TOTAL HOURS OF DRIVING
2 NO----->Skip to question A26, ne:xt page

A24. What type of vehicle did you use to get to the Grand Canyon?
(CIRCLE ONE NUMBER)
1 FULL SIZED AUTOMOBILE
2

INTERMEDIATE SIZED AUTOMOBILE

3

COMPACT AUTOMOBILE

4

SMALL TRUCK (Toyota, Chevy S10, Bronco II, etc.)

5

R.V., FULL SIZE TRUCK, VAN

6 OTHER, (please specify) --------------~
A25. How many people travelled with you (in the same vehicle) to the
Grand Canyon?
MYSELF AND _ _ OTHER PEOPLE

E.10
11

page 8
A26. Please estimate how much your trip cost (COSTS FOR YOU
INDIVIDUALLY, EITHER PAID BY YOURSELF OR BY OTHERS). Include only
money spent on items specifically for this trip. If a certain item
was not purchased for this trip, please put $0.
Gas and Oil for vehicle

$- - - - -

Airfare

$"_ _ _ __

Car Rental

$"_ _ _ __

Food and Beverages

$"_ _ _ __

Personal gear (suntan lotion, sun
glasses, film for camera)

$_ _ _ __

Lodging, Camping (before and after
white water trip)

$- - - - -

Boat Gear (oars, lines, etc.)

$_ _ _ __

Equipment rental

$_ _ _ __

Take out at Diamond Creek

$

Vehicle shuttle
Tow across Lake Mead

----$----$-----

Other (please specify) ~------$_ _ _ __

TOTAL .AK>URT TRIP OOST (Please add all
payments and fill in the total on this
line)

$

-----

A27. Would you still have gone on the Grand Canyon white water trip
if your costs had been $
•ore than the total you just
calculated in Question A26? (CIRCLE ONE NUMBER)
1 YES, I WOULD PAY THIS AMOUNT TO TAKE THE TRIP
2

NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

t

E. 11

.
PLEASE READ CAREFULLY

Y.ia.ny factors influence the quality of Grand Canyon white water
trips. For example, in a recent survey white water boaters told us that
things like good weather, good social interaction, good guides and trip
leaders, the number of layovers at attraction sites, running rapids,
good food, and many other things would contribute to an excellent or
perfect trip. The present survey, however, is focusing specifically on
those aspects of the trip that are affected by water flows in the
Colorado River.
Water flows in a river like the Colorado are often measured in cubic
feet per second (cfs) passing a given point. For our study, water flows
are being described in terms of four categories: low flows (5,000 cfs),
moderate flows (13,000 cfs), moderately high flows (22,000 cfs), and
high flows (40,000 cfs) as measured by releases at Glen Canyon Dam, the
last dam above the Grand Canyon. These flow levels are only a few of
the many alternative flows that are possible given legal restrictions on
releases from Glen Canyon Dam and they are being used here to find out
about your preference for various Colorado river flows through the Grand
Canyon.
The amount of water being released from Glen Canyon Dam can also
vary from time to time within any one day. These daily fluctuations,
when they occur, typically follow a regular pattern. Flow releases from
the Dam increase during the morning to provide high water during the
afternoon, and decrease in the late afternoon and evening, resulting in
low water at night and in the early morning hours.
In the case descriptions that follow, we will describe the effects
of each of these types of flow patterns. For each type of flow we would
like you to tell us how it would affect the quality of a Grand Canyon
white water trip for you. A previous study of boating in the Grand
Canyon shows that white water boaters tend to give a high rating to
their trip regardless of the flow they actually experienced. However,
most boaters were able to indicate a preference for one type of flow
over others. Information from this previous survey is presented as an
aid in your evaluation of different river conditions and represents the
general opinion of boaters in our previous study. Your opinion about
water levels, however, may be different. For each type of condition, we
would like you to tell us how the river flow would affect the quality of
your white water trip.
Your white water trip in the Grand Canyon started on ~~~~­
Records show that during your trip the average water level was about
~~~~~ cfs, with daily changes ranging from an average daily low of
~~~~~ cfs to an average daily high flow of
cfs.

E. 12

..
CASE 1

At a constant flow of 5,000 cfs, the speed of the river is relatively
slow, reducing time for side canyon visits and other attractions.
Boaters must break camp early to stay on schedule. Although rapids are
present at this low water level, the waves are smaller and do not produce
the big "roller coaster" ride created by higher flows. Due to exposed
rocks, some rapids may be so difficult that it is likely passengers would
have to walk around them. However, camping opportunities are abundant
with many large sandy beaches exposed.

l

B1. Do you think a Grand Canyon white water trip under the conditions

described for Case 1 above would be better or worse than your last
Grand Canyon white water trip? (CIRCLE ONE NUMBER)
MUCH BETTER
2 SOMEWHAT BETTER
3

ABOUT THE SAME

4

SOMEWHAT WORSE

5

MUCH WORSE

We would now like you to imagine that you are presently deciding whether
or not to go on a Grand Canyon white water trip. Imagine that the trip
would be the same as your last trip (e.g., the same people, same food,
etc.) with two exceptions:
The water level would be constant at 5,000 cfs (see case 1 above)
AND

Your individual costs for the trip increased by $~~~­ (over the
total cost you calculated on page 8, question A26)

B2. Would you go on this trip?

(CIRCLE ONE NUMBER)

1 YES, I WOULD PAY THIS Al-DUNT TO TAKE THE TRIP
2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

•

E. 13

CASE 2

I

With flows fluctuating daily from 1,000 to 17,000 cfs, around an
average daily flow of 5,000 cfs, most people are aware of changes in the
water level. Trip speed is relatively slow, reducing time for side
canyon visits, and boaters must break camp early to stay on schedule.
Large sandy beaches are generally abundant, but boatmen must take care
selecting mooring sites. Occasionally, due to low water in the morning,
gear will have to be carried a long ways (perhaps across slippery rocks)
to be loaded on the boats. Boatmen may have to wait above certain rapids
for the water to rise, or hurry to get to a rapid before the water
falls. Due to exposed rocks, some rapids may be so difficult that it is
likely passengers would have to walk around them. At other rapids,
however, higher flows may produce large waves and a bigger "roller
coaster" ride than at a low constant flow.

B3. If you had to choose, which would you prefer:

low water with small
or no fluctuations or low water with large daily fluctuations?
(CIRCLE ONE NUMBER)
1 LOW WATER WITH SMALL OR NO FLUCTUATIONS
2 LOW WATER WITH LARGE DAILY FLUCTUATIONS
3 MAKES NO DIFFERENCE TO ME

Now imagine that you are deciding whether or not to go on a Grand Canyon
white water trip. Imagine that the trip would be the same as your last
trip (e.g., the people, food, etc.) with two exceptions:
There would be large daily fluctuations from a low flew of 1,000 cfs
to a high flow of 17,000 cfs around an average of 5,000 cfs (see
description for Case 2 above)
AND

Your individual costs for the trip increased by $---- (over the
total cost you calculated on page 8, question A26)

B4. Would you go on this trip?

(CIRCLE ONE NUMBER)

YES, I WOULD PAY THIS AMOUNT TO TAKE THE TRIP
2

NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

E. 14

•
CASE 3

At moderate water levels (around 13,000 cfs), the pace of the river
is slightly faster than at low flows, leaving a little more time for
hiking in side canyons and stops at attractions. Most boating groups
will not have a problem staying on schedule. Rapids tend to have larger
waves and provide a little more of a "roller coaster" ride than at low
water. Passengers may have to walk around only a few rapids. Campsites
are still large and plentiful.
C1. Do you think a Grand Canyon white water trip under the conditions
described for Case 3 above would be better or worse than your last
Grand Canyon white water trip? (CIRCLE ONE NUMBER)
1 MU CH BETTER
2 SOMEWHAT BETTER

3 ABOUT THE SAME
4

SOMEWHAT WORSE

5

MUCH WORSE

We would now like you to imagine that you are presently deciding whether
or not to go on a Grand Canyon white water trip. Imagine that the trip
would be the same as your trip (e.g., the same people, same food, etc.)
with two exceptions:
The water level would be constant at 13,000 cfs (see description for
Case 3 above)
AND

Your individual costs for the trip increased by $---- (over the
total cost you calculated on page 8, question A26)

C2. Would you go on this trip?

(CIRCLE ONE NUMBER)

1 YES, I WOULD PAY THIS AMOUNT TO TAKE THE TRIP
2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

i

E.15

..
CASE 4

At moderately high water levels (around 22,000 cfs), the pace of the
river is faster than at lower flows, leaving more time for side canyons
and stops at attractions. Boating groups do not have a problem staying
on schedule. Rapids have larger waves and provide a bigger "roller
coaster" ride than at moderate water. Only a few passengers choose to
walk around some of the bigger rapids for their safety. Some potential
campsites are under water in some areas of the canyon, but generally
campsites are plentiful al though a bit smaller in size.
D1. Do you think a Grand Canyon white water trip under these
conditions (Case 4 above) would be better or worse than your

last Grand Canyon white water trip?

(CIRCLE ONE NUMBER)

1 MUCH BETTER
2 SOMEWHAT BETTER
4

ABOUT THE SAME
SOMEWHAT WORSE

5

MUCH WORSE

3

We would now like you to imagine that you are presently deciding whether
or not to go on a Grand Canyon white water trip. Imagine that the trip
would be the same as your last trip (e.g., the people, food, etc.) with
two exceptions:
The water level would be constant at 22,000 cfs (see description for
Case 4 above)
AND

Your individual costs for the trip increased by $'--~~­ (over the
total cost you calculated on page 8, question A26)

"

D2. Would you go on this trip?

(CIRCLE ONE

NU~IBER)

YES, I WOULD PAY THIS AM:>UNT TO TAKE THE TRIP
2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

E.16

CASE 5
With large daily fluctuations from 10,000 cfs - 31,500 cfs,
around an average daily flow of 22,000 cfs, most people are aware of
water level changes. The boatmen will have to take more care in
selecting mooring and camping sites. Due to low water levels in the
morning, gear may have to be carried (perhaps across rocky areas) to
be loaded on the boats. Boatmen may decide to wait above certain
rapids for the water level to rise or may have to hurry to get to a
certain rapid before the water level falls. In addition, some rapids
may be difficult due to exposed rocks at low water levels and other
rapids might be quite large at high water levels, and it is likely
that passengers may have to walk around a few rapids. When the water
is high or rising, however, the standing waves in some of the major
rapids become larger, resulting in a bigger "roller coaster" ride.

I

D3. If you had to choose, which would you prefer: moderately high
water with small or no fluctuations or moderately high water with
large daily fluctuations? (CIRCLE ONE NUMBER)
1 MODERATELY HIGH WATER WITH Sl-"i.ALL OR NO FLUCTUATIONS
2

MODERATELY HIGH WATER WITH LARGE DAILY FLUCTUATIONS

3 MAKES NO DIFFERENCE TO ME
Now imagine that you are deciding whether or not to go on a Grand
Canyon white water trip. Imagine that the trip would be the same as
your last trip (e.g., the people, food, etc.) with two exceptions:
There would be large daily fluctuations from a low flow of
10,000 cfs to a high flow of 31,500 cfs around an average of
22,000 cfs (see description for Case 5 above)
AND

Your individual costs for the trip increased by $
the total cost you calculated on page 8, question A26)

---=~-

DJJ. Would you go on this trip?

(CIRCLE ONE NUMBER)

1 YES, I WOULD PAY THIS AMOUNT TO TAKE THE TRIP
2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

(over

•

E. 17

.
CASE 6

At high water levels (around 40,000 cfs), the current is fast.
Trips are able to stop at additional side canyons and spend additional
time at attraction sites. Fewer rapids are present, as some of the
smaller rapids are "washed out." In other rapids, however, the waves
are very large and some passengers, especially those on oar powered
trips, face an increased likelihood of having to walk around one or more
of the major rapids for their safety. Campsites become more scarce as
sandbars and shore areas are flooded, and campsites are much smaller.
In some areas of the Canyon, there is an increased chance of camping
with or near other groups.
E1. Do you think a Grand Canyon white water trip under the conditions

described above for Case 6 would be better or worse than your last
Grand Canyon white water trip?
1 MU CH BETTER
2 SOMEWHAT BETTER
3

ABOUT THE SAME

4

SOMEWHAT WORSE

5

MUCH WORSE

We would now like you to imagine that you are presently deciding whether
or not to go on a Grand Canyon white water trip. Imagine that the trip
would be the same as your last trip (e.g., the people, food, etc.) with
two exceptions:
The water level would be constant at 40,000 cfs (see Case 6 above)
AND

Your individual costs for the trip increased by $~~~­ (over the
total cost you calculated on page 8, question A26)

E2. Would you go on this trip? (CIRCLE ONE NUMBER)
1 YES, I WOULD PAY THIS AMOUNT TO TAKE THE TRIP
2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

E.18

..
CASE 7

There are indications that certain types of flow patterns in the
long run may reduce the number of sandy beaches in the Grand Canyon. At
present, the area between Hance Rapids and Havasu has fewer beaches than
other parts of the canyon. Trip leaders must plan schedules very
closely to ensure a good campsite in this area. As beaches disappear,
this careful planning would have to be extended to other parts of the
canyon.
This planning might mean missing some attraction sites to get to
camp early or longer stops at some attraction sites. Fewer beaches
would increase the likelihood of camping near other parties and perhaps
sharing a beach with other parties. Some camps might have to be made in
areas without any sand.

F1. If the number of beaches in the Grand Canyon were substantially

reduced, the effects described above would become much more
likely. We would like you to imagine that you are presently
deciding whether or not to go on a Grand Canyon white water
trip. All of the details of this trip would be the same as your
last trip with two exceptions:
The number of beaches was substantially reduced (see Case 7
above)
ARD

Your individual costs for the trip increased by $'--~~­ (over
the total cost you calculated on page 8, question A26)

F2. Would you go on this trip?

(CIRCLE ONE NUMBER)

YES, I WOULD PAY THIS AMOUNT TO TAKE THE TRIP
2 NO, I WOULD NOT PAY THIS AMOUNT TO TAKE THE TRIP

E.19

•
In the previous case descriptions there were a number of different
questions where we asked whether you would still take the white water
trip if your expenses increased by a certain dollar amount.
To help us better understand your responses, we would like to know the
extent to which various factors affected your answers. Please tell us
whether the following statements were true or not for you when answering
those questions. (CIRCLE ONE RESPONSE FOR EACH STATEMENT)
Definitely Probably Probably Definitely
True
True
False
False

"

G1. My main concern was that the
Park Service might start charging
a fee for private Grand Canyon
white water trips.

1

2

3

4

G2. My main concern was that rafting
companies might increase the price
of commercial Grand Canyon white
water trips.

1

2

3

4

G3. My responses reflected my best
guess as to whether I would pay
the specified amounts for the
Grand Canyon white water trip.

2

3

4

G4. My responses reflected the fact
that I feel the Grand Canyon white
water boating experience should be
maintained for all people to enjoy.

2

3

4

GS. I just don't know how much more I
would really pay for a Grand Canyon
white water trip regardless of the
conditions.

2

3

4

G6. I just don't want to have to pay
more for a Grand Canyon white water
trip regardless of what the
conditions would be.
1

2

3

4

G7. Other, please describe:

E.20

•
In this section we would like to know how you evaluate the chance of a
boat flipping in specific rapids at specific flow levels.

,
H1. It has been suggested that the flow level in the river might
affect the likelihood of boating accidents in the rapids. We
would like your judgment as to the risk of flipping a boat (the
type of boat you used for your 1985 Grand Canyon white water trip)
in two specific rapids, under four different flow conditions.
Imagine that 100 boats of your type were run, by "average"
boaters, through Crystal rapid at flows in the range of 3,000 to
9,000 cfs. How many of these boats do you think would flip? You
would answer "zero" if you felt that none of these boats would
flip in Crystal at these flows, and "100" if you feel that all of
these boats would flip at these flows. Answering "50", for
example, would indicate you believe that roughly half of the boats
like yours would flip in Crystal at these flows. Please give us
your best judgment (between 0 and 100) of the chance of flipping
in these two rapids at the four flow conditions shown below.
(PLEASE FILL IN THE BLANK FOR EACH CATEGORY)

BANCE RAPID:

boats out of every 100 would flip at daily flows
between 3,000 and 9,000 cf's
boats out of every 100 would flip at daily flows
between 10,000 and 15,000 cf's
boats out of every 100 would flip at daily flows
between 16,000 and 31,000 cf's
boats out of every 100 would flip at daily flows
greater than 32,000 cf's

CRYSTAL RAPID:

boats out of every 100 would flip at daily flows
between 3,000 and 9,000 cf's
boats out of every 100 would flip at daily flows
between 10,000 and 15,000 cf's
boats out of every 100 would flip at daily flows
between 16,000 and 31,000 cf's
boats out of every 100 would flip at daily flows
greater than 32,000 cf's

•

E.21

.
H2. For question H1 you filled in eight nunbers describing how many
boats like yours might flip under various conditions. We would
now like to know, in your judgment, if any of the nunbers you
reported in question H1 are so high that you would have serious
concerns about running the rapid(s) under those conditions.
Please indicate whether the chances you reported of flipping are
acceptable or unacceptable. (CIRCLE ONE NUMBER FOR EACH
CATEGORY).
The chances I reported of flipping in llAHCE RAPID are:
Acceptable

Unacceptable

1

2

at a daily flow of 3,000 to 9,000
cf's

1

2

at daily flows between 10,000 and
15,000 cf's

1

2

at daily flows between 16,000 and
31,000 cf's

1

2

at daily flows greater than 32,000
cf's

The chances I reported of flipping in CRYSTAL RAPID are:
Acceptable

Unacceptable

1

2

at a daily flow of 3,000 to 9,000
cf's

2

at daily flows between 10,000 and
15,000 cf's

2

at daily flows between 16,000 and
31,000 cf's

II

1

2

at daily flows greater than 32,000
cf's

E.22

In this final section, we would like to ask some questions about your
background which will help us compare your answers with those of other
people. We stress that all of your answers are strictly confidential.

I1. How old are you? ___ YEARS OLD
I2. Are you:

1

MALE

2

FEMALE

13. How many years of school have you completed? (CIRCLE OR CHECK THE
HIGHEST YEAR OR LEVEL)
1

2

3

4

5

6

---'Some college
_ _B. A. or equivalent

7

8

9

10

11

12

_ _ M.A., M.S.
---'Advanced degree (J. D., M. D.,
Ph. D)

I4. Please circle the response that comes closest to your total family
income before taxes.

(CIRCLE ONE NUMBER)

1 Less than $10,000
2
3
4
5
6
7
8

$10,000 to $14,999
$15,000 to $19,999
$20,000 to $24,999
$25,000 to $29,999
$30,000 to $34,999
$35,000 to $39,999
$40,000 to $44,999

9
10

$45, 000 to $49,999
$50,000 to $59,999

11
12

$60,000 to $69,999
$70,000 to $79,999

13
14

$80,000 to $89,999
$90,000 to $99,999

15

$100,000 or more

E.23

15. With reference to your primary occupation, are you currently
(CIRCLE ONE NUMBER):

1

EMPLOYED FULL-TIME

2

EMPLOYED PART-TIME

3
4

FULL-TIME HOMEMAKER
TEMPORARILY UNEMPLOYED

5

NOT EMPLOYED, NOT LOOKING FOR WORK

6 RETIRED, NOT WORKING
7

RETIRED, WORKING PART-TIME

_ _ Check here i f you would like a copy of results

THAllK YOU FOR ?OUR HELP!

Please return this survey in the enclosed envelope to:
l

HBRS

4513 Vernon Boulevard
Madison, WI 53705

F. 1

•

APPENDIX F
GLD CA.HYOR ARGLERS OR-SITE An'RmUTE SURVEY

•

F.2
Lee·s Ferry Fishing Survey
l) overall, how was your fishing trip today?
POOR
FAIR, it just didn't work out very well
GOOD, but I wish a number of things
could have been different
VERY GOOD, but could have been better
EXCELLENT, only minor problems
PERFECT
2l

How important were each of the following
reasons in your decision to come to Lee's
Ferry to fish rather than going elsewhere?
(CIRCLE ONE NUMBER FOR EACH REASON)

Thought I would

Not
Somewhat
Very
Important Important Important

catch a lot of fish

10)

How important would each of the following be
in contributing to a ~ fishing trip at
Lee's Ferry for you
Not
Somewhat
Very
Important Important Important

Not catching your
limit
Not catching a
trophy fish
Catching no fish
Poor weather
High water level

Low water level

Thought I would
catch big fish

Rising water level

Wanted to fish
in Glen Canyon

Falling water level
Seeing many others

Close to home

2

Not being able to get
upstream to fish

Few other trout
areas in Arizona

3 l

'

Boat/Motor trouble due
to the water level
l

How many fish did :t.2!!.• personally, keep
today?
_ _ _FISH

Not being able to camp
alony thEI river
l

What was the biggest fish :t.2!!_ kept? _ _ _ lbs.
4 )

What do ~consider to be a •trophy size•
rainbow trout? _ _ _ lbs.

S ) Counting this year, how many years have you
been fishing at Lee's Ferry? _ _ _YEARS
6)

11) Have you ever fiched at Lee's Ferry under
the following conditions?
Don't
Know
Medium water (9,000-16,000 CFS)
Low water (9,000 CFS or less)
Fluctuating water levels

How many days have you fished at Lee's Ferry

in 1984?

_ _ _DAYS

7 ) Have you camped upstream along the river
during 1984?
NO

12) Did you know today's expected water level
before your trip?
NO, DIDN'T TRY TO FIND OUT
NO, TRIED TO FIND OUT BUT COULDN'T
YES--~What was your source of informationi

YES----..How many nights? _ __

8 ) On average, how crowded did you feel the river
was when you were fishing today? (CIRCLE
THE NUMBER BEST REPRESENTING YOUR FEELI!IGS)

13} How would a lower water level affect each of
the followin~0
Don't
Effect ~ ~ Know

My chances of
catching fish
Not at

all
Crowded
9 l

Slightly
Crowded

Moderately
Crowded

Extremely
Crowded

How important would each of the following be
in contributing to an excellent or perfect
fishing trip at Lee's Ferry for you? {CIRCLE
ONE NUMBER FOR EACH ITEM)
Not
Somewhat
Very
Important Important Important

Catching a trophy fish

My chances of
catching a trophy
fish
Allloun t of time
spent fishing
My chances to fish

certain areas
I like
Problems with
boat/motor

Catching your limit
Good Weather
High water level
Low water level

14) Did you go all the way upstream (as far as
you were allowed) today?
YES
NO---..-.Why not? ~~~~~~~~~~-

Camping along the river l
Seeinq few others
Rising water level
during the day
Falling water level
during the day
Other ~~~~~~~~~~~~~~~~~~-

15) What i• your zip coda?
16) Please circle the response that comes closest
~~I~~~~ 6~~a~~~~fy income before taxes.

LESS THAN $10,000
$10,000 - $24,999
$25,000 - $39,999

4
5
6

$40,000 - $59,999
$60,000 - $79,999
$80,000 O!'. "ORE

a

G.1

'

APPENDIX G
GLD C.ARYOH OGLERS OH-SITE PRE-OOHTIHGEHT-V.ILUATIOH SURVEY

'

G. 2
1985 LEE'S FERRY ANGLERS SURVEY

This survey asks you several questions about your current fishing trip. In
answering these questions, please assume that this current trip began when you
left home to come to Lee's Ferry to fish.

'

How many days have you been fishing at Lee's Ferry on this current trip,
so far?

1.

___ days, so far

2.

Overall, how would you rate your fishing trip so far?

(CIRCLE ONE NUMBER)

1

POOR

2

FAIR, it just hasn't worked out very well
GOOD, but a number of things could have been different
VERY GOOD, but could have been better
EXCELLENT, only minor problems
PERFECT

3
4
5
6

3. Compared to your expectations, how has the fishing been so far?

ONE NUMBER)
l
2
3

4
5

(CIRCLE

MUCH WORSE THAN I EXPECTED
SOMEWHAT WORSE THAN I EXPECTED
ABOUT WHAT I EXPECTED
SOMEWHAT BETTER THAN I EXPECTED
MUCH BETTER THAN I EXPECTED

4. On average, how crowded did you feel on the river while you were fishing?
(CIRCLE THE NUMBER BEST REPRESENTING YOUR FEELINGS)
1

2

Not at all
Crowded

3

4

Slightly
Crowded

5

6

7

Moderately
Crowded

8

9

Extremely
Crowded

5. Have you, personally, caught any fish on this trip so far?
l

NO

2

YES------>How many fish did you catch? ___ fish
------>What was the largest fish you caught?
--~lbs

--~

(CIRCLE ONE)

inches

6. Have you kept any fish that you personally caught on this trip so far?

(CIRCLE ONE)
1

NO

2

YES------>How many fish have you kept?
------>What is the largest fish you kept?
--~lbs

--~

inches

7. How would you rate the water conditions today?

1
2
3

fish

(CIRCLE ONE)

WATER WAS TOO HIGH
WATER WAS ABOUT RIGHT
WATER WAS TOO LOW

Name

S t a t e - - - - - - - Zip - - Telephone _.__ __,_________

i

H.1

'
..
APPENDIX B
GLER cumB ABGLERS OOBTIBGENT-VALUATIOB SURVEY

II

H.2
I

Earlier this year, on

you filled out a short

~~~~~~~-'

survey when you were fishing at Lee's Ferry.

When answering

questions in this survey, we would like you to think about that
trip.

On that particular trip you had already caught

fish

(which included any fish you may have caught and released).

The

largest fish you had caught was about

inches

lbs. and was

long.

i

H.3
I

Pa e 1
We would like you to recall the trip when you filled out the short
questionnaire at Lee's Ferry. How did that trip turn out for you?

A1. Overall, how would you rate the fishing on that trip?
NUMBER)
1 POOR
2 FAIR, IT JUST DIDN'T WORK OUT WELL

(CIRCLE ONE

3 GOOD, BUT I WISH A NUMBER OF THINGS COULD HAVE BEEN
DIFFERENT

4 VERY GOOD, BUT COULD HAVE BEEN BETTER
5 EXCELLENT, ONLY MINOR PROBLEMS
6 PERFECT
A2. On that trip, what was your main method of fishing?
NUMBER)

(CIRCLE ONE

1 FROM A BOAT
2 FROM THE BANK
A3. How many days, in total, did you spend fishing at Lee's Ferry
during the trip when you filled out the short survey?
_ _ TOTAL DAYS OF FISHING ON THAT TRIP
A4. How many fish did you, personally, catch on that trip? (Include
those you reported the day you filled out our survey at Lee's
Ferry, as well as any fish you may have released.)

---

•

FISH

A5. What was the biggest fish you, personally, caught on that trip?
(Give your best estimate) •

- - LBS.

___ INCHES

A6. How many fish did you, personally, keep on that trip?
_ _ FISH

H.4
Page 2
A7. How would you best describe your reason(s) for taking the trip you
were on when you filled out our survey at Lee's Ferry'? (CIRCLE ONE
NUMBER)
FISHING AT LEE'S FERRY WAS THE ONLY REASON FOR THE TRIP
2 FISHING AT LEE'S FERRY WAS THE MOST IMPORTANT REASON FOR
THE TRIP
3 FISHING AT LEE'S FERRY WAS ONE OF SEVERAL EQUALLY
IMPORTANT REASONS FOR MAKING THAT TRIP
4 FISHING AT LEE'S FERRY WAS NOT AN IMPORTANT REASON FOR
MAKING THAT TRIP

AB. How would you describe the destinations of the trip you were on when
you filled out our survey at Lee's Ferry'? (CIRCLE ONE NUMBER)
1 LEE'S FERRY WAS THE SOLE DESTINATION OF THAT TRIP
2 LEE'S FERRY WAS THE MOST IMPORTANT DESTINATION OF THAT TRIP
3 LEE'S FERRY WAS ONLY ONE OF SEVERAL EQUALLY IMPORTANT
DESTINATIONS ON THAT TRIP
4 LEE'S FERRY WAS JUST AN INCIDENTAL STOP ON THE WAY TO
SOME OTHER DESTINATION
A9. Was any part of your trip to Lee's Ferry by airplane'?
1 YES---->Skip to Question 14, next page
2 NO
A10. About how long did it take you to travel from your home to Lee's
Ferry'?
___ DAY(S) _ _ _ TOTAL HOURS OF DRIVING
A11. What type of vehicle did you use to get to Lee's Ferry'?
NUMBER)
1
2
3
4
5

FULL SIZED AUTOMOBILE
INTERMEDIATE SIZED AUTOMOBILE
COMPACT AUTOMOBILE
SMALL TRUCK (Toyota, Nissan, Courier, etc.)
R.V., FULL SIZE TRUCK, VAN

(CIRCLE ONE
I

H.5

'

Page 3
A12. Were you pulling a boat or trailer?
1 YES
2

NO

A13. How many people travelled with you (in the same vehicle) to Lee's
Ferry?
MYSELF AND _ _ OTHER PEOPLE
A14. As near as you can recall for the trip when you filled out our short
survey, about how much was your share of total trip expenses for the
following items? (Include only money you personally spent. If you
didn't spend money on a certain item, please put $0). [PLEASE
CALCULATE AND FILL IN THE TOTAL ON THE LAST LINE].
Gas and Oil for vehicle

$,_____

Food and Beverages

$_ _ _ __

Lodging, Camping

$._____

Fishing equipment/bait/license $'------Guide fees

$'-------

Boat/equipment rental

$'-------

Airfare

$._____

Car rental

$._____

Other

$._ _ _ __

TOTAL YOU SPENT OH THIS TRIP

$._ __

A15. Would you still have gone on that particular trip to Lee's
Ferry if your expenses had been $
more than the total you just
calculated? (CIRCLE ONE NUMBER)
YES, the trip would still be worthwhile
2 NO, it would not be worthwhile

H.6

PLEASE READ CAREFULLY

River flows at Lee's Ferry are affected by the operation of Glen
Canyon Dam. River flows, in turn, may have an impact on the quality
of a fishing trip at Lee's Ferry in many ways. Fish might be easier
or harder to catch, access along the banks may be better or worse, or
boats may be easier or harder to handle.
In the next series of questions we will describe several types of
river conditions at Lee's Ferry. For each condition, we have
described the pctential effects on fishing. Our description of these
effects are based on previous studies of fishing at Lee's Ferry.
They are included to help you evaluate the fishing under the various
conditions. Since these descriptions are based on the general
opinions of Lee's Ferry anglers, your own personal opinions about the
effects may differ. For each type of condition, we would like you to
tell us how the river flow would have affected the quality of the
fishing trip you were on when you filled out the short survey.
Records from Glen Canyon Dam show that the average water flow that
day was
cubic feet per second (cfs), with a high flow of
cfs and a low flow of
cfs.
To help put these numbers in perspective, bank and boat anglers
sometimes have difficulty fishing at high water levels (over 25,000
cfs) because of the swift current. In addition, some anglers may
experience trouble in handling their boats in flows above 25,000
cfs. In the past, the National Park Service has imposed a 25
horsepower requirement on motors when flows have exceeded 40,000
cfs. On the other end of the scale, low water levels (below 9,000
cfs) may tend to concentrate the fish. While low water levels may
make access easier for bank anglers, boat anglers start experiencing
more damage to boats and motors. In addition, it is known that very
few boat anglers can pass over the sand and gravel bar three miles
upstream from Lee's Ferry when the flow is less than 3,000 cfs.

H.7
CASE 1

Boat anglers have said that getting upstream to fish can sometimes be a
problem at low water (3,000 cfs or less). At a constant flow of 3,000
cfs, large boats can't get past the sand and gravel bar three miles
upstream from Lee's Ferry, while even very small boats may have to be
dragged over slippery rock gravel bars. Consequently, nearly all of
the fishing would occur in the three miles just above Lee's Ferry. In
addition, damage to boats and motors is somewhat more frequent than at
higher water levels. However, low water tends to concentrate fish, and
bank anglers can find large areas of exposed gravel and rocks leaving a
great deal of space between the water and the edge of the vegetation.
B1.

Do you think a fishing trip under the conditions described above
(water level constant at 3,000 cfs) would be better or worse than
the trip you took when you filled out our short survey at Lee's
Ferry? (CIRCLE ONE NUMBER)
1

BETTER

2

WORSE

3 ABOUT THE SAME
B2. If the river conditions were always like those described above for
Case 1 (constant flows of 3, 000 cfs), would it affect the number of
times you would go to Lee's Ferry to fish in a typical year?
1

YES----->How so?

(CIRCLE ONE NUMBER)

1

I would make FEWER trips per year--How many?
FEWER TRIPS PER YEAR

2

I would make MJRE trips per year--How many?
MORE TRIPS PER YEAR

2

NO, IT WOULDN'T AFFECT HOW OFTEN I COME TO LEE'S FERRY TO FISH

We would like you to imagine that you are deciding whether or not to
take a trip to Lee's Ferry. The conditions (weather, number of anlil.ers
coming to Lee's Ferry, etc.) on this trip would be the same as when you
filled out our survey with two exceptions:
The water level would be constant at 3,000 cfs (see Case 1)
Your expenses increased by
calculated on page 3)
B3. Would you go on this trip?

$~~­

(over the total expenses you

(CIRCLE ONE NUMBER)

1 YES, it would still be worthwhile
2 NO, it would not be worthwhile

H.8
CASE 2

The questions above asked about a relatively constant flow of 3,000
cfs. Daily changes in the water flow may have other effects, in
addition to those described in case 1, on the quality of fishing. With
flows changing daily from a low of 1,000 cfs to a high flow of 15,000
cfs (around an average flow of 3,000 cfs), boats may get swamped if they
are tied too tightly to the bank during a fluctuation. There is also a
chance of getting stranded above 3 mile bar if the water drops
substantially. On the other hand, biological studies give some
indication that rising water may cause the fishing to improve as fish
begin to feed on the debris stirred up by the rising water.
B4. How do you think the fishing experience would be if there were large
daily fluctuations (with the conditions described in Case 2 above)
around a base of 3,000 cfs as opposed to being constant at 3,000
cfs? (CIRCLE ONE NUMBER)

1 I THINK IT WOULD MAKE NO DIFFERENCE
2 THE FLUCTUATIONS WOULD HURT THE FISHING EXPERIENCE

3 THE FLUCTUATIONS WOULD IMPROVE THE FISHING EXPERIENCE
B5. If the river conditions were always like those described above for
Case 2 (daily changes around an average flow of 3,000 cfs), would
it affect the number of times you would go to Lee's Ferry to fish
in a typical year?
1 YES----->How so?

(CIRCLE ONE NUMBER)

1 I would make FEWER trips per year--How many?
_ _ FEWER TRI PS PER YEAR
2

I would make MORE trips per year--How many?
_ _ MORE TRI PS PER YEAR

2

NO, IT WOULDN'T AFFECT HOW OFTEN I COME TO LEE'S FERRY TO FISH

Now we would like you to imagine that you are deciding whether or not to
take a fishing trip to Lee's Ferry. The conditions (weather, number of
anglers on the river, etc.) on this trip would be the same as the trip
when you filled out our survey with two exceptions:
There would be large daily fluctuations from a low flow of 1,000
cfs to a high flow of 15,000 cfs around an average flow of 3,000
cfs (see descriptions for case 2 above).
AND

Your expenses increased by $---- (over the total you
calculated on page 3)

B6. Would you go on this trip? (CIRCLE ONE NUMBER)
1 YES, it would still be worthwhile
2 · NO, it would not be worthwhile

H.9
CASE 3

t

Boat anglers seem to experience fewer problems with damage to their
boats and motors when the water is at least 10,000 cfs, and boats can
get up and downstream with no difficulty. At a flow of 10,000 cfs, bank
anglers would still find exposed gravel and rock bars and some room
between the water's edge and shore vegetation. In previous studies,
about 40 percent of the anglers have said that they feel the fishing is
generally better at constant flows of 10,000 cfs than when the water
level is higher.
C1. Do you think a fishing trip under the conditions described above for
Case 3 (constant flow of 10,000 cfs) would be better or worse than
the trip when you filled out our short survey at Lee's Ferry?
(CIR CLE ONE NUMBER )

BETTER
2 WORSE
3

ABOUT THE SAME

C2. If the river conditions were always like those described above in
Case 3 (constant at 10,000 cfs), would it affect the number of times
you would go to Lee's Ferry to fish in a typical year?
YES----->How so?

(CIRCLE ONE NUMBER)

I would make FEWER trips per year--How many?
FEWER TRIPS PER YEAR
2

I would make MORE trips per year--Hcw many?
_ _ t-'.ORE TRI PS PER YEAR

2

NO, IT WOULDN'T AFFECT HOW OFTEN I COME TO LEE'S FERRY TO FISH

We would like you to imagine you are deciding whether or not to take a
trip to Lee's Ferry. The conditions (weather, number of anglers coming to
Lee's Ferry, etc.) on this trip would be exactly the same as the trip when
you filled out our short survey with two exceptions:

f

The water flow would be constant at 10,000 cfs (as described
in Case 3 above)
AND

Your travel expenses increased by $
expenses you calculated on page 3).
C3. Would you make this trip?

~~­

(CIRCLE ONE NUMBER)

YES, it would still be worthwhile
2 NO, it would not be worthwhile

(over the total

H.10
CASE I&

The questions above asked about a relatively constant flow of 10,000 cfs.
Daily changes in the flow may have other effects on the quality of fishing
in addition to those described above for Case 3. At moderate flows, large
daily fluctuations from a low flow of 1,000 cfs to a high flow of 22,000
cfs (around an average of 10,000 cfs) may contribute to the swamping of
boats tied to the bank or dragging anchors. There would still be a chance
of getting stranded above 3 mile bar. Again, however, there is some
indication that the rising water may improve fishing as fish begin to feed
on debris stirred up by the rising water.
Cl&. How do you think the fishing experience would be if there were large
daily fluctuations (with the conditions described in Case 4) around a
base of 10,000 cfs as opposed to being constant at 10,000 cfs? (CIRCLE
ONE NUMBER)

1 I THINK IT WOULD MAKE NO DIFFERENCE
2 THE FLUCTUATIONS WOULD HURT THE FISHING EXPERIENCE

3 THE FLUCTUATIONS WOULD IMPROVE THE FISHING EXPERIENCE
CS. If the river conditions were always like those described above for
Case 4 (daily changes around an average flow of 10,000 cfs), would it
affect the number of times you would go to Lee's Ferry to fish in a
typical year?
1 YES----->How so?

(CIRCLE ONE NUMBER)

1 I would make FEWER trips per year--How many?
___FEWER TRIPS PER YEAR
2
2

I would make IDRE trips per year--How many?
___MORE TRIPS -PER YEAR

NO, IT WOULDN'T AFFECT HOW OFTEN I COME TO LEE'S FERRY TO FISH

Now we would like you to imagine that you are deciding whether or not to
take a fishing trip to Lee's Ferry. The conditions (weather, nunber of
anglers on the river, etc.) on this trip would be the same as the trip
when you filled out our survey with two exceptions:
There would be large daily fluctuations from a low flow of 1,000
cfs to a high flow of 22,000 cfs around an average flow of 10,000
cfs (see Case 4 above) •
.ARD

Your expenses were increased by $---- (over the total you
calculated on page 3)
C6. Would you make this trip?

(CIRCLE ONE NUMBER)

1 YES, it would still be worthwhile
2 NO, it would not be worthwhile

H. 11
CASE 5

Next, consider a constant flow of 25,000 cfs. There is no minimum motor
horsepower restriction, although motors with 10 hp or less may have
problems getting upstream. The chance of damage to boats and motors due
to obstructions in the water are small, but the high flows of 25,000 cfs
may increase the chances of swamping a boat while dragging an anchor,
especially for inexperienced boaters.
Fish may be less concentrated at this higher flow level. Bank anglers
may have less space between the waters edge and bank vegetation, but
eddies along the shoreline are often larger and more pronounced. About
12 percent of the anglers in a recent survey felt that fishing was better
at 25,000 cfs than at lower water levels in Glen Canyon.
D1. Do you think a fishing trip under the conditions described for Case
5 above (constant flow of 25,000 cfs) would be better or worse than

the trip when you filled out our short survey at Lee's Ferry?
1

BETTER

2

WORSE

3

ABOUT THE SAME

D2. If the river conditions were always like those described above in
Case 5 (constant at 25,000 cfs), would it affect the number of times
you would go to Lee's Ferry to fish in a typical year? (CIRCLE ONE
NUMBER)
YES----->How so? (CIRCLE ONE NUMBER)
I would make FEWER trips per year--How many?
_ _ FEWER TRIPS PER YEAR
2

I would make MORE trips per year--How many?
_ _ MORE TRIPS PER YEAR

2

NO, IT WOULDN'T AFFECT HOW OFTEN I COME TO LEE'S FERRY TO FISH

We would like you to imagine you are deciding whether or not to take a
trip to Lee's Ferry. The conditions (weather, ni..mber of anglers coming
to Lee's Ferry, etc.) on this trip would be exactly the same as the trip
when you filled out our short survey with two exceptions:

'

The water flow would be constant at 25,000 cfs (see Case 5 above)
AND

Your expenses increased by $ _ __
calculated on page 3).
D3. Would you make this trip?

(over the total expenses you

(CIRCLE ONE NUMBER)

1 YES, it would still be worthwhile
2 NO, it would not be worthwhile

H.12
CASE 6
The impacts of large daily fluctuations are somewhat different at higher
water than at lower water levels. With fluctuations from low flows of
12,000 cfs to high flows of 32,000 cfs (around an average flow of 25,000
cfs), it is very unlikely boats would get stranded above 3 mile bar.
Boats tied too tightly to the shoreline, however, may be flooded. Rising
water might also trigger more feeding by fish, but fish become more
difficult to find because of the higher water and faster current. In the
long run, large daily fluctuations at this flow level may wash away many
of the campsites upstream from Lee's Ferry.

'

D4. How do you think the fishing experience would be if there were large
daily fluctuations (with the conditions described in Case 6) around an
average flow of 25,000 cfs as opposed to being constant at 25,000
cfs? (CIRCLE ONE NUMBER)
I THINK IT WOULD MAKE NO DIFFERENCE
2

THE FLUCTUATIONS WOULD HURT THE FISHING EXPERIENCE

3 THE FLUCTUATIONS WOULD IMPROVE THE FISHING EXPERIENCE
D5. If the river conditions were always like those described above for Case
6 (daily changes around an average flow of 25,000 cfs), would it affect
the number of times you would go to Lee's Ferry to fish in a typical
year?
1 YES----->How so?

(CIRCLE ONE NUMBER)

1 I would make FEWER trips per year--How many?

--2
2

FEWER TRIPS PER YEAR

I would make M:>RE trips per year--How many?
_ _ MORE TRIPS PER YEAR

NO, IT WOULDN'T AFFECT HOW OFTEN I COME TO LEE'S FERRY TO FISH

Now we would like you to imagine that you are deciding whether or not
to take a fishing trip to Lee's Ferry. The conditions (weather, nunber
of anglers coming to Lee's Ferry, etc.) on this trip would be the same
as the trip when you filled out our survey with two exceptions:
0

There would be large daily fluctuations from a low flow of
12,000 cfs to a high flow of 32,000 cfs around an average flow
of 25,000 cfs (see Case 6).
AND

Your expenses increased by $---- (over the total you
calculated on page 3)
D6. Would you make this trip?

(CIRCLE ONE NUMBER)

YES, it would still be worthwhile
2 NO, it would not be worthwhile

'

H.13
CASE 7
At constant flows of 40,000 cfs, the current is swift and the Park
Service requires all boat motors to have at least a 25 horsepower
motor. Large boats can get up and down the river more easily than
smaller boats. The chances of damage to boats and motors due to
obstructions in the water are smaller than at lower flows. However,
for inexperienced boaters the high water increases the chances of boats
being swamped while dragging anchors. Eddies along the shoreline are
larger and well defined, but bank anglers find the water is up into the
bank vegetation and this may make bank fishing more difficult for
them. At these high flows, fish feeding patterns may change since fish
would generally stay out of the main current. Fish may be harder to
find.
E1. Do you think a fishing trip under the conditions described above
for Case 7 (constant flow of 40,000 cfs) would be better or worse

than the trip when you filled out our short survey at Lee's Ferry?
1 BETTER
2 WORSE

3 ABOUT THE SAME
E2. If the river conditions were always like those described above in
Case 7 (constant at 40,000 cfs), would it affect the number of
times you would go to Lee's Ferry to fish in a typical year?
(CIRCLE ONE NUMBER)
1 YES----->How so?

(CIRCLE ONE NUMBER)

1 I would make FE.WER trips per year--How many?
_ _ FEWER TRIPS PER YEAR
2
2

I would make MORE trips per year--How many?
_ _ MORE TRIPS PER YEAR

NO, IT WOULDN'T AFFECT HOW OFTEN I COME TO LEE'S FERRY TO FISH

We would like you to imagine you are deciding whether or not to take a
trip to Lee's Ferry. The conditions (weather, nl1Dber of anglers coming
to Lee's Ferry, etc.) on this trip would be exactly the same as the
trip when you filled out our short survey with two exceptions:
The water flow would be constant at 40,000 cfs (see Case 7 above)
AND

Your expenses increased by $ _ __
calculated on page 3).
E3. Would you make this trip?

(over the total expenses you

(CIRCLE ONE NUMBER)

1 YES, it would still be worthwhile
2 NO, it would not be worthwhile

H. 14

In addition to short term effects on fishing success, access to the
river, and ease of boat handling, river flows may have longer term
impacts on the size and number of fish in the river around Lee's Ferry.

These last two cases describe two possible changes in fishing success.
After reading the description of each change, we would like you to tell
us how that change would affect the quality of your fishing trip(s) at
Lee's Ferry.

.

.

,--~-~------------

---

H.15

.

CHANGE 1

A survey of anglers at Lee's Ferry last year showed that about 15
percent of them reported catching a fish larger than three pounds, and
only 3 percent reported catching a fish larger than four pounds.
These numbers reflect how an average angler might do on any particular
day at Lee's Ferry. We realize that no one is exactly average, but we
would like you to suppose that the fishery at Lee's Ferry changed in
such a way that your chances of catching one of these bigger fish were
to double. If you feel you are an average fisher, your chances of
catching a fish bigger than three lbs. would now be about 30 percent,
while your chances of catching a fish bigger than four lbs. would now
be about 6 percent. If you think you are not an average fisherman at
Lee's Ferry, your chances would vary accordingly.
F1. If the fishery changed in this way (so that your chances of

catching a fish bigger than three pounds had doubled) would it
affect the nunber of times you would travel to Lee' Ferry to fish,
in an average year? (CIRCLE ONE NUMBER)
1 YES----->How so? (CIRCLE ONE NUMBER)
1 I would make FEWER trips per year--How many?
_ _ FEWER TRIPS PER YEAR
2
2

I Would make M:JRE trips per year--How many?
_ _ MORE TRIPS PER YEAR

NO, IT WOULDN'T AFFECT HOW OFTEN I COME TO LEE'S FERRY TO FISH

Now we would like you to imagine you are deciding whether or not to take
a trip to Lee's Ferry. On this trip all of the fishing conditions (water
levels, weather, number of other anglers coming to Lee's Ferry, etc.)
would be the same as the trip when you filled out our survey with two
exceptions:
Your chances of catching a big fish (over 3 pounds) would be
doubled
AND

""'

Your expenses of a trip to Lee's Ferry increased
(over the total you calculated on page 3)
F2. Would you go on this trip?

(CIRCLE ONE NUMBER)

YES, it would still be worthwhile
2

NO, it would not be worthwhile

by$~~­

H.16
CHARGE 2

We are sure that almost every angler has experienced, at one time
or another, "getting skunked" (catching no fish at all). In fact,
about 20 percent of our respondents to a previous survey at Lee's Ferry
reported they had not yet caught a fish. This number reflects how an
average angler might do on any particular day at Lee's Ferry. No one
is exactly average, but we would like you to suppose that the fishery
at Lee's Ferry changed in such a way that your chances of getting
skunked were to double. In other words, if you feel you are an average
angler and came repeatedly to Lee's Ferry, you could expect that on
four out of ten trips you would catch no fish at all. If you were an
above average angler, your chances of getting· skunked would be less,
and if you were below average, the chances would be greater than 4 out
of 10 times.

G1. If the fishery changed in this way (so that your chances of catching

no fish doubled) would it affect the number of times you would travel
to Lee's Ferry to fish in an average year? (CIRCLE ONE NUMBER)
1 YES----->How so?

(CIRCLE ONE NUMBER)

I would make FEWER trips per year--How many?
FE.WER TRIPS PER YEAR
2

I would make MORE trips per year--How many?
MORE TRIPS PER YEAR

2

NO, IT WOULDN'T AFFECT HOW OFTEN I COME TO LEE'S FERRY TO FISH

Now we would like you to imagine that you are deciding whether or not to
make a trip to Lee's Ferry. On this trip, all of the fishing conditions
(water levels, weather, number of other anglers coming to Lee's Ferry,
etc.) would be the same as the trip when you filled out our survey, with
two exceptions:
Your chances of catching no fish would double
AHD

Your expenses increased by $ _ __ (over the total expenses you
calculated on page 3)
G2. Would you go on this trip?

(CIRCLE ONE NUMBER)

1 YES, it would still be worthwhile
2

NO, it would not be worthwhile

H.17
In the previous sections there were a number of questions where we asked
whether you would still make the trip to Lee's Ferry to fish if your
expenses increased by a certain amount.
To help us better understand your responses, we would like to know the
extent to which various factors affected your answers to the expense
questions. Please tell us whether the following statements were true or
not for you when answering the previous questions (CIRCLE ONE RESFONSE
FOR EACH STATEMENT).
Definitely Probably Probably Definitely
True
True
False
False
My main concern was that
license fees may be increased
for fishing at Lee's Ferry.

1

2

3

4

My responses reflected my
best guess as to whether the
described trip would have been
worthwhile.

1

2

3

4

I just don't want to have
to pay more to fish at Lee's
Ferry, regardless of the
conditions.

1

2

3

4

My responses reflected the
fact that I don't think dollar
values should be put on the
fishing experience at Lee's
Ferry.

1

2

3

4

Are there any other factors
that affected your answers?

..

(please describe: - - - - - - - - - - - -

H.18

In this section we would like to find out about your fishing background.

I1. Was 1985 your first year of fishing at Lee's Ferry?

(CIRCLE ONE)

1 YES
2

NO----->Including this year, how many years have you fished
at Lee's Ferry?
___ YEARS
----->About how many trips do you make to Lee's Ferry in
an average year?
_ _ TRIPS

I2. Was the fishing trip when you filled out our short survey the only
trip you have made to Lee's Ferry in the last 12 months? (CIRCLE
ONE)
1 YES
2

NO----->Including the trip when you filled out our short
survey, how many trips have you made in the last 12
months?
_ _ TRIPS
----->What is the average number of days you spend at
Lee's Ferry on a typical fishing trip there?
_ _ DAYS

I3. Some people have many other activities which they enjoy as much as
fishing at Lee's Ferry. Others have very few. Which of the
following statements most closely reflects how you feel? (CIRCLE
ONE)
If I couldn't go fishing at Lee's Ferry I probably would
not miss it at all and would find something else that was
just as enjoyable
2

If I couldn't go fishing at Lee's Ferry, I would miss it,
but not as much as a lot of other things I now enjoy

3

If I couldn't go fishing at Lee's Ferry, I would miss it
more than most of the other interests I now enjoy

4

If I couldn't go fishing at Lee's Ferry, I would miss it
more than all of the other interests I now enjoy

H.19

•
In this final section, we would like to ask sane questions about your
background and occupation which will help us compare your answers with
those of other people. We stress that all of your answers are
strictly confidential.

J1.

How old are you? ___ YEARS OLD

J2.

Are you:
1

MALE

2

FEMALE

J3. How many years of school have you completed? (CIRCLE OR CHECK THE
HIGHEST YEAR OR LEVEL)
1

4
3
Some college
2

5

6

7

8

9

10

11

12

_M.A., M.S.
Advanced degree (M.D., Ph.D)

B.A. or equivalent

J4. Please circle the response that comes closest to your total family
income before taxes. (CIRCLE ONE NUMBER)
1 Less than $10,000
2 $10,000 to $19,999

J5.

..

$20,000 to $29,999

3
4

$30,000 to $39,999

5

$40,000 to $49,999

6
7
8

$50,000 to $59,999
$60,000 to $69,999
$70,000 to $79,999
$80,000 to $89,999

9
10 $90,000 to $99,999
11 $100,000 or more

With reference to your primary occupation, are you currently (CIRCLE
ONE NUMBER) :
1

EMPLOYED FULL-TIME

2

EMPLOYED PART-TIME

3 FULL-TIME HOMEMAKER
4 TEMPORARILY UNEMPLOYED
5

NOT EMPLOYED AND NOT LOOKING FOR WORK

6

RETIRED, NOT WORKING
RETIRED, WORKING PART-TIME

7

I. 1

•

APPENDIX I

DAY-USE RAFTERS ATTRIBUTE SURVEY

I. 2

**********************************************************************

In this section, we would like to find out about your reasons for
taking this raft trip and how you enjoyed it.

**********************************************************************
1.

Do you live in Northern Arizona?
1 NO
2 YES---->Skip to Question 6

2. What was the main reason you came to Northern Arizona?

3. How many days did you stay in Northern Arizona?
_ _ day(s)
4. Were you aware of the one-day Glen Canyon raft trip before you came
to Northern Arizona? (CIRCLE ONE NUMBER)
NO---->Where did you learn about it? (CIRCLE ONE NUMBER)
1 IN THE PAGE AREA
2

AT THE SOUTH RIM OF THE GRAND CANYON

3

OTHER_~~~~~~~~~~~~~

2 YES---->How did you learn about the raft trip?
ONE NUMBER)

(CIRCLE

1 AD IN TRAVEL MAGAZINE
2 FROM OTHERS WHO HAD TAKEN THE TRIP
RAFI' TRIP COMPANY BROCHURE
4 WROTE TO GRAND CANYON NATIONAL PARK
3

5 OTHER---------------

•

I. 3

5. Was the chance to go on this one-day raft trip an important reason
to you in deciding to come to the Northern Arizona? (CIRCLE ONE
NUMBER)
1 NOT AT ALL IMPORTANT REASON
2 SOMEWHAT IMPORTANT REASON
3 VERY IMPORTANT REASON
4 I WASN'T AWARE OF THE ONE-DAY RAFI' TRIP WHEN I DECIDED

6. If you had the opportunity, would you take this raft trip again?
(CIRCLE ONE NUMBER)
DEFINITELY NOT
2 PROBABLY NOT
3 PROBABLY YES
4 DEFINITELY YES

7. Overall, how would you rate your raft trip? (CIRCLE ONE NUMBER)
1 POOR
2 FAIR, it just didn't work out very well
3 GOOD, but a number of things could have been different
4 VERY GOOD, but could have been better
5 EXCELLENT, only minor problems
6 PERFECT

8. What was your main reason for taking the Glen Canyon raft trip?
PLEASE TRY TO BE AS SPECIFIC AS YOU CAN.

'

•

I.4

9. What things would contribute most to an excellent or perfect
one-day raft trip in the Glen Canyon area for you?

10. What things would contribute most to a poor one-day raft trip in
the Glen Canyon area for you?

11. Including yourself, about how many people were there on this raft
trip? (INCLUDE YOUR GUIDE OR TRIP LEADER AND ALL OF THE RAFTS
THAT WERE WITH YOUR RAFT.)
___PEOPLE

12. Would you have liked a tour of the Glen Canyon Dam as part of

your raft trip?
1

(CIRCLE ONE NUMBER)

NO

2 YES---->How do you think this would have improved your
raft trip experience?

'

•

I. 5

...

•

13. How many white water raft or kayak trips have you taken? Do not
include the one-day raft trip in Glen Canyon. (CIRCLE ONE
NUMBER)
1

NONE

2

1-2

3
4

3-5

5

11-20

6

MORE THAN 20

6-10

14. Do you think the Glen Canyon river trip is a good substitute for a
river trip through the Grand Canyon?
NO
2

'

•

YES

(CIRCLE ONE NUMBER)

I.6

15. Raft trips through the Glen Canyon area have a number of features
and people differ in what they feel is important to them
personally. In this next section, we list a number of features of
a Glen Canyon raft trip. Please indicate how important each
feature was for you on your trip. (CIRCLE ONE NUMBER FOR EACH
ITEM)
Not at all Somewhat Very
Important Important Important

Did not
Experience

Being in a natural setting

1

2

3

0

Seeing wildlife

1

2

3

0

Being able to say I've
been to the Glen Canyon

1

2

3

0

Being with family or friends

1

2

3

0

Interacting with my guide
or trip leader

1

2

3

0

Learning about the
history of Glen Canyon

1

2

3

0

Relaxing; getting away
from it all

1

2

3

0

Stopping for lunch along
the river

1

2

3

0

Well paced and organized trip 1

2

3

0

Starting the trip at the dam

1

2

3

0

Good weather

1

2

3

0

Stopping to explore along
the river

2

3

0

Seeing few others while
on the river

2

3

0

2

3

0

Learning about others on
the trip

1

•

I. 7

Not at all Somewhat Very
Important Important Important

•

Did not
Experience

Feeling safe

1

2

3

0

Having confidence in my guide
or trip leader

1

2

3

0

Not having to make advance
pl ans for the river trip

1

2

3

0

Seeing the Canyon in one day

1

2

3

0

Seeing archeological sites
in the Canyon

1

2

3

0

2

3

0

2

3

0

2

3

0

2

3

0

Floating without the motor on
a quiet stretch of the river
Learning a bout the Glen
Canyon Dam

1

Being on the Colorado River
Other

1

16. On average, how crowded did you feel the river was when you were
rafting on this trip? (Circle the number on the scale which best
represents your feelings.)
2

Not at all
Crowded

3

4

Slightly
Crowded

5

6

Moderately
Crowded

7

8

9

Extremely
Crowded

r.8

••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

These next questions are about the water level on the river the day
you took your trip •

••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••
1. When you signed up for this trip, did you know the expected water
level on the Colorado river for the date of your trip?
1 NO
2 YES----->Did this information about the expected water
level have any influence on yoilr decision WHEN
to take this trip?
1

NO

2

YES (please explain)

------

2. Did you notice the water level changing during your raft trip?
NO
2

YES

3. Was the speed of the water (the current) during your raft trip:
(CIRCLE ONE NUMBER)
1

TOO SLOW

2

ABOUT RIGHT

3

TOO FAST

4

DON'T KNo.i

4. If you had your choice, would you have preferred the water level
be:

(CIRCLE ONE NUMBER)

1

LOWER

2

ABOUT THE SAME

3 HIGHER
4

DON IT KNOW OR 00 ESN IT MATTER

•

I.9

5. What a guide or trip leader does and says during a trip can also
affect a person's trip. During your trip, how often did your
guide or trip leader do the following: (CIRCLE ONE NUMBER FOR
EACH ITEM)
Never

Sometimes Often

Didn't
Notice

Turned the motor on to make time

1

2

3

0

Manuevered the raft around rocks

1

2

3

0

Commented that the water
level was too low

1

2

3

0

Commented that the water level
was too high

1

2

3

0

Commented that the current was
too fast

1

2

3

0

Commented that the current was
too slow

1

2

3

0

Shut motor off to talk about
the River/Canyon

1

2

3

0

Shut off the motor because of
low water or rocks

1

2

3

0

2

3

0

Pointed out archeological or
other attractions

6. Did your guide or trip leader discuss how the Glen Canyon
affected your trip?
1 NO

Dam

2 YES---->What did he/she say? - - - - - - - - - - - -

I. 10

••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••

In this last section we would like to ask you some questions about
your background which will help us compare your answers to those of
other people. We would stress that all of your answers are strictly
confi den ti al •

••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••
1. How old are you?
2. Are you:

--~years

1

MALE

2

FEMALE

old

3. How many years of school have you completed? (CIRCLE OR CHECK THE
CORRECT RF.SPONSE)
1

2

3

4

5

6

8

7

9

10

11

12

__some college or technical school
__B.A. or equivalent
__M.A. or equivalent
__Advanced degree (M.D. Ph.D., etc.)

4. Please circle the response that comes closest to your total family
income before taxes. If you are a student and unmarried, please
give your parents' income. (CIRCLE ONE NUMBER)
1 Less than $10,000

$50,000 to $59,999
$60,000 to $69,999

2

$10,000 to $17,499

7
8

3
4

$17,500 to $24,999
$25,000 to $32,499

9
10

$70,000 to $79,999
$80,000 to $89,999

5
6

$32,500 to $39,999
$40,000 to $49,999

11

$90,000 to $99,999
$100,000 or more

12

TBARIC YOU FOR YOUR HELP!

If you would like a copy of the results, please check here:

•

J. 1

APPENDIX

J

DAY-USE RAFTERS COHTIBGEMT-VALUATIOH SURVEY

•

J.2

In this first section, we would like to find out about your raft trip
in Glen Canyon and how much you enjoyed it.

1. Overall, how would you rate your raft trip?
1 POOR
2

(CIRCLE ONE NUMBER)

FAIR, it just didn't work out very well

3 GOOD, but a number of things could have been different
4 VERY GOOD, but could have been better
EXCELLENT, only minor problems
6 PERFECT

5

2. If you had the opportunity, would you take this raft trip again?

(CIRCLE ONE NUMBER)
1

DEFINITELY NOT

2

PROBABLY NOT

3

PROBABLY YES
DEFINITELY YES

4

3. Do you live in Northern Arizona?
1

NO

2

YES---->Skip to Question 6 ·

4. Were you aware of the one-day Glen Canyon raft trip before you
came to Northern Arizona?
1 NO
2 YES

(CIRCLE ONE NUMBER)

•

J.3
5. Was the chance to go on this one-day raft trip through Glen
Canyon an important reason to you in deciding to come to Northern
Arizona? (CIRCLE ONE NUMBER)
1 NOT AN IMPORTANT REASON
2 A SOMEWHAT IMPORTANT REASON
3 A VERY IMPORTANT REASON
4 I WASN'T AWARE OF THE ONE-DAY RAFT TRIP WHEN I DECIDED

6. What was your main reason for taking the Glen Canyon raft trip?
(CIRCLE ONLY ONE NUMBER)
1 TO TAKE A TRIP THROUGH GLEN CANYON
2 TO TAKE A TRIP ON THE COLORADO RIVER
3 TO SEE SCENERY
4 TO TAKE A RAFT TRIP
5 TO RELAX AND ENJOY NATURE
6 OTHER, please s p e c i f y - - - - - - - - - - - - - - - -

7. Including yourself, about how many people were there on this raft
trip? (INCLUDE YOUR GUIDE OR TRIP LEADER AND PEOPLE ON OTHER
RAFTS THAT WERE WITH YOUR RAFT.)
_ _PEOPLE.

8. About how many people were on the same raft as you on your trip?

- -PEOPLE

•

J.4

9. How many of your friends or family members accompanied you on
this raft trip?
_ _PEOPLE

10. On average, how crowded did you feel the river was when you were
rafting on this trip? (Circle the number on the scale which best
represents your feelings.)
1

2

Not at all
Crowded

6

3

5

Slightly
Crowded

Moderately
Crowded

7

8

9

Extremely
Crowded

11. If you had a choice, what type of Glen Canyon raft trip would you

prefer:
1

(CIRCLE ONE NUMBER)
A TRIP STARTING WITH A TOUR OF GLEN CANYON DAM AND
FLOATING DOWNSTREAM FROM THE BASE OF THE DAM TO LEE'S
FERRY

2 A TRIP STARTING AT THE BASE OF GLEN CANYON DAM AND
FLOATING DOWNSTREAM TO LEE'S FERRY (A TOUR OF THE DAM IS
NOT INCLUDED)
--

3 A TRIP STARTING AT LEE'S FERRY AND MOTORING PART OF THE
WAY TO THE DAM BEFORE FLOATING BACK DOWNSTREAM (PASSENGERS
CANNOT VIEW THE DAM FROM THE RIVER)

•

J.5

12. Please estimate how much your raft trip cost (COSTS FOR YOU
INDIVIDUALLY, NOT OTHERS FOR WHOM YOU MIGHT HAVE PAID). (Include
only money spent on items specifically for this raft trip. If a
certain item was not purchased for this trip, please put $0).
Payment to Rafting Company

$_ _ _ __

Food and Beverages (not supplied by raft
company)

$~----

Personal Gear (sunglasses, suntan lotion,
film for camera, etc.)

$_ _ _ __

Other, please s p e c i f y - - - - - - - - $_ _ __

TOTAL AH>URT TRIP COST [Please add all costs
and fill in the total on this line.]

$~----

13. Would you still have gone on the Glen Canyon raft trip if your
costs had been $
K>RE than the total you just calculated in
Question 12? (CIRCLE ONE NUMBER)
1 YES, I WOULD PAY THAT AMJUNT TO TAKE THE TRIP
2 NO, I WOULD NOT PAY THAT AMOUNT TO TAKE THE TRIP

•

•

J.6

14. To help us better understand your responses to question 13, we
would like to know the extent to which various factors affected
your answers. Please tell us whether the following statements were
true or not for you when answering question 13. (CIRCLE ONE
RESPONSE FOR EACH STATEMENT)
Definitely
True
My main concern was that
the Park Service might start
charging a fee for Glen
Canyon raft trips.
My main concern was that the
rafting company might
increase the price of Glen
Canyon raft trips.

1

My response reflected my
best guess as to whether I
would really pay the
specified amount for a Glen
Canyon raft trip.

Probably Probably
True
False

Definitely
False

2

3

4

2

3

4

2

3

4

My response reflected the
fact that I feel the Glen
Canyon rafting experience
should be maintained for all
people to enjoy.

1

2

3

4

My response reflected the fact
that I just don't know how much
more I would really pay for
a Glen Canyon raft trip.

1

2

3

4

2

3

4

My response reflected the fact
that I just don't want to pay
more for a Glen Canyon
raft trip.

•

In this last section we would like to ask you some questions about
your backgro\llld which will help us compare your answers to those of
other people. We would stress that all of your answers are strictly
confidential.

1.
2.

How old are you'?
Are you:

MALE
2

3.

years old

FEMALE

How many years of school have you completed'? (CIRCLE OR CHECK THE
CORRECT RESPONSE)
2

1

3

4

5

6

8

7

9

10

11

12

some college or technical school
B.A. or equivalent
M.A. or equivalent
Advanced degree (L. L. D., M. D., Ph.D., etc.)

4.

Please circle the response that comes closest to your total
family income before taxes. If you are a student and unmarried,
please give your parents' income. (CIRCLE ONE NUMBER)
Less than $10,000
2
3
4

•

•

$10,000 to $14,999$15,000 to $19,999

9
10

$45,000 to $149,999
$50,000 to $59,999

11

$60,000 to $69,999
$70,000 to $79,999

12

5
6

$20,000 to $24 '999
$25,000 to $29,999
$30,000 to $34,999

7
8

$35,000 to $39,999
$40,000 to $44,999

15

13
14

$80,000 to $89,999
$90,000 to $99 '999
$100,000 or more

THAHK YOU FOR YOUR HELP!

K. 1

APPENDIX IC
GLEN CARYON OGLER OONTIBGENT-VALUATIOB
PRETF.ST SURVEY AND RESULTS

•

K.2

APPENDIX K
GLEN CAll10R ARGLER OONTIRGENT-VALUATIOR
PRETF.sT SURVEY AND RESULTS
Introduction
In this appendix we present the results of the contingent-valuation
pretest survey of Glen Canyon Anglers. The purpose of this survey
was two-fold. First, the pretest allowed us to refine the
contingent-valuation survey instrument, as was done for the
administration of all attribute and contingent-valuation (CV) surveys
in the study. The second objective, and the unique aspect of the
angler CV pretest survey, is that it was used to identify which of
three CV questioning formats, iterative-bidding, open ended or
dichotomous-choice, would be best to employ in the final CV surveys
of all three user groups (anglers, white-water boaters and day-use
rafters).
In the remainder of this appendix, we will present the results of the
angler CV pretest survey and discuss the relative strengths and
weaknesses of each of the three techniques of asking the CV question
that were employed. After presenting the procedures used for this
survey, we will move to a discussion of the empirical results and
will close with a conclusion regarding the selection of a single CV
technique to be used in the remaining surveys.
Procedures
Sampling. The sampling frame for this pretest survey consisted of
anglers who fished at Lee's Ferry on 75 selected days between April
29 and December 19, 1985. A sampling period of this length was
chosen to minimize the potential for a seasonal bias in the types of
anglers selected to participate in the study. The same procedures,
as were used for the angler on-site attribute surveys, were used to
contact anglers and to solicit their names and addresses for the CV
mail survey.

Our field personnel estimated that 986 anglers were eligible for
selection on the 75 specified sampling days, and they were able to
contact 900 (91 percent). Some anglers were missed during busy times
at the dock, while others had not returned by dark. For the 900
anglers contacted, 774 completed the on-site questionnaire (86
percent) and provided a usable name and address for the mail survey.
The remaining 126 anglers either declined to complete the on-site
questionnaire or listed insufficient or illegible address

•

K.3

information. The sampling frame, then, contained 86 percent of the
anglers contacted and 78 percent of the estimated total number of
anglers.
Three hundred of the anglers who provided usable address information
during the period April 29 through July 29, 1985 were selected to
participate in the pretest survey.
Survey Design. Respondents were asked to answer a total of five
contingent-valuation questions in the surveys. First, they were
asked to evaluate their actual trip when they were contacted for
their on-site preliminary interview. Three flow scenarios were then
evaluated, 3,000, 10,000 and 40;000 cfs. Constant flows, and two
environmental impact scenarios (doubling the chances of catching a
trophy fish and halving the probability of catching no fish, getting
"skunked").

•

To compare the three techniques of asking the CV question,
respondents were randomly assigned to three groups and all
individuals within a subgroup responded to the same CV technique.
The comparison of techniques was based on six criteria of
performance. The first we will refer to as the "performance relative
to preferences" criterion. This requires a CV technique to rank
alternatives in the same order as direct measures of preferences. As
an example, if a substantial majority of anglers prefer a constant
flow of 15,000 cfs to a fluctuating flow with an average of 5,000
cfs, then the 15,000 cfs constant flow should generate a higher CV
value. Second, CV techniques must perform well relative to each
other. If, for example, two techniques provided comparable results,
but a third technique yields dramatically different results, then the
third technique, all other factors equal, would be judged inferior.
Third, a "good" response rate to a survey is a necessary, although by
no means a sufficient, condition for a sample to adequately represent
the population from which it is drawn. A CV technique which
generates a low response rate relative to other techniques, either
for the survey as a whole for the CV questions themselves (so-called
item nonresponse), would be deemed inferior. The fourth criterion is
particular to the current study. As explained above, respondents
were asked to evaluate several different scenarios, as well as their
actual trips. Thus, if a CV technique must be capable of being
applied to individuals to evaluate several exercises within the same
survey. The fifth criterion involves the degree of complexity of
data analysis associated with a technique, and the complexity of
explaining the technique, data analysis and results to decision
makers. The more difficult the data analysis and/or the harder the
results are to explain for a given technique, the lower this format
of asking the CV question would be rated. Finally, the sixth
criterion deals with the expense of applying a CV technique. The

K.4

consideration here is that if two CV techniques produce comparable
results, yet one technique is considerably cheaper to apply, the less
expensive technique would be preferred.
CV Questioning Formats. Before proceeding it will be helpful to
briefly explain each of the three CV question formats to be compared
here. Such a discussion will facilitate the presentation of results
in the remainder of this appendix and will help to identify the basic
differences between each of the questioning formats.

Iterative-bidding is the oldest and most commonly used CV technique.
The bidding process is conducted in the following manner. As is done
in all contingent-valuation studies, the first step is to describe
the item to be valued and a hypothetical market for trading the item
to a respondent. A traditional bidding application begins when the
interviewer suggests an initial (starting) bid to the respondent. If
the respondent is willing to pay the initial bid, the interviewer
revises the bid upward until a negative response is obtained. A
negative response to the initial bid results in the interviewer
revising the bid downward until an acceptable amount is found. The
final bid is a measure of the respondents' surplus value for the item
in question.
In a study that uses the dichotomous-choice technique, the nonmarket
item to be valued and a hypothetical market for trading the item are
described to the respondent, just as is done for iterative-bidding.
Subsequently, a respondent is asked to state whether he or she is
willing to accept or reject a single take-it-or-leave-it offer for
the item being valued. Respondents are not asked to state a specific
dollar value. Respondents' yes and no responses to the dollar
offers, and the offers themselves, are used to derive estimates of
value.
A technique that has been used in several contingent-valuation
studies is an open-ended questioning format. Open-ended questions
simply entail asking respondents what the maximum is that they would
pay for an item being valued.
Survey Procedures. Individuals in the pretest were randomly assigned
to receive a mail survey with one of the three types of CV
questions. Each respondent was informed of the average flow and
range of fluctuations on the day when the interview was completed.
This information was designed to give respondents a point of
reference in considering the flow scenarios. Two members of the
dichotomous-choice group had to be deleted because their interview
days fall into a gap in our data on flows. Advance letters to all
members of the open-ended group and half of the members of the
bidding game and dichotomous-choice groups were mailed on August 7,
1985. One week later, a questionnaire and cover letter were mailed,

•

K.5
•
followed by a reminder/thank you postcard a few days after that. Two
weeks after the first questionnaire mailing nonrespondents were
mailed a second questionnaire. A third questionnaire and letter were
sent by certified mail in mid-September to remaining nonrespondents.
Responses from the first group were used to initiate bidding for the
iterative-bidding application and the offers for the second half of
the dichotomous-choice group. Dichotomous-choice offers need to
fully cover the range of relevant values. Questionnaires were mailed
to these subsamples on August 27, followed by a reminder/thank you
postcard. The second questionnaire went out on September 14 and the
certified mailing on September 25. The response rates for the mail
questionnaires as a percent of delivered questionnaires was 82
percent for the bidding game group, 77 percent for the
dichotomous-choice group, and 86 percent for the open ended group,
for an overall response rate of 81 percent.
Pretest Survey Results
The estimated surplus values are quite similar for each of the
questioning formats (Table K-1). Pairwise statistical comparisons of
the bidding-games and the open-ended estimates failed to show
significant differences for the actual trip and for each of the
scenarios. The dichotomous-choice values are derived from estimated
logit models, so that direct comparisons of means across valuation
techniques was not possible for these values. However, it is
worthwhile noting that the dichotomous-choice values were well within
the 95 percent confidence intervals for the respective
iterative-bidding means and open-ended means. Thus, there seemed to
be no statistical grounds for rejecting the null hypothesis that, for
the actual trip and each of the scenarios, the three techniques
yielded essentially the same values per trip except for sampling
error.

•

Confidence in this conclusion is bolstered by observing that, with
one minor exception, the three methods give the same rankings for the
actual trip and scenarios: All rank the actual trip first, the
trophy fish scenario second, the "reduce chances of getting skunked"
scenario third, and so on. The exception involves whether the 3,000
cfs scenario is ranked just above or just below the 40,000 cfs
scenario. Given the closeness in the values of these two scenarios
(and the lack of a statistically significant difference in values),
this discrepancy is not like attributable to sampling error rather
than any real differences in values .

Table K-1.

Results f'or Three CV Techniques Used in Pretest

Iterative-Bidding

Open-Ended
(Protest Zeroes Excluded)

Dichotomous- Choice

Value

Rank

Value

Rank

Value

Rank

$124

2

$121

2

$147

2

3,000 cfs Scenario

66

6

76

5

70

5

10,000 cfs Scenario

109

4

99

4

96

4

40,000 cfs Scenario

86

5

64

6

68

6

Double Chances of
Trophy

146

1

128

1

163

1

Reduce Chances of
Catching Nothing

106

3

106

3

100

3

Actual Trip
(averaged 33,000 cfs)

.::-::

ffi

•

...

...

...

K.7

To compare this ranking with preferences, three measures of
preferences will be used. The first has to do with how each
respondent compared the flow scenario with her or his actual trip.
Results for all three subsamples combined are shown in Table K-2.
This table is a little hard to follow at first, but provides some
very relevant information. It says, for example, that 62 percent of
the respondents considered 10,000 cfs superior to the flow during
their actual trip. A total of 53 percent considered 3,000 cfs to be
worse and 47 percent thought 40,000 cfs would be worse. The average
daily flow actually experienced averaged 33,000 cfs, but varied
between 19,000 cfs and 45,000 cfs.
Table K-2.

Percent of Respondents Rating the Scenarios "Better
Than,• "Vorse Than,• and •About the Same As" Their Actual
Trip

3,000 cfs
Scenario

10, 000 cfs
Scenario

40,000 cfs
Scenario

Better

30%

62%

Worse

53

9

47

About the Same

17

29

46

7%

Table K-3 presents additional data on preferences. These data are
from the on-site interviews in the Lee's Ferry Area so there had not
been any opportunity yet for material in the scenarios or elsewhere
in the mail survey to influence people's views. The interview form
asked how the person would rate the water level that day. While 59
percent responded "about right," those who preferred a different flow
indicated that the flow that day was "too high." Furthermore, this
tendency was even more pronounced when respondents were divided into
those who actually experienced less than 30,000 cfs and those that
experienced more than 30,000 cfs. Here 24 percent of the lower flow
group thought the water was too high while the comparable percentage
for those experiencing higher flows was 54 percent .

•

K.8

Table K-3.

Ratings of Water Level During Actual Trip (in percentages
of respondents)

Overall

Those Experiencing
Less Than
30, 000 cfs

Those Experiencing
More Than
30,000 cfs

Too High

36%

24%

54%

About Right

59

69

45

5

7

Too Low

A third bit of evidence comes from questions posed as part of each
scenario asking whether the respondent would take more or fewer trips
under the scenarios than he or she had taken in an average year.
People "vote with their feet," as the saying goes. As we have
already seen in reviewing the history of the Glen Canyon fishery,
angler participation rates can fluctuate widely with changes in
fishing conditions. Thus, while the absolute magnitude of the
numbers may not be terribly reliable, expressions of intention about
number of trips do indicate preferences. Responses to the question
about changes in participation are given in Table K-4. As might be
expected the anglers in our survey were never unanimous about how
their participation would change in response to the scenarios. For
all scenarios, some said they would take fewer trips and some more
trips. The total numbers of fewer trips taken by all embers of the
sample are given in the first column of numbers. Others said they
would take more trips and the total additional trips by all members
of our sample combined are given in the second column of figures.
The net changes in trips for the sample are given in the third
column. Finally, the percentage changes are calculated in the last
column based on the 544 trips respondents reported having taken to
Lee's Ferry in the preceding 12 months.

•

•

Table K-4.

~

.P

l-

Ef:fects o:f Scenarios on Trips Taken In An Average Year

Fewer Trips Taken
by Sample
3,000 cfs

134 trips

More Trips Taken
by Sample
69 trips

Net Change In
Trips Taken
- 65 trips

10,000 cfs

17

169

+152

40,000 cfs

138

47

- 91

Double chances of trophy

21

336

+315

Reduce chances of no fish

66

115

+ 49

*Based on 544 trips reported actually taken in the last 12 months

~

"°

K.10

Most impressive, although not surprising given the attribute survey
results, is the huge increase in intended trips in response to a
doubling of chances to catch a trophy fish. Also, there is a clear
mandate in favor of 10,000 cfs and against 3,000 and 40,000 cfs.
Reducing the chances of getting skunked does not result in a
particularly strong positive reaction.
Though Glen Canyon anglers are by no means unanimous, these three
tables combine to show a definite preferences for flows in the middle
range between say, 10,000 and 20,000 cfs and a tendency to judge
extremely low and high flows as inferior. These results are
consistent with those from the attribute survey.
Returning now to the criterion of consistency between preferences and
CV values, there are both favorable and unfavorable observations to
be made. On the favorable side, all three CV techniques were
consistent with our preference measures in ranking the three flow
scenarios. That is, 10,000 cfs had larger values than either 3,000
cfs or 10,000 cfs. A few of the value differences across the flow
scenarios were not statistically significant, but enough were
significant and the level of consistency across CV techniques was so
great that we are convinced of overall consistency between
preferences and flow scenario values regardless of the CV technique
used. Also favorable were the relatively large value for the trophy
fish question. This result-is consistent with the history of the
fishery, the results of the attribute survey, and the results are
presented in Table K-4.
On the other hand, the actual trip value is problematical. It turns
out consistently higher than the value of the 10,000 cfs scenario and
the scenario where the chances of getting skunked were cut in half.
This is an anomaly since the average actual flow was 33,000 cfs and
getting skunked is definitely a negative attribute. However, this
appears to be a problem with all three techniques. One technique is
not necessarily superior to another in this regard. One plausible
explanation is that people tend to give higher values for actual
experiences than for scenarios. Doubling the chances of catching a
trophy fish is sufficiently attractive to overcome this tendency, but
the relative attractiveness of 10,000 cfs is not large enough to do
so. This same tendency will be somewhat visible in the white water
boater results although statistical testing there will be difficult.
Perhaps people tend to be more conservative across the board when
expressing values for scenarios. The higher value for the actual
trip relative to the getting skunked scenario would also be
consistent with this explanation although here the differences are
not strong enough to show up as statistically significant.
At any rate, our conclusion was that except for anomalies associated
with the actual trip value, the CV techniques were all consistent
with preferences and none was more so than the others.

•

K. 11

Concerning internal consistency among the CV methods -- the second
criterion -- the nearly equivalent rankings of the actual trip and
the scenarios indicates that there too the three CV methods are
performing similarly. An additional test is nevertheless of
interest. Random variables may have similar means yet have very
different distributions. Table K-5 shows statistics for testing
hypotheses that the distribution of bids from the iterative-bidding
and open-ended responses are the same as the distributions of values
implied by the dichotomous-choice responses. It turns out that the
bidding games and dichotomous-choice responses imply statistically
indistinguishable distributions in all cases except the actual trip,
while differences are ·significant for the actual trip and two
scenarios in the case of the open-ended responses. Stated
differently, people's behavior in answering iterative-bidding and
dichotomous-choice questions appears to be quite consistent, except
for the actual trip question, statistically significant differences
in behavior turn up for the actual trip and two scenarios when
dichotomous-choice and open-ended responses are compared. These
results tend to support iterative-bidding and dichotomous-choice
questions over open-ended questions.
Table K-5.

Test Statistics from Test of Hypothesis That Distribution
of Bids in Open-Ended or Iterative-Bidding Formats is
Consistent with the· Estimated Dichotomous-Choice
Distributions

Iterative-Bidding
Actual experience
Low water
Moderate Water
High Water
Doubling Chances
Reducing Chance of Getting Skunked

* Indicates

. 2314 *
.1193
• 1495
• 0935
• 1179
.0859

Open Ended

.2683 *
. 1201 *
.2557
.1539*
.2461
.1428

a significant difference from the dichotomous-choice

model

•

The third criterion is response rate. Here dichotomous-choice
appears to have an advantage. Because the Glen Canyon fishery
attracts anglers from a broad geographic area and because it appeared
to be infeasible to conduct the full survey by personal interview at
Lee's Ferry, the best practical alternative for conducting the

K. 12

bidding games was a combined mail survey and telephone interview. As
noted previously, the combination of nonresponse to the mail survey
and difficulty in re-contacting all mail survey respondents by
telephone made for a relatively low response rate. Further efforts
to contact people by phone might have improved the response rate
somewhat, but we are not optimistic. One alternative would have been
to eliminate the mail survey component altogether and do the entire
survey by telephone. However, to do perhaps ten bidding games and
gather the needed additional inf'ormation would have made for a very
long interview.
The problem with open-ended questions is the number of zero
responses. When zeroes imply that no surplus is present they
represent valid values, but zeroes can also imply an unwillingness to
answer. Such protest responses are a form of item nonresponse.
Little can be done about them other than excluding them entirely from
the analysis. To be sure, some of the same people may simply answer
"no" to dichotomous-choice questions regardless of the offer
amounts. However, we suspect that this is much less prevalent than
zero responses to open-ended questions. Open-ended questions ask
people to- state their maximum values, and this is likely to be much
more difficult and seem much more unrealistic than simply responding
"yes" or "no" to a specific amount.
The CV technique finally chosen for this study needed to be capable
of application to the actual trip and several scenarios. We proposed
that dichotomous-choice is the easiest of the three CV techniques for
respondents to deal with and hence should be the most amenable to
surveys where several CV exercises must be conducted. The pretest
results did not produce any empirical evidence to support or refute
this argument. Respondent fatigue did not appear to affect the
results under any of the three CV techniques. Thus, our endorsement
of dichotomous-choice techniques in this regard remains an
intuitively appealing hypothesis for further research.
Regarding the ease with which results can be analyzed and explained
to decision makers and the public, bidding games and open-ended
questions_have a definite edge. While we have developed computer
programs to analyze dichotomous-choice data and estimate values
easily, the problems with explaining and interpreting the results to
noneconomists become readily apparent in Chapter 3.
On the expense side, the open-ended and dichotomous-choice techniques
have the advantage. Bidding games, if applied in this study, would
require telephone interviews. This would make bidding games
substantially more expensive than the other two techniques.

•

K.13

Conclusion
In the end, the decision was to go with dichotomous-choice in the
angler, white-water boater, and day-use rafter CV surveys.
Dichotomous-choice had shown itself to work at least as well as other
techniques in the research on validity of CV reported in Chapter ~'
particularly in the valuation of Sandhill deer hunting permits. In
the- present study, it gave values-that were roughly equivalent to the
bidding-game and open-ended results that corresponded equally well to
preference rankings. Furthermore, responses to bidding games and
dichotomous-choice questions involving scenarios produced empirical
value distributions that could not be distinguished from each other
statistically. Prospects for a relatively high response rate seemed
better with dichotomous-choice, and dichotomous-choice is cheaper.

L. 1

APPENDIX L
TECHNICAL CONSIDERATIONS FOR DERIVING SURH.US VALUES FROM
RESPONSES TO DICHOTOtl>US-CHOICE VALUATION QUESTIONS
The surplus values presented in the body of this report were derived
from the analyses of respondents',answers to the actual trip and the
scenario contingent-valuation questions.

The procedure used to ask

these valuation questions is commonly referred to as the
"dichotomous-choice" technique.

The application of this technique

involved asking respondents whether they would pay a specific amount
(offer), above and beyond their actual trip expenses, to take their
actual trip and a number of scenarios of plausible Grand Canyon
white-water trips or Glen Canyon fishing trips.
question was used for each of the trips.

A separate valuation

The offers for each of

these valuation questions were randomly assigned to questionnaires
based on the findings frcm the analyses of response to comparable
questions in the CV pretest surveys.
Respondents' answers to the valuation questions were analyzed using
logi t models.

A legit model is a special case of a general group of

models, known as probabilistic models, that can be used to analyze
qualitative response data.

The qualitative responses here are

respondents' "yes" and "no" answers to the valuation questions.

The

general form of a legit model can be specified as
Pr(YES)

= [1

+

exp(f{x))] -1

where Pr(YES) is the probability that a respondent will answer yes to
a specific valuation question, exp indicates exponential notation
(e), and f(x) is a function of variables, including the offer, which
may influence respondents' answers to the valuation questions.

A

separate logit equation is estimated for each valuation question so
the number and types of variables included in the f(x) term may vary

L.2

with the situation being evaluated.

For the current analyses the

following functional form of the f(x) term was used:
n

f(x)

bo +

where bo and the bi are parameters to be estimated, the xi are
that are hypothesized to influence respondents' answers to

va~iables

a valuation question, and n is the number of

variab~es

included in a

specific legit equation (see Hanneman, 1984, for a discussion of the
choice of a functional specification of the f(x) term).
The logit equations were estimated using a maximum likelihood
procedure and the estimated equations were used to calculate expected
surplus values.
Pr(NO)

That is,

1 - Pr(YES)

since yes and no are the only possible answers to the valuation
questions and the two answer categories are mutually exclusive
events.

Since Pr(NO) represents a cumulative distribution function

(c.d.f.), the above equation can be rewritten as
Pr(NO)
where F( ) is a c.d.f., X1 is the offer variable from the valuation
question and the Xj are fixed levels of the other variables in the
legit equation.

Thus, F( ) represents the probability that a typical

respondent will answer no to a valuation question with a specific
offer amount, given certain levels of the other variables (Xj's).

,.

L.3

Surplus values that are conditioned on specified levels of the Xj
variables are calculated from the estimated legit equations as:

- Joo
0

J
00

=

Pr(YES)dx 1

0

where E ( ) denotes expected value and the right-hand side of the
equation is the integral of the appropriate legit equation with the
"
estimated parameters entered in the equation [Pr(YES)] and variables
other than the offer are evaluated at the specified levels.

Rather

than solving this integral explicitly, numerical integration
procedures were used to derive surplus values.

This was done to

simplify the analysis and to facilitate the derivation of surplus
values using a computer.

We did compare some of the surplus values

calculated by numerical approximation with those from the exact
solutions and found that the margin of error was less than 1 percent.
An unconditional surplus value is calculated as

00

E(X )
1

=

l:

j,k

cf
0

Pr(YES)dx Jg.(x.k)
1

J

J

.

where this unconditional surplus value is simply the weighted average
of all of the conditional surplus values calculated in the manner
outlined above, gj( ) are the observed probability distributions of
the Xj variables, and Xjk is the kth value of jth variable.

This

calculation is based on the implicit assumption that all of the Xj
variables have discrete probability distributions.
analyses this assumption will be true.

For the current

M. 1

"

APPENDIX M
ANAL YSF.S OF RF.SPONDENTS' ANSWERS TO THE WHITE-WATER BOATER
CONTINGENT-VALUATION QUF.STIOHS

·ti

Actual Trip
The variables examined in the legit analysis of respondents' answers
to the actual trip valuation question are outlined below.

These

variables are:

x,

= OFFER =

the dollar amount from the actual trip
valuation question;

x2

= EXPENSE =

the amount a respondent spent to take thej_r
actual trip;

x3

= ID TOR =

1 if a respondents' trip was in a motor raft
and 0 i f not;

X4 = PADDLE

=

1 if a respondents' trip was in a paddle raft
and 0 if not;

x5 = DORY =

1 if a respondents' trip was in a dory and 0 if
not;

x6 = KAYAK =

1 if a respondent used a kayak and 0 if not;

x7

= DAYS =

the number of days spent on the river;

x8 = CROWD =

an integer scale, ranging from 1 to 9,
reflecting how crowded a respondent felt the
river was with other boaters during his or her
trip;

x9

1 if a respondent had to walk around a rapid
and 0 if not;

x10

= WALK =
= WATERLVL =

an integer scale, ranging from -1 to 1,
reflecting a respondents' preference for an
optimum flow level relative to what they
actually experienced-(-1 lower, O=same, and
1=higher);

M.2

· X11

= HIKING =

1 if a respondent felt that she or he had
enough time for hiking and seeing
attraction sites, and 0 if not;

x12 = SHARBEACH =

= FEE

1 if a respondent ever had to share a
beach for camping and O if not;
1 if a respondent felt his or her answers
to the valuation questions would affect
the cost of Grand Canyon white water trips
and O if not;

-

x14 = CONFIDENCE =

1 if a respondent was not confident in his
or her answers to the valuation questions
and 0 if not;

x15 = FEEOFFER =

FEE multiplied by OFFER;

X = CONFIDENCEOFFER =
16

CONFIDENCE multi plied by OFFER;

x17 = FLOW =

average flow (in cf s) experienced by a
respondent divided by 1,000; and

x18 = FLOWSQ =

FLOW squared.

Four types of variables were included in the analysis.

Variable X1

is included because it is the dollar amount respondents were asked to
consider in the actual trip valuation question.
from $4 to $1729, with an average of $670.

The offers ranged

The assignment of

specific dollar amounts to the surveys was random.

The second group

of variables, X2 through X12' are characteristics of respondents'
actual trips, each of which may affect how a respondent answered the
valuation question.

The expense variable (X2) can be thought of as

representing the price of a Grand Canyon white water boating trip.
Contingent-valuation data sets are typically examined to identify
responses to the valuation question that are deemed to be invalid.
To address this issue we included

v~riables

X13 through X16·

The

purpose of these variables was to determine whether respondents
concerns about costs and confidence in their answers to the valuation
questions would significantly affect surplus values.

If the

•

M.3

estimated parameters for either of these variables turns out to -be
significant, the variable(s) will be evaluated at a value of zero in
the. computation of surplus values.

This will be done to control for

these types of effects which should not enter into the computation of
surplus values.
The average daily flow levels experienced were modeled as average
flow and average flow squared to account for the fact that surplus
values decline after some optimum flow level.
diffe~ent

We examined several

functional specifications of the flow variables and found

that the specification using flow and flow squared fit the data
best.

We did not model fluctuating flow levels for the actual trip

because only 12 percent of the respondents experienced a daily
fluctuation in excess of 10,000 cfs, and the largest fluctuation
experienced was 16,600 cfs.
We found that only a few of the variables outlined above had a
statistically significant effect on respondents' answers to the
actual trip valuation question.

The variables with significant

parameters for commercial passengers were:
WATERLVL, FEE, FLOW, and FLOWSQ.
with significant parameters were:
FEE, FLOW, and FLOW SQ.

OFFER, EXPENSES,

For private boaters, the variables
OFFER, EXPENSES, CRCWD, SHARBEACH,

Logi t equations which only include these

variables with significant parameters are presented in Table M-1.
Statistical significance is denoted by an asterisk to the upper right
of an estimated parameter.

The constant term (bo) is statistically

different from zero for commercial passengers, but is not for private
boaters.

Variables with significant parameters are interpreted as

having a significant effect on respondents answers to the valuation

•

question and, consequently, will affect calculated surplus values •
Variables which have insignificant parameters, on the other hand, do
not have an effect.

It is important, and interesting, to note that

the type of boat a respondent used did not have a significant effect

M.4
Table H-1.

Variable

Estimated Paralleters for Respondents' Actual Trip Logit

Equa~ions

Equation

Parameter

Ccmmercial Passengers

Private Boaters

3.4505•a
( 1. 6913 )b

0.4188
(2.2630)

•

0.0037
(0.0005)

OFFER

•

0.0052
(0.0013)

•

EXPENSES

-o. 0011 •
(0.0003)

-0.0009
(0.0005)

CRCMD

-------

0.5633
(0.1619)

WATERLVL

0.6691
(0.3134)

SHARBEACH

-------

fF:E

1.3898
(0.2909)

•

•

•

•

-0.3115
(0.1117)

•

N
~/

------0.9737
(0.5030)

•

FL CM SQ

"'

•

2.1228
(0.5300)

•

-0.4118
(0. 1631)

•

0.0047
(0.0018)

0.0070
(0.0030)

320.95c

112.88

303

150

An asterisk denotes significance of the parameter at the 0.10 level.

]21 Numbers in parentheses are asymptotic standard errors.

-cl The Chi-square statistics are used to test the null hypothesis that all of the estimated
parameters in an equation are zero simultaneously. The degrees of freedom are ccmputed by
subtracting one from the nUDber of parameters estimated. The null hypothesis is rejected
i f the reported Chi-square statistic exceeds the table value for the appropriate degrees of
freedom.

•

M. 5

on estimated surplus values.

The estimated logit questions were used

to compute the actual trip surplus values for constant flow levels
plotted in Figure 5-1 of the text.
The omission of insignificant variables did not appear to have
affected the magnitude of the estimated parameters for the included
variables.

The logic for estimating a logit equation that only

includes variables with significant parameters is as follows.

For

survey data, each variable may have some missing observations
associated with it since some respondents do not answer all of the
questions in a survey.

As a result, these missing responses censor

the number of observations that can be used for estimation purposes.
To make the best use of our data for estimating the actual trip logit
equation and

th~

calculation of an actual trip surplus value, we

created a data set for estimation purposes which only included the
variables with statistically significant parameters and consequently
were determined to have a significant effect on respondents' answers
to the actual trip valuation question.

Thus, there would be fewer

observations censored by missing data and we could use more
respondents' answers to the actual trip valuation question.
Most of the parameters in the estimated equation have the expected
signs.

It is important to note that due to the specific functional

form of a logit equation, the signs on the parameters are reversed
from what intuition might lead one to expect, based on a linear
regression model.

That is, one would expect the probability of a yes

response to the valuation question to decline as the magnitude of the
"

offer increases.

For this result to occur in a logit equation, the

-

parameter on the OFFER variable must have a positive sign.
An unexpected finding was that respondents' surplus values increased
with the amount they spent to take their actual trip.

This result

contradicts what economic theory would tell us the sign on this

M.6

variable should be.

That is, the more an individual pays for their

trip the lower should be their surplus values, all other factors
equal.

However, this is not the case here.

We also found that surplus values were significantly lower for
respondents who felt their answers to the valuation questions would
affect the cost of boating in the Grand Canyon.

To control for this

undesirable effect, the FEE variable was evaluated at zero for the
computation of surplus values.
Finally, the optimum flow levels for both groups of respondents are
computed by substituting the estimated legit parameter into the f(x)
term from the logit equation and by taking the first derivative of
[f(x)] with respect to the Fl.OW variable.

This derivative when

evaluated at zero can be used to solve for the optimum flow levels.
The resulting flow is an optimum only if the second derivative of
f(x) with respect to flow is negative.

Scenarios
A separate legit equation was estimated for each of the seven
scenarios, and each of the equations included the same set of
variables.

The variables included in the legit equations for the

scenarios are:

X1
X2
X3
X4
X5
X6

= OFFER;
= EXPENSE;
= FEE;
= CONFIDENCE;
= FEEOFFE R; and
= CONFIDENCEOFFER.

•

M.7

'
All of these variables were defined for the actual trip logi t
equation so we will not repeat those definitions here.

It is

important to note, however, that the distribution of OFFERS varied
across contingent-valuation questions for each of the scenarios.
The estimated legit equations for each of the scenarios are presented
in Table M-2 for commercial passengers and those for private boaters
are presented in Table M-3.

These equations were used to calculate

the scenario surplus values reported in the text.
The reported equations do not include the FEE, FEEOFFER, CONFIDENCE,
and CONFIDENCEOFFER variables as the analysis revealed that the
parameters on these variables were not generally significant.

So, we

used the subset of variables with significant parameters for the
estimated equations reported here..

This is the same procedure we

used for the analysis of the actual trip valuation data to make use
of a larger number of respondents' answers to the valuation ·question.
The parameters for the OFFER variable in Tables M-2 and M-3 are
significant in all of the equations, and the parameters on the
EXPENSE variable are significant in nine of the fourteen equations.
The parameter for the expense variable has the wrong sign in 13 of
the 14 equations, a result we also observed in the estimates for the
actual trip legit equations.

For the equation where the parameter

for the expense variable does have the correct sign, the parameter is
insignificant.

Table M-2.

Estimated Logit Parameters

~or

Scenario Equations -

ec-ercial Passengers

5,000 cfs
Constant
Flow

5 ,000 cfs
W/Fluc

13,000 cfs
Constant
Flow

22, 000 cfs
Constant
Flow

0.6880
(0.4499)

-0.5772
( o. 4320)

·-o.4373
(0.4637)

Variable

Parameter

-------

bo

1.5982•a
(0.5279)b

OFFER

b1

0.0024
(0.0006)

EXPENSE

b2

-0.0007
(0.0003)

-0.0007
(0.0003)

-0.0010
(0.0003)

-0.0013
(0.0003)

257 .15c

295.20

348.59

325d

323

321

-------------

N

•
•

I

0.0036
(0.0006)

•

•

0.0043
(0.0006)

•

•

22,000 cfs
W/Fluc

•

-0.7511
(0.4301)

•

Beaches
Reduced

-0.4896
(0.4333)

0.0659
(0.4094)

•

•

0.0031
(0.0004)

0.0024
(0.0005)

-0.0007
(0.0003)

-0.0004
(0.0003)

-0.0004
(0.0002)

354.95

323.58

341.40

405.33

322

326

326

324

0.0039
(0.0005)

•

0.0041
(0.0005)

40,000 cfs
Constant
Flow

•

•

~/ An asterisk denotes significance of the parameter at the.0.10 level •
..!1.I Numbers in parentheses are asymptotic standard errors.
£/ The Chi-square statistics are used to test the null hypothesis that all of the estimated parameters in an
equation are zero simultaneously. The degrees of freedom are computed by subtracting one from the n1.111ber
of i::erameters estimated. The null hypothesis is rejected if the reported Chi-square statistic exceeds the table
value for the appropriate degrees of freedom.

g/ The sample sizes vary across logit equations because some ·of the respondents did not answer all of the
valuation questions in the survey.

.
,

ti

;:::

ro

'

"'

.

.
Table ~3.

.

(1

Estillated Logit Paraaeters ror Scenario Eqmtions -

5,000 eta
Constant
Fl.ow

5, 000 cfs
WIFl.uc

13,000 eta
Constant
Fl.ow
-1. 7769
(0.4553)

Variable

Parameter

-------

bo

-0.3302
(0.3972)a

-0.2868
(0.3996)

OFFER

b1

0.0054.b
(0. 0011)

0.0050
(0.0010)

EXPENSE

b2

-0.0011
(0.0004)

-0.0010
(0.0005)

134.16c
166d

-------------

N

•

•

•

•

0.0049
(0.0008)

PriYate Boaters

22,000 eta
Constant
Flow

•

-1.8916
(0.5176)

•

22,000 cfs
WIFluc

•

-1.4730
(0.5483)

•

40,000 cfs
Constant
Flow

Beaches
Reduced

•

•

-1. 7424
(0.4585)

-1.9645
(0.4693)

•

•

0.0044
(0.0008)

0.0055
(0.0009)

0.0047
(0.0008)

0.0055
(0.0010)

-0.0011
(0.0005)

-0.0006
(0.0004)

-0.0010
(0.0007)

-0.0003
(0.0004)

0.0001
(0.0004)

154. 17

158.90

171.83

150.33

147.04

164

165

166

164

165

•

•

•

~

175.88
166

P:.I Numbers in parentheses are asymptotic standard errors.
bl
An asterisk denotes significance of the parameter at the 0.10 level.
cl
- The Chi-square statistics are used to test the null hypcthesis that all of the estimated parameters in an equation
are zero simultaneously. The degrees of freedom are computed by subtracting one from the number of parameters
estimated. The null hypcthesis is rejected if the reported Chi-square statistic exceeds the table value for the
appropriate degrees of freedom.

di
- The sample sizes vary across logit equations because some of the respcndents did not answer all of the valuation
questions in the survey.

.:s:

l.l)

M.10

Evaluation of the Effect of Scenario Sequence on Surplus Values
Since we asked a total of eight valuation questions in the survey, we were
concerned that the placement of any specific scenario in the sequence of
white water boating experiences to be evaluated might have affected
respondents' answers to that valuation question.

For example, if the

5,000 cfs constant flow scenario was the first white water boating
experience evaluated and there was a sequence effect, we might expect
respondents to give different answers to the associated valuation
questions than they would if this were the last

exp~rience

evaluated.

However, if there is not an ordering effect, respondents' answers should
be the same regardless of the placement of any specific scenario within
the sequence of scenarios to be evaluated.
To examine the potential for this type of problem we randomly assig]led
respondents to two groups and reversed the order in which these two groups
evaluated the scenarios.

Individuals in both groups were asked an actual

trip valuation question first. ·The sequence in which the scenarios were
presented varied between the two groups.

The exact order in which the

scenarios were presented is shown in Table M-4.

Note that a constant flow

scenario always preceded the corresponding fluctuating flow scenario
regardless of the overall sequence of the scenarios.
Table M-4.

Sequence in Which Scenarios were Evaluated

Ascending Order

Descending Order

Actual Trip
5,000 cfs Constant Flow
5, 000 cfs With Fluctuations
13,000 cfs Constant Flow
22,000 cfs Constant Flow
22,000 cfs With Fluctuations
40,000 cfs Constant Flow
Beaches Reduced

Actual Trip
Beaches Reduced
40,000 cfs Constant Flow
22,000 cfs Constant Flow
22,000 cfs With Fluctuations
13,000 cfs Constant Flow
5,000 cfs Constant Flow
5,000 cfs With Fluctuations

r

-

M. 11 .

To address the issue of whether the sequence of scenarios affected
the surplus values we estimated separate logit equations for each of
these groups for each of the scenarios and statistically tested for
differences between the estimated logit coefficients for the Acending
Order group and the comparable estimates from the Descending Order
group.

The Chi-square statistics for all pairwise comparisons are

reported in Table M-5.

A statistically significant difference was

identified for two of the comparisons for commercial passenger and
three of the private boater scenario comparisons, i.e., the
Chi-square statistics exceed 5.99 implying a significant difference
at the 0. 05 level.

However, there does not appear to be a pattern to

the occurrence of these significant differences and we would conclude
that the sequence i.n which respondents evaluated the scenarios did
not affect the calculated surplus values.
Tab1e M-5.

Comparison of Logit Estimates for Ascending Order and
Descending Order Groups

Scenario

Chi-square Statistics
Commercial
Passengers

5,000 cfs Constant Flow
5,000 cfs With Fluctuations
13,000 cfs Constant Flow
22,000 cfs Constant Flow
22,000 cfs With Fluctuations
40,000 cfs Constant Flow
Beaches Reduced

4. 19a
0.46
5.39*b

8.56
5.91.
12.61
0.37

Private
Boaters
6.16 *
5.41.
11.14*

8.80
4.39
2.61
4.29

E,.I The degrees of freedom corresponding to all of the Chi-square
statistics are 2 •

.QI An asterisk denotes a significant difference at the 0.05 level.

N. 1

APPENDIX B
ABALYSF.S OF RF.SFOBDEBTS' ABSWERS TO THE ANGLER
COBTINGENT-VALUATIOH QUF.STIOHS

Actual Trip
The variables examined in the logit analysis of respondents' answers
to the actual trip valuation question are outlined below.

These

variables are:
X

1

X2

•

= OFFER =

the dollar amount from the actual trip
valuation question;

= EXPENSE =

the amount a respondent spent to take his or
her actual fishing trip;

x3 = GUIDE =

1 if a respondent hired a guide to take them
fishing and 0 if not;

x4 = SEASON =

1 if a respondents' trip occurred during the
winter season (January 1 through April 15 and
October 1 through December 31) and 0 if not;

x5 = DAYS =

the number of days fished during trip;

x6 = BOAT =

1·if a respondent fished from a boat and 0 if
not;

x7

the number of fish a respondent caught;

= FISH =

x8 = SIZE =

1 if a respondent caught a fish larger than 3.0
pounds and 0 if not;

x9 = COMMITMENT =

an integer scale, ranging from 1 to 4,
reflecting respondents' commitment to fishing
in Glen Canyon, with an increase in the integer
reflecting an increase in commitment;

x10 = FEE =

1 if a respondent felt his or her answers to
the valuation questions would affect the cost
of fishing in Glen Canyon and 0 if not;

N.2

x11

= CONFIDENCE =

1 for respondents who indicated that they
were not confident in their answers to the
valuation questions and 0 if not;

x12

= FEEOFFER =

FEE multiplied by OFFER; and

x13

= CONFIDENCEOFFER =

CONFIDENCE multiplied by OFFER.

Five types of variables were included in the analysis.

;

Variable X1

is included because it is the dollar amount respondents were asked to
consider in the actual trip valuation question.

The offers ranged

from $1 to $459, with an average of $156.

The assignment of specific

dollar amounts to the surveys was random.

The second group of

variables, X2 through Xa, are characteristics of respondents'
actual trips, each of which may affect how a respondent answered the
valuation question.

The expense variable can be thought of as

representing the price of a fishing trip.

Variable Xg relates to

respondents' subjective feelings about Glen Canyon angling.
Contingent-valuation data sets are typically examined to identify
responses to the valuation question that are deemed to be invalid.
To address this issue we included variables X10 thru X13 in the
equation.

The purpose of these variables was to determine whether a

concern about costs would result in significantly lower surplus
values or if respondents were not confident of their answers to the
valuation questions.

If the estimated parameters for either of these

variables turns out to be significant, the variable(s) will be
evaluated at a value of zero in the computation of surplus values.
This will be done to control for these types of effects which should
not enter into the computation of surplus values.
The average daily flow levels experienced, and also the amount of
fluctuations in daily flow levels experienced, may also affect
respondents' answers to the actual trip valuation question.
examined several different functional specifications of flow

We

N.3

variables, such as average flow and average flow squared and splits
of the data set between constant flow and fluctuating flow levels to
determine whether the flow levels and fluctuations experienced had a
statistically significant effect on actual trip surplus values.
We found that only a few of the variables outlined above had a
statistically significant effect on respondents' answers to the
actual trip valuation question.
parameters were:

The variables with significant

OFFER, EXPENSES, and FEE.

A legit equation which

only includes these variables with significant parameters is
presented in the first column of Table N-1.

Statistical significance

is denoted by an asterisk to the upper right of an estimated
parameter.

The constant term (bo) is also statistically different

from zero.

Variables with significant parameters are interpreted as

having a significant effect on respondents' answers to the valuation
question and, consequently, will affect calculated surplus values.
Variables which had insignificant parameters, on the other hand, do
not have an effect.

The omission of insignificant variables did not

appear to have affected the magnitude of the estimated parameters for
the included variables.
The logic for estimating a legit equation that only includes
variables with significant parameters is as follows.

For survey

data, each variable may have some missing observations associated
with it since some respondents do not answer all of the questions in
a survey.

As a result, these missing responses censor the number of

observations that can be used for estimation purposes.

'

To make the

best use of our data for estimating the actual trip legit equation
and the calculation of an actual trip surplus value, we created a
data set for estimation purposes which only included the variables
with statistically significant parameters and consequently were
determined to have a significant effect on respondents' answers to
the actual trip valuation question.

Thus, there would be fewer

Table R-1.

Estillated Parameters ror Respondents' Actual Trip Logit Equations

Equation

Parameter

Variable

All Flows

•

Constant Flow

•

-1.8406
(0.6330)

Fluctuating Flow
-0.3416
(0.5939)

bo

-1. 1297 a
(0.4059)b

OFFER

b1

0.0178
(0.0027)

0.0237
(0.0047)

EXPENSES

b2

-0.0052
(0.0012)

-0. 0036
(0.0018)

-0.0074
(0. 0021)

FEE

b10

0.6675
(0. 3932)

o. 3080
(0.5764)

1. 1921
(0.6097)

202.49c

89.27

N

•

•

230

•

•

114

•

0.0130
(0.0036)

•

•

104.69
116

~/An asterisk denotes significance of the parameter at the 0.10 level.
_QI Numbers in ~rentheses are asymptotic standard errors.
s_/ The Chi-square statistics are used to test the null hypothesis that all of the estimated

parameters in an equation are zero simultaneously. The degrees of freedom are computed by
subtracting one from the number of ~rameters estimated. The null hypothesis is rejected
if the reported Chi-square statistic exceeds the table value for the appropriate degrees of
freedom.

•·

""

..

z
~

.,

N.5

observations censored by missing data and we could use more
respondents' answers to the actual trip valuation question.
It is important to note that due to the specific functional form of a
logit equation, the signs on the parameters are reversed from what
intuition might lead one to expect, based on a linear regression
model.

That is, one would expect the probability of a yes response

to the valuation question to decline as the magnitude of the offer
increases.

For this result to occur in a logit equation, the

parameter on the OFFER variable must have a positive sign.
An unexpected finding was that respondents' surplus values increased
with the amount they spent to take their actual Glen Canyon fishing
trip.

This result contradicts what economic theory would tell us the

sign on this variable should be.

That is, the more an individual

pays for their trip the lower should be their surplus values, all
other factors equal.

However, this is not the case here.

We also found that surplus values were significantly lower for
respondents who felt their answers to the valuation questions would
affect the cost of fishing in Glen Canyon.

To control for this

undesirable effect, the FEE variable was evaluated at zero for the
computation of surplus values.
Finally, we found that none of the estimated parameters for the flow
-

variables, for any of the functional specifications examined for
these variables, were statistically different from zero.

That is, we

were not able to make a direct link between the actual trip surplus
value and the flow levels experienced by respondents within the
estimated logit equation.

We believe that this result is due to the

limited variation in flow levels experienced by anglers for their
actual trip.

About one half of the respondents experienced a

constant flow for their trip, while the other half experienced

N.6

fluctuating flows.

Of those who experienced a constant flow level,

none experienced an average flow below 10,000 cfs, the optimum flow
level derived from the analysis of scenarios, and 75 percent
experienced an average daily flow between 20,000 and 30,000 cfs.

In

contrast, only two of the respondents who experienced a fluctuating
flow level were on the river when the average flow exceeded 20,000
cfs.

In fact, 75 percent of these respondents experienced an average

daily flow between 5,000 and 15,000 cfs, and about 50 percent
experienced an average daily flow between 8,000 and 12,000 cfs.
Another complicating factor is that 11 percent of the respondents
hired a guide to take them fishing and, as a result, they did not
have to worry about maneuvering a boat on the river.

One could also

argue that a guide might know how to catch fish regardless of the
flow level so a client would be likely to feel the experience was
acceptable regardless of the flow regime during the time they were on
the river.

We believe that these examples highlight the types of

perturbations in the data that complicated the derivation of a
relationship between average flow levels and actual trip surplus
values.
Even though we were not able to identify a significant relationship
between flow variables and respondents' answers to the actual trip
valuation question, we were able to split respondents into two groups
according to whether they had experienced a constant or fluctuating
flow for their actual trip.
manner.

This split was made in the following

Respondents who experienced daily fluctuations in flows of

less than 10,000 cfs were put in the constant flow group, while all
other respondents were classified as having taken a trip under
conditions of fluctuating flows.

We estimated a separate legit

equation for each of these two groupings of respondents.

The

estimated equations are reported in the second and third columns of
Table N-1 and each equation includes the same set of explanatory
variables as were reported in column one.

We tested the two

N.7

equations to see if they were statistically different and were able
to conclude that there was a significant difference at the 0.01
leve1 •.1/

Since there is a statistical difference according to

whether respondents experienced a constant or fluctuating flow level
for their actual trip, the legit equations in columns two and three
of Table N-1 were used to compute the actual trip surplus values
reported in the text.
Scenarios
A separate legit equation was estimated for each of the nine
scenarios, and each of the equations included the same set of
variables.

The variables included in the legit equations for the

scenarios are:
X1 = OFFER;
X2 = EXPENSE;
X3 = FEE;
X4 = COMMITMENT;
X5 = FEEOFFER;
X6 = COMMITMENTOFFER.

All of these variables were defined for the actual trip legit
equation so we will not repeat those definitions here.

It is

important to note, however, that the distribution of OFFERS varied
across contingent-valuation questions for each of the scenarios.

1/ The Chi-square statistic for this test is 8.53 with three
degrees of freedom, indicating that the null hypothesis of no
difference can be rejected at the 0.10 level.

---------

--~

----

N.8

The estimated logit equations for each of the scenarios are presented
in Table N-2 under the respective scenario headings.

As reported in

the text, we did not find a statistically significant difference in
respondents' answers to valuation questions for the

3,ooq

cfs

constant flow and the 3,000 cfs fluctuating flow scenarios.

Thus, we

only report a logit equation for which respondents answers to these
two scenarios were combined for estimation purposes.

The equations

reported in Table N-2 were used to calculate the constant flow
surplus values reported in the text.
The reported equations do not include the FEE, FEEOFFER, CONFIDENCE,
and CONFIDENCEOFFER variables as the analysis revealed that the
parameters on these variables were not generally significant.

So, we

used the subset of variables with significant parameters for the
estimated equations reported here.

This is the same procedure we

used for the analysis of the actual trip valuation data to make. use
of a larger number of respondents' answers to the valuation question.
The parameters for the OFFER variable in Table N-2 are significant in
all of the equations, and the parameter on the EXPENSE variable is
significant in six of the nine equations.

The parameter for the

EXPENSE variable has the wrong sign in seven of the eight equations,
a result we also observed in the estimates for the actual trip logit
equations.

For the equation where the parameter for the EXPENSE

variable does have the correct sign, the parameter is insignificant.

""

Table H-2.

Variable

.,

~

..,

Est:t.ated Logit Paraaeters for Scenario Equations

Parameter

3,000 cfs
Constant or
W/Fluc

10,000 cfs
Constant
Flow

•

10,000 cfs
W/Fluc

bo

-o.7759•a
(0.2348)b

-0.6594
(0.3358)

OFFER

bl

0.0163 •
(0.0022)

0.0127
(0.0025)

EXPENSE

b2

0.0003
(0.0007)

-0. 0047
(0.0012)

-0.0046
(0. 0011)

471.61c

256.91

446d

225

N

•

•

0. 1750
(0.3414)

•

25,000 cfs
Constant
Flow

•

-0.8841
(0.3247)

•

25,000 cfs
W/Fluc

40, 000 cfs
Constant
Flow

-0.0887
(0.3166)

0.6514
(0.3197)

•

•
•

Bigger
Fish

Skunked

•

-1.1694
(0.3591)

•

-0.0248
(0.3067)

•

0.0142
(0.0029)

0.0167
(0.0027)

0.0139
(0.0028)

0.0091
(0. 0027)

0.0125
(0.0021)

-0.0029
(0.0010)

•

-0.0025 •
(0.0010)

-0.0012
(0.0011)

-0.0024
(0.0011)

-0.0025
(0. 0011)

235.72

243.12

230.38

204.s_o

235.74

222.91

221

226

223

225

220

223

0.0115
(0.0027)

•

•

•

J!/ An asterisk denotes significance of the parameter at the 0.10 level.

~I Numbers in parentheses are asymptotic standard errors.
~I The Chi-square statistics are used to test the null hypothesis that all of the estimated parameters in an equation are zero
s:iJJJultaneously. The degrees of freedom are computed by subtracting one from the number of parameters estimated.
rejected if the reported Chi-square statistic exceeds the table value for the appropriate degrees of freedom •

The null hypothesis is

.QI The sample sizes vary across logit equations because some of the respondents did not answer all of the valuation questions in the

survey.

z
~

...

N.10

Evaluation of the Effect of Scenario Sequence on Surplus Values
Since we asked a total of ten valuation questions in the Glen Canyon
Anglers' Survey, we were concerned that the placement of any specific
scenario in the sequence of angling experiences to be evaluated might
have affected respondents' answers to that valuation question.

For

example, if the 3,000 cfs constant flow scenario was the first
angling experience evaluated and there was a sequence effect, we
might expect respondents to give different answer& to the associated
valuation questions than they would if this angling experience were
evaluated last.

However, if there is not an ordering effect,

respondents' answers

shoul~

be the same regardless of the placement

of any specific scenario within the sequence of scenarios to be
evaluated.
To examine the potential for this type of problem we randomly
assigned respondents to two groups and reversed the order in whi.ch
these two groups evaluated the scenarios.

Individuals in both groups

were asked an actual trip valuation question first.

The sequence in

which the scenarios were presented varied between the two groups.
The exact order in which the scenarios were presented is shown in
Table N-3.

Note that a constant flow scenario always preceded the

corresponding fluctuating flow scenario regardless of the overall
sequence of the scenarios.

N. 11

Table H-3.

Sequence in Which Scenarios were Evaluated

Ascending Order

Descending Order

Actual Trip
3,000 cf s Constant Flow
3,000 cfs With Fluctuations
10,000 cfs Constant Flow
10,000 cfs With Fluctuations
25,000 cfs Constant Flow
25,000 cfs With Fluctuations
40,000 cfs Constant Flow
Bigger Fish
Getting Skunked

Actual Trip
Getting Skunked
Bigger Fish
40,000 cf s Constant Flow
25,000 cfs Constant Flow
25,000 cfs With Fluctuations
10,000 cfs Constant Flow
10,000 cf s With Fluctuations
3,000 cfs Constant Flow
3,000 cfs With Fluctuations

To address the issue of whether the scenario sequence affected
surplus values we estimated separate logit equations for each of
these groups for each of the scenarios and statistically tested for
differences between the estimated logit coefficients for the Normal
Order group and the comparable estimates from the Reversed Order
group.

The Chi-square statistics for all pairwise comparisons are

reported in Table N-4.

A statistically significant difference was

identified for only one of the scenario comparisons.

All of these

statistics, except for the Getting Skunked scenarios, are less than
5.99, indicating that a significant difference does not exist between

,

the estimated equations for these pairwise comparisons at the 0.05
level.

Thus, we would conclude that the sequence in which

respondents evaluated the scenarios did not affect the calculated
surplus values.

N.12

Table H-4.

Comparison or Logit Estimates ror Ascending Order and
Descending Order Groups

Scenario
3,000 cfs Constant Flow
3,000 cfs With Fluctuations
10,000 cfs Constant Flow
10,000 cfs With Fluctuations
25,000 cfs Constant Flow
25,000 cfs With Fluctuations
40,000 cfs Constant Flow
Bigger Fish
Getting Skunked

Chi-square
Statistics

2.18a
5.50
1. 76
0.45
5.85
5.27

5. 19
2.61*b
9.26

'2,.I The degrees of freedom corresponding to all of the Chi-square
statistics are 2 •

.QI An asterick denotes a significant difference at the 0.05 level.

j

0. 1

(

APPENDIX 0
ANALlSF.S OF RESFOBDEBTS' AH~ERS TO THE GLEN CANYON DAY-USE RAFTER
CONTINGENT-VALUATION QUF.STIOH
Actual Trip
The variables examined in the logit analysis of respondents' answers
to the actual trip valuation question are outlined below.

These

variables are:

,

x1

= OFFER =

the dollar a.mount from the actual trip
valuation question;

x2

= EXPENSE =

the a.mount a respondent spent to take their
actual trip;

x3

= CROWD

an integer scale, ranging from 1 to 9,
reflecting hew crowded a respondent felt the
river was with other boaters during his or
her trip;

=

x4 = FEE =

1 if a respondent felt his or her answers to
the valuation questions would affect the
cost of Grand Canyon white water trips and 0
if not;

x5 = CONFIDENCE =

1 if a respondent was not confident in his
or her answers to the valuation questions
and 0 i f not;

x6 = FEEOFFER =

FEE multiplied by OFFER;

x7 = CONFIDENCEOFFER =

CONFIDENCE multiplied by OFFER;

x8 = FLOW =

average flow (in cfs) experienced by a
respondent divided by 1,000; and

x9 = FLOWSQ =

FLOW squared.

All of the variables listed above were previously defined in
Appendices L and M so we will not repeat these definitions here.

0.2

)

As was done in the estimation of the legit equations for white-water
boaters and anglers, we only report an equation which includes
variables with significant parameters.

This was done to make the

best use of our data because some respondents did not answer certain
questions in the survey that were used to comput.e the various
variables in the equation and, in turn, these nonrespondents censored
the sample size used to estimate the first equation.

Thus, by

removing variables with insignificant parameters we increased our
sample size for estimation purposes from 173 to 182.

In turn, we

were able to make use of more respondents answers to the valuation
question.

It is also important to note that the removal of the

variables with insignificant parameters from the equation does not
appear to have affected the magnitude of the estimates of b1 or

b4·
The estimated parameters are presented_ in Table 0-1.

Only the

estimated parameter for the OFFER and FEE variables were
significantly different from zero at the 0.10 level and have the
appropriate signs.1/

An insignificant parameter indicates that the

corresponding variable did not have a statistically significant
effect on respondents' answers to the valuation question.

This

equation was used to calculate the surplus value reported in the
text.

j/ Because of the functional form of a legit equation, the signs on the
parameters are the opposite of what one would expect from a regular
linear regression model. Thus, as the magnitude of the OFFER
increases, the probability of a yes response to
the valuation question will decrease even though the sign of the
parameter for this variable (b ) is positive.
1

j

0.3
I
Table 0-1.

Variable

Estimated Parameters For Respondents' Actual Trip Logit
Equation

Parameter

Equation
2

-------

bo

OFFER

b1

FEE

b4

-------------

-1. 7540

*a

(0.5333)

o.0131*b
(0.0184)
0.6430 *
(0.3274)
218.68c

N

182

~/An asterisk denotes significance at the 0.10 level .

.QI Numbers in parentheses are asymptotic standard errors for the
respective estimated parameters
fl/ The Chi-square statistics are used to test whether all of the

estimated parameters for a single equation are simultan~ously
statistically different from zero. Both of the Chi-square
statistics reported here reveal that we can reject the hypothesis
that all of the estimated parameters in each of the respective
equations are simultaneously equal to zero.

LITERATURE CITED

\

•

Hanneman, ~uchael W. 1984. Welfare Evaluations in Co~tingent
Valuation Experiments With Discrete Responses. American Journal of
Agricultural Economics.· 66: 332-341.

'