SUPPLY AND DEMAND OF “STEM”
OCCUPATIONS IN ARIZONA
A Report from the Productivity and Prosperity Project (P3)

0B

November 2008

1B

Dennis Hoffman, Ph.D.
Professor of Economics, University Economist,
and Director, L. William Seidman Research Institute
Eva Madly, M.S. Economics
Research Economist, L. William Seidman Research Institute
Tom R. Rex, MBA
Associate Director, Center for Competitiveness and Prosperity Research

Center for Competitiveness and Prosperity Research
L. William Seidman Research Institute
W. P. Carey School of Business
Arizona State University
Box 874011
Tempe, Arizona 85287-4011
2B

(480) 965-5362
FAX: (480) 965-5458
EMAIL: Dennis.Hoffman@asu.edu or Tom.Rex@asu.edu
www.wpcarey.asu.edu/seid

INTRODUCTION
The demand for new workers in STEM (science, technology, engineering and mathematics)
occupations in Arizona is a function of net job growth in these occupations in the state and the
number of workers leaving STEM jobs in Arizona. Employees may leave their position for
various reasons, including to retire, to move out of the state, or to change their profession.
The supply of new workers in STEM occupations in Arizona is a function of the number of new
college graduates in related majors, the number of workers moving to the state who are qualified
to fill STEM jobs, and the number of Arizonans who are qualified to fill STEM jobs moving
from a non-STEM to STEM occupation.
Serious data limitations prevent the overall estimation of supply and demand of STEM workers
in Arizona:
• The change in employment in broad categories of STEM occupations is available, but
the figures are estimates based on a survey subject to substantial sampling error. The
definitions of the broad categories of STEM occupations do not coincide with the
definitions of categories of STEM college graduates.
• Use of employment data on individual STEM occupations would allow a better match to
be made to the categories of college graduates, but much of the occupational data are
withheld and those occupational data that are available demonstrate unacceptably large
amounts of survey error.
• The number of college graduates in STEM-related majors is available, but the number
entering the Arizona workforce in STEM occupations is unknown. Some graduates leave
Arizona, some continue their education rather than entering the workforce, and some
take jobs in other occupations.
• In- and out-migration of workers in broad categories of STEM occupations is available,
but the data are old and the figures are estimates based on a survey subject to substantial
sampling error.
• The age profile of STEM workers is available with which to estimate the number of
retiring workers, but the data are old and the figures are estimates based on a survey
subject to substantial sampling error.
• Data on workers who change professions — into or out of STEM occupations — do not
exist for Arizona. Similarly, data on those who leave or re-enter the workforce for other
reasons, such as maternity leave, also do not exist for STEM occupations in Arizona.

1

EMPLOYMENT
Annual estimates of employment and wages by occupation by state are produced from surveys of
employers conducted by the Occupational Employment Statistics program of the U.S.
Department of Labor, Bureau of Labor Statistics (BLS). The standard occupational classification
identifies some 800 occupations aggregated into 23 major occupational groups. For states,
however, much of the data for individual occupations are not released due to limited sample size.
Even when released, sampling error is so large as to make the data for individual occupations
unusable. The accuracy of the data for the 23 occupational groups is better, but sampling error
still is a concern.
The earliest occupational data are for 1999, with the latest for 2007. Sampling error appears to be
substantial in the early years even by occupational group, so data for this analysis have been
restricted to the 2002 through 2007 period. This period does not cover an entire economic cycle,
with 2002 being a recovery year from the 2001 recession and 2007 being the peak of the cycle.
Thus, the annual average change in employment over this period overstates that of an entire
economic cycle. The annual average change in employment shown in Table 1 actually is
calculated on a 4.5 year period, since the occupational surveying was done in November 2002
and in May 2007.
Most STEM occupations are included in one of four occupational groups: (1) computer and
mathematical, (2) architecture and engineering, (3) life, physical, and social science, and (4)
healthcare practitioners and technical. Because of the educational focus of the STEM Education
Center, data for the education, training, and library occupational group also are displayed.
Employment growth has been uneven across these occupational groups. Between 2002 and 2007,
the employment gain in the architecture and engineering group was modest, with the group’s
share of total Arizona employment falling. A modest gain in employment in this occupational
group is consistent with the slump in the state’s high-technology industries. (For more
information on the high-technology industries, see the report at
http://www.azcommerce.com/doclib/prop/hightechstudyupdate.pdf .)

HU

UH

The share of employment in the education occupational group also fell (apparently due mostly to
a decline in the share of jobs in the teacher assistants occupation). Unlike most occupations,
which experience much faster job growth during economic expansions than during economic
slumps, employment gains in education occupations tend to be steadier from year-to-year,
meaning that changes in employment share are countercyclical.
In contrast, employment growth in the other STEM occupational groups was relatively rapid
between 2002 and 2007. Numeric employment growth was much greater in the health care
practitioners group than in the others.

2

TABLE 1
EMPLOYMENT BY OCCUPATIONAL GROUP IN ARIZONA
2007
Occupational Group
Computer and Mathematical
Architecture and Engineering
Life, Physical, and Social Science
Healthcare Practitioners and Technical
Education, Training, and Library
TOTAL, All Occupations

Number
56,630
59,680
18,520
115,360
137,700
2,648,000

Share
2.1%
2.3
0.7
4.4
5.2
100.0

Annual Average
Change, 2002-2007
Number
Share
2,444
2.7%
878
1.0
1,216
1.4
4,902
5.5
1,704
1.9
89,522
100.0

Note: the subcategories do not sum to the occupational group total because of withheld data.
Source: U.S. Department of Labor, Bureau of Labor Statistics.

3

DEGREES
The number of university graduates was obtained from the Integrated Postsecondary Education
Data System (IPEDS), available online from the National Center for Education Statistics
(NCES). Completion numbers (the number of graduates) for all campuses of Arizona’s three
public universities were obtained for academic years 2002-03 through 2006-07, providing five
years of data that match the five years of employment data used. Graduates include those with
bachelor’s, master’s and doctoral degrees, as well as first professional degrees in the field of
health care. Degrees related to STEM fall into several categories, as seen in Table 2.
The number of degrees in the computer and information sciences and support services category
fell from the 2004 peak, but the number of mathematics and statistics degrees rose faster than the
overall number of degrees over the five years. The number of degrees in the engineering
technologies and technicians category fell substantially, while the number of engineering degrees
fluctuated within a narrow range. Engineering as a share of the total degrees slid. Degrees in the
biological and biomedical sciences steadily rose, faster than the rise in overall degrees, as did the
number in the physical sciences and science technologies and technicians categories. The
number of degrees in the health professions and related clinical services also rose considerably.
The number of education degrees rose slowly until a large increase in 2007.

TABLE 2
NUMBER OF GRADUATES AT ARIZONA'S PUBLIC UNIVERSITIES
BY ACADEMIC YEAR AND PROGRAM
Program
Computer and Information Sciences and
Support Services
Mathematics and Statistics
Engineering Technologies and Technicians
Engineering
Biological and Biomedical Sciences
Physical Sciences
Science Technologies and Technicians
Health Professions & Clinical Sciences
Education
TOTAL, All Degrees

2003

2004

2005

2006

2007

Annual
Average

379
140
213
1,601
968
393
62
935
4,569
23,940

418
165
179
1,683
997
407
80
1,085
4,966
25,074

360
153
179
1,582
1,069
453
110
1,185
4,812
25,542

310
172
163
1,587
1,122
456
124
1,205
4,980
25,647

265
200
49
1,651
1,252
523
147
1,224
6,223
26,103

346
166
157
1,621
1,082
446
105
1,127
5,110
25,261

Source: National Center for Education Statistics.

4

MIGRATION
The Census Bureau provides an annual estimate of migration to and from each state, but
information on the characteristics of the migrants historically has been limited to the decennial
census. The American Community Survey, which is replacing the long form of the decennial
census, will provide such information, but the survey error of the limited amount of data
available thus far from the ACS is too large to make use of these data. Thus, the migration data
are seven years older than the employment and degree data.
To estimate the number of workers who migrated into and out of Arizona in the occupational
groups of interest, the 2000 Census Public Use Microdata Sample Files (PUMS) were used. The
number of in-migrants to Arizona from other states or countries between 1995 and 2000 was
obtained from the 5 percent Arizona state sample. The number of out-migrants was obtained
from the 1 percent national sample, by looking at how many people who at the time of the census
resided in any state except Arizona and had lived in Arizona in 1995.
The ratio of in-migrants to out-migrants was particularly high in the health care practitioners
occupational group. Thus, net in-migration of those employed in this occupation was
disproportionately high. The ratio was almost as high in the sciences occupational group, and
was above average in the engineering group. In contrast, in the education and computer groups,
the ratio was near the overall average. These figures from the late 1990s may be different from
the current situation. For example, since the demand for engineers has declined as employment
in some of the high-technology industries has fallen, the number of net migrants in this
occupational group likely has decreased since the late 1990s.

TABLE 3
ARIZONA MIGRATION BETWEEN 1995 AND 2000, BY OCCUPATIONAL GROUP
Annual Average
Occupational Group
Computer and Mathematical
Architecture and Engineering
Life, Physical, and Social Science
Healthcare Practitioners and Technical
Education, Training, and Library
TOTAL, All Employed Migrants

In-Migrants
2,788
2,535
863
4,045
3,917
80,920

Out-Migrants
1,757
1,304
386
1,662
2,323
48,382

Net Migrants
1,031
1,231
478
2,383
1,594
32,538

Ratio of Into OutMigration
1.59
1.95
2.24
2.43
1.69
1.67

Source: U.S. Department of Commerce, Census Bureau, 2000 Census Public Use Microdata Sample.

5

AGE OF WORKFORCE
As with the migration data, the only source of data about the age of the workforce in Arizona is
from the decennial census PUMS. Thus, the age data are several years old and subject to
sampling error.
The focus of the age analysis is to provide insight on the number of workers who might be
retiring soon from STEM occupations. The age of the Arizona workforce in 2000 is summarized
in Table 4. In 2000, the baby-boom generation was aged 36 to 54, accounting for the higher
proportion of workers between age 50 and 54 than between 55 and 59. Presumably, most of
those 60 or older in 2000 already have retired, and some of those who were 55-to-59 years old
likely have retired as well. Those who were between 50 and 54 probably are beginning to retire.
Relative to the entire workforce, those working in the computer and mathematical occupations
are relatively young. Architects and engineers also have a slightly younger than average age
profile. In contrast, those working in the education occupations are older than average. In
addition, a greater than average share of scientists and health care practitioners were 50-to-54
years old in 2000.

TABLE 4
AGE OF ARIZONA WORKFORCE IN 2000, BY OCCUPATIONAL GROUP
Percentage of Workers by Age
Occupational Group
Computer and Mathematical
Architecture and Engineering
Life, Physical, and Social Science
Healthcare Practitioners and Technical
Education, Training, and Library
TOTAL, All Occupations

50 to 54
8.2%
9.0
12.6
12.5
13.8
9.5

55 to 59
4.2%
6.2
5.4
6.6
8.1
6.5

60 or Older
2.9%
5.5
5.3
6.3
6.6
6.8

Source: U.S. Department of Commerce, Census Bureau, 2000 Census Public Use
Microdata Sample.

6

ESTIMATING SUPPLY AND DEMAND
Job growth creates a need for new workers. Those entering the workforce after completing their
education make up one source of new workers. A comparison of these two components of supply
and demand is presented in Table 5. Those moving to Arizona from other states and nations
comprise another source of new workers. These data also are displayed in the table, but are seven
years older than the employment and degree data. Data on other types of movement into and out
of the workforce are lacking.
The data that are displayed in the table indicate that conditions vary substantially by occupational
group. In the education occupations, the number of new degrees is triple the employment change,
and net migration also is larger than employment change. These much larger numbers of degrees
and net in-migrants than employment change suggest that a substantial number of workers are
leaving the education profession, as noted in other studies.
In the healthcare occupations, the number of new degrees in Arizona is far less than the change
in the number of jobs. With the annual employment change accelerating since 2000, the net
migration of workers in this occupational group likely is currently higher than shown in the table.
Still, the figures suggest that employers of health care workers may have a difficult time finding
workers, at least in some occupations. As the baby-boom generation reaches retirement age (the
oldest members of that group currently are 62 years old), the low production of new health care
degrees in Arizona may result in a substantial supply-demand imbalance.
In the science occupational group, the number of degrees exceeds the employment change, and a
modest number of migrants also boosts the supply. In reality, the number of new degrees is
understated relative to the occupational group employment, since the occupational group
includes social sciences, while social science degrees are not counted in the degree totals. It may
be that many individuals obtaining a life or physical science degree go on for additional
education instead of entering the workforce, or that many move out of Arizona, or that many take
jobs in other occupations.
In the architecture and engineering occupational group, the number of degrees greatly exceeds
the employment change. Net migration currently probably is less than during the 1995-to-2000
period since job growth in this occupation has slowed since then.
The situation is the opposite in the computer and mathematical group, with the sum of new
degrees and net migration well below the employment change. However, many of those working
in computer occupations probably have a degree in something other than the NCES category of
computer and information sciences and support services; some likely have an engineering
degree. Thus, the computer/mathematical and architecture/engineering groups have been
combined in the table.

7

TABLE 5
SUMMARY OF DATA BY OCCUPATIONAL GROUP
Occupational Group
Combined Computer and Engineering
Computer and Mathematical
Architecture and Engineering
Life, Physical, and Social Science
Healthcare Practitioners and Technical
Education, Training, and Library

Employment
Change
3,322
2,444
878
1,216
4,902
1,704

New
Degrees
2,290
512
1,778
1,633
1,127
5,110

Difference
-1,032
-1,932
900
417
-3,775
3,406

Net
Migration
2,262
1,031
1,231
478
1,594
2,383

Notes:
Employment change is the annual average from 2002 through 2007.
New degrees is the annual average of 2003 through 2007.
Net migration is the annual average from 1995 to 2000.
Source: U.S. Department of Labor, Bureau of Labor Statistics; National Center for Education Statistics;
U.S. Department of Commerce, Census Bureau, 2000 Census Public Use Microdata Sample.

8