THE LATINO POPULATION IN ARIZONA —
GROWTH, CHARACTERISTICS, AND OUTLOOK —
WITH A FOCUS ON LATINO EDUCATION

September 2011

Tom R. Rex
Associate Director
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
(480) 965-5362
FAX: (480) 965-5458
EMAIL: Tom.Rex@asu.edu
www.wpcarey.asu.edu/seid

	
  

SUMMARY
In 2001, the issue of Latino educational achievement and attainment was raised in the Morrison
Institute report Five Shoes Waiting to Drop on Arizona’s Future. Though the educational
performance and attainment of Hispanics was substantially less than that of non-Hispanic whites
and Asians, it was in line with other minority groups. The special focus on Hispanics was due to
the large and rapidly growing size of the Hispanic population in Arizona.
In terms of educational attainment and achievement, the situation has not changed much in the
last decade. Hispanics continue to make only slow gains relative to non-Hispanic whites and
Asians. Among Hispanics, the relative percentage of high school graduates has increased, but the
relative college graduation rate has not improved.
Demographically, however, significant changes have occurred since 2001. The growth rate of the
Hispanic population has slowed — reducing the magnitude of the Latino education issue — as a
result of two factors. First, the number of undocumented immigrants living in Arizona has
dropped, primarily as a result of the state’s anti-immigration legislation. The long and deep
recession followed by a very slow economic recovery also has caused some undocumented
immigrants to leave the state and has reduced the number of undocumented immigrants crossing
the border. Second, the fertility rate of Hispanics has fallen considerably — by more than that of
any other racial/ethnic group.
Going forward, the magnitude of the Latino education issue in Arizona will depend largely on
the size of the Hispanic population relative to the state’s entire population. The expansion or
restriction of educational programs tailored to students learning English as a second language or
to Hispanic/minority/low-income students generally also would have an impact.
The fertility rate of Hispanics will play an important role in determining the number of Hispanics
living in Arizona, particularly in the number going to school. However, the single largest factor
that will influence Latino educational achievement and attainment in Arizona in coming years is
the number and characteristics of Latino immigrants who will live in the state. Since few changes
have occurred over the last 20 years in the characteristics of Hispanic immigrants and Arizona’s
Hispanic population as a whole, the significance of the Latino education issue in coming years
will largely be determined by the number of immigrants, particularly undocumented immigrants.
Immigration to Arizona in the future is dependent largely on policy decisions made in Arizona,
most notably in the enforcement of the state’s employer sanctions law. Certain changes in federal
immigration law also would have an effect. Economic and demographic conditions, typically the
most important factors influencing immigration — particularly undocumented immigration —
may also play a role, but the presence of Arizona’s employer sanctions law likely will relegate
them to secondary importance.
Changes in the age distribution in both the United States and Mexico almost certainly will result
in less undocumented immigration in coming decades. Economic conditions in countries from
which the United States receives immigrants (mostly Mexico), in the United States, and
specifically in Arizona could have an effect on the number of immigrants settling in Arizona.

	
  

1

INTRODUCTION
The primary purpose of this paper is to examine the educational achievement and attainment of
Latinos living in Arizona. (The terms Latino and Hispanic are used interchangeably in this
report.) In short, the educational performance and attainment of Hispanics lags considerably
behind that of the state’s non-Hispanic white residents. Moreover, the growth in the number of
Hispanics, especially children, has far outpaced that of the non-Hispanic white population,
raising the importance of this educational gap. This issue was highlighted in the 2001 Morrison
Institute report Five Shoes Waiting to Drop on Arizona’s Future.
The specific goal of this report is to assess the future of Hispanic educational performance and
attainment in Arizona, but this outlook is dependent to a very sizable extent on the characteristics
of the Hispanics living in Arizona. In particular, the educational achievement and attainment of
Hispanic immigrants has been substantially less than that of Hispanics born in the United States.
Forecasting Hispanic immigration specifically and the characteristics of the Hispanic population
generally is extremely difficult for two reasons: severe data limitations, and recent legal and
policy changes related to immigration.
To make forecasts using any standard modeling technique requires a long time series of
consistent data on a range of related indicators. However, demographic and socioeconomic data
for Hispanics are largely limited to a data point every 10 years from the decennial census. Solid
data on undocumented immigrants is virtually nonexistent, yet it appears that this component of
the Hispanic population became substantially more significant during the 1990s.
Even if more and higher quality data existed, the recent legal and policy changes in Arizona
related to immigration, and the potential for additional action and for legal challenges to affect
legislation that has been passed, would render the historical data to be of limited use. Rather than
economic and demographic conditions in the United States and in the countries supplying the
immigrants (mostly Mexico) being the most important factors to consider in making projections
of the number of Hispanic immigrants to Arizona, future immigration is more likely to be
affected by immigrant-related laws passed by the Arizona Legislature since 2007.
The key legislation is the “employer sanctions law” passed in July 2007 and implemented at the
beginning of 2008. The number of undocumented immigrants living in Arizona appears to have
dropped in 2007 and 2008 in direct response to this legislation. The other significant piece of
legislation is “Senate Bill 1070” that was passed in 2010. Court challenges have blocked
significant parts of this legislation from being implemented, but a number of cases remain open
in the court system.
Data Limitations
The population count from the decennial census is the only source of accurate figures on the
number of residents at any geographic level, including the nation and states. Annual estimates of
the total population in years between the decennial censuses are produced by the U.S. Census
Bureau and by other organizations, including the Arizona Department of Administration’s Office
of Employment and Population Statistics. (This unit formerly was within the Department of
Economic Security and, most recently, the Department of Commerce.) The accuracy of these
estimates varies widely over time and by geographic area. Generally, estimates are most accurate

	
  

2

for large population centers and least accurate for lightly populated areas. The Census Bureau’s
2010 population estimate for Arizona was much higher than the decennial census count, with the
overestimate being the largest in the nation on a percentage basis. Estimates from other sources
also were substantially too high.
While the Census Bureau will revise its annual intercensal estimates from 2000 to 2010 to match
the 2000 and 2010 census results, those revisions are not yet available. Further, since the Census
Bureau does this for every state and county in the nation, it must use a simplistic approach that
will not consider, for example, the effects of the implementation of the state’s employer
sanctions law in 2007 and 2008 on the number of Hispanics living in Arizona.
Beyond the overall population, counts are available from the 2010 decennial census by gender,
race/ethnicity (including Hispanic), age, and household relationship. The annual population
estimates produced by the Census Bureau are subdivided by gender, age, and race/ethnicity, but
these figures are highly derived. For example, in addition to the overestimate of the total
population, the Hispanic share of the Arizona population was overestimated compared to the
2010 decennial census results.
Other demographic and socioeconomic information (such as educational attainment) was
available from decennial censuses prior to 2010. These measures were derived from responses to
the long form of the decennial census questionnaire that was sent to only about one-in-six
households. Thus, sampling error was a significant issue when looking at small geographic areas
or at small subsets of the population in more populous geographic areas (such as the number of
Hispanic immigrants in Arizona).
The Current Population Survey (CPS) has for decades provided annual estimates of most of the
demographic indicators reported by the decennial census, but the CPS was designed to be
accurate only for the nation and four broad regions of the country. Another survey, the American
Community Survey (ACS), was launched nationwide in 2005 to provide demographic and
socioeconomic data for states and smaller geographic areas. In order to fund the ACS, the longform questionnaire of the decennial census was abandoned — such data are not available from
the 2010 census.
American Community Survey
The ACS has the advantage of providing ongoing data updates (annually), rather than an update
every 10 years. However, survey error in the ACS is very large. Except for very populous areas,
only the broadest indicators can be considered reliable using the annual data. As discussed
below, a comparison of the results of the 2009 ACS and the 2010 census suggests that even the
estimates/counts of the broadest indicators — such as gender — are noticeably different for
Arizona.
Three-year and five-year averages, which have less sampling error, also are produced from the
ACS. However, even for the five-year average, sampling error is larger than the error from the
long form of the earlier decennial censuses. Moreover, multiyear averages from the ACS are
more difficult to interpret than the point-in-time estimates from the decennial censuses. For
example, the first two years of the latest five-year average for 2005 through 2009 were part of

	
  

3

the real estate/economic boom while the last two years were marked by the deepest and longest
recession since the 1930s.
Most of the survey results from the ACS and from earlier decennial censuses are expressed in
two ways: number of people and percentage of the total. In the decennial censuses prior to 2010,
the number of people in a given category (such as the number of people speaking Spanish at
home) was calculated from the percentage of those surveyed and the overall population count.
The ACS has no population count to use. Instead, the Census Bureau’s annual population
estimates are used as the control for the ACS. However, since the Census Bureau’s population
estimates for Arizona in recent years were substantially too high, the population estimates by
category published in the ACS have significant error beyond that of the survey error. Thus, only
percentages from the ACS should be used.
Beyond using the margin of error published by the Census Bureau for every ACS estimate, a
sense of the reasonableness of the ACS results can be obtained by comparing the 2009 (single
year) ACS figures to the 2010 census results for those few indicators that are available from the
2010 census. (ACS results for 2010 are not yet available.) The 2009 ACS data were collected
throughout the calendar year and thus on average represent a period nine months earlier than the
decennial census data that are expressed as of April 1, 2010. It is unlikely that during these nine
months significant demographic changes would have occurred in Arizona’s population. Thus, it
is assumed that most of the differential in results between the 2009 ACS and the 2010 census are
due to sampling error in the ACS.
The comparison reveals that non-Hispanic men (relative to non-Hispanic women) are marginally
underrepresented in the 2009 ACS, but Hispanic men are overreported. By age, the ACS
underportrays non-Hispanics between the ages of 55 and 74 (relative to other ages) and slightly
overstates the percentage in most other age brackets. Among Hispanics, the ACS underdepicts
the percentages in the 45-to-74 and 15-to-19 age groups. Those 25-to-34 years old and children
less than 5 years old are overrepresented.
By race, the ACS results are close to those of the census among non-Hispanics. Among
Hispanics, however, the ACS greatly overreports whites, offset by a much lower proportion in
the “other” race category. This is inconsistent not only with the 2010 census, but with earlier
censuses.
The ACS overstates the proportion of households headed by someone between the ages of 25
and 44, especially for Hispanics. Very young and old householders are underdepicted in the
ACS. The sampled households in the ACS had more children and more other relatives (such as
parent or sibling of the householder) than did the count of households in the decennial census.
Average household size was very substantially overstated in the ACS. Because of these
significant differences between the 2009 ACS and 2010 census results, considerable caution
should be exercised in interpreting the results from the 2005-09 ACS that are included later in
this report.

	
  

4

Effects of Legal and Policy Changes
The passage of immigration-related legislation by the Arizona Legislature may have permanently
changed the pattern of growth in the Hispanic population in Arizona beginning in late 2007.
Several factors contribute to the uncertainty of this statement. While the employer sanctions law
has passed legal muster, if it is not vigorously enforced some employers may hire undocumented
workers, allowing illegal immigration to the state to occur. Other uncertainties involve pending
court decisions, subsequent legislation, and actions taken by other states and/or by the U.S.
Congress.
It is difficult to quantify the effects of the legislation passed so far. The only “symptomatic”
indicators that are available by race/ethnicity are births and school enrollment, but the latter
series underwent a methodological change in the key year when the employer sanctions law was
implemented and otherwise appears to include inaccuracies. The birth and enrollment data are
analyzed in a later section of this report.
A further complication to understanding the effects of the Arizona legislation is that the Arizona
and national economies fell into a deep recession at the same time that the employer sanctions
law went into effect. When Senate Bill 1070 went into effect, the Arizona economy had barely
begun to recover. Conceptually, it is possible to disentangle the effects of the legislation from the
economic recession on the number of Hispanics living in Arizona, but the quality of the data
needed to perform such an analysis is limited. Such a study, performed by the Public Policy
Institute of California, is reviewed in this report.

	
  

5

GROWTH OF THE HISPANIC POPULATION IN ARIZONA
The growth of the Arizona population over the last three decades is shown in Table 1 by
ethnicity. Numeric population growth was less between 2000 and 2010 than in the preceding
decade, almost entirely due to a slowing of non-Hispanic growth. The Hispanic share of the
population did not rise as much between 2000 and 2010 as during the 1990s.
Population change consists of net natural increase — the number of births less the number of
deaths — plus net migration. Total net migration consists of net domestic migration (migration
to and from elsewhere in the United States) plus net international migration (immigration less
emigration).
While an accurate count of the number of births and deaths is available by race/ethnicity, counts
of the other components of population growth do not exist. Total net migration between two
decennial censuses can be calculated by subtracting net natural increase from the change in the
population, but it is not possible to derive the number of domestic versus international migrants.
Net migration from state to state within the United States is available annually from the Internal
Revenue Service (IRS), but no demographic detail, such as race/ethnicity or age, is available.
Further, these IRS figures are not complete counts and are not available promptly.
The number of legal immigrants is reported annually by state by the U.S. Department of
Homeland Security’s Office of Immigration Statistics, but in many cases the immigrants entered
the country earlier than the year in which they are reported (for example, by first entering on a
temporary visa). Estimates from various sources of the number of undocumented immigrants by
state are available, but these figures are derived from surveys with a relatively small sample size.
Further, the estimated numbers by state are not subdivided by ethnicity.
Based on their estimated numbers, undocumented immigrants accounted for around 70 percent
of Arizona’s total immigrants during the 1990s. The undocumented share was smaller from 2000
through 2007, but still exceeded 50 percent. Since then, Arizona probably has experienced a net
outflow of undocumented immigrants.
TABLE 1
NUMBER OF ARIZONA RESIDENTS BY ETHNICITY
1980
1990
2000
2010
Change:
1980-90
1990-2000
2000-10

Total
2,718,215
3,665,228
5,130,632
6,392,017
947,013
1,465,404
1,261,385

Non-Hispanic
2,277,514
2,976,890
3,835,015
4,496,868
699,376
858,125
661,853

Hispanic
440,701
688,338
1,295,617
1,895,149
247,637
607,279
599,532

Source: U.S. Department of Commerce, Census Bureau, decennial censuses.

	
  

6

Hispanic Share
16.2%
18.8
25.3
29.6
2.6
6.5
4.3

Symptomatic Indicators
Because of the unavailability of most components of population growth, in order to make
intercensal estimates of the population, a variety of “symptomatic” indicators are used. These
include building permits, school enrollment, driver licenses, and Medicare and Social Security
enrollments. Births and deaths are sometimes used as indicators of broader population change.
Of these indicators, the only ones that provide information by state by race/ethnicity are births
and deaths and school enrollment.
School Enrollment
Operationally, school enrollment figures are primarily used to estimate the size of the school-age
population. However, changes in enrollment from a given grade in one year to the next grade in
the following year provide insight into net migration, not only of children, but of their parents.
Public school enrollment figures, as of October 1, are reported annually by the Arizona
Department of Education (ADE). Prior to 2008, a student enrolled in more than one school —
primarily high school students taking classes at their regular high school and at a vocational
school — was counted more than once. Since then, a student is counted only once. Since the
ADE has not published the counts for 2008 under both methodologies, a discontinuity in the time
series exists in a key year — when the employer sanctions law went into effect.
In addition to this methodological change, questions regarding the accuracy of the enrollment
figures have been present for years. An examination of annual enrollment counts by county and
by grade reveals numerous instances of unexpected changes. There are a number of cases of an
unusually large increase in enrollment in one year followed by an unusually small increase, or a
decrease, in the following year, which suggests that enrollment was overstated in the first year.
The opposite also occurs, suggesting that the enrollment figure in a given year was understated.
In other cases, it is impossible to evaluate the likelihood of an unexpected change of being
accurate.
Another limitation of using the school enrollment figures as an indicator of population change is
that the counts are limited to public schools. If the popularity of private schools and home
schooling changes over time, then the public enrollment figures are less indicative of the change
in the school-age population. Taken together, the change in methodology, the questions
regarding accuracy, and the limitation to public schools greatly reduce the usefulness of this
school enrollment dataset as an indicator of the change in the Hispanic population.
From 2000 through 2006, the increase in Hispanic enrollment ranged from 3.4-to-6.4 percentage
points per year higher than that of non-Hispanics (see Chart 1). In the four years since then, the
difference has ranged from 0.0-to-3.1 percentage points. (The 2008 figures are based on an
estimate of 2008 enrollment under the old methodology.) The slowdown in Hispanic growth
relative to other groups begins in 2007 and precedes the implementation of the employer
sanctions law on January 1, 2008. However, the law was passed in July 2007. Anecdotal
evidence suggests that some undocumented Hispanic immigrants left the state before the law
took effect — before the school year started — rather than make their children change schools
during the school year.

	
  

7

CHART 1
PUBLIC SCHOOL ENROLLMENT IN ARIZONA, ANNUAL PERCENT CHANGE

	
  
* Estimated
Source: Arizona Department of Education, Research and Evaluation Section.

Without being able to precisely state how the enrollment of Hispanics changed once legislation
targeting undocumented immigrants began to be passed, the enrollment data strongly suggest
that the growth in number of school-age Hispanics has slowed substantially.
Births
The final count of births reported by the Arizona Department of Health Services is believed to be
highly accurate. Annual data are available for several decades. The numbers are reported by the
race/ethnicity and age of the mother.
Not only do births represent a component of population change, changes in the number of births
can be used to estimate changes in the size of the broader population. Even if a long-term trend is
present, fertility rates — the number of births relative to the number of women of child-bearing
age — typically do not change very much from year to year. Generally, short-term fluctuations
are related to the economic cycle, since some people postpone family additions during periods of
economic weakness. Thus, an increase or decrease in the number of births that cannot be
explained by economic conditions may signal a change in the population size.
In order to calculate fertility rates, data must be available on the number of females of childbearing age. Since the overall fertility rate is affected by the age distribution of women —
fertility rates vary widely by age of the mother — it is more useful to examine fertility rates by
age of the mother. Thus, at a state or local level, accurate fertility rates can only be calculated
based on decennial census counts.
More commonly reported than the fertility rate is the crude birth rate, which divides the number
of births by an estimate of the entire population size. Crude birth rates are very high for
	
  

8

Hispanics in Arizona, but in large part this is due to such a high percentage of the Hispanic
population being of child-bearing age.
The fertility rates displayed in Table 2 use the census counts in 2010 of the number of women of
child-bearing age and the number of births in 2009 (births by age and race/ethnicity have not yet
been published for 2010), The rates are expressed as the number of births per 1,000 women in
each age group.
The Hispanic fertility rate is higher than the overall rate of all races/ethnicities for women under
the age of 30 and approximately equal to the overall rate among women 30 or older. In each age
group under the age of 30, the Hispanic fertility rate is higher than that of non-Hispanic whites
and non-Hispanic Asians, but is lower than that of non-Hispanic Native Americans.
Fertility rates fell between 2000 and 2010 among women under the age of 30. Fertility rates fell
particularly among Hispanics — in every age group they had the largest decline or smallest
increase in fertility rate among the race/ethnic groups.
TABLE 2
FERTILITY RATES IN ARIZONA BY AGE AND RACE/ETHNICITY OF MOTHER
(Number of Births Per 1,000 Women)
Age of
Mother
Total
Hispanic
2009/2010*
Total
54.6
65.4
10-14
0.6
0.9
15-17
26.6
41.6
18-19
79.2
113.4
20-24
112.1
143.2
25-29
123.3
138.9
30-34
96.5
95.9
35-39
45.9
47.6
40-44
10.0
11.0
45-49
0.8
0.7
Change Between 2000 and 2009/10
Total
-4.5
-19.9
10-14
-0.6
-1.4
15-17
-14.5
-34.8
18-19
-26.5
-48.7
20-24
-27.0
-48.8
25-29
-5.8
-17.6
30-34
4.1
-1.1
35-39
5.4
4.6
40-44
1.6
0.5
45-49
0.3
0.1

White

Non-Hispanic
Black
Am Indian

45.9
0.2
11.3
47.6
86.3
113.4
96.5
43.0
8.7
0.7

63.5
0.7
30.9
99.3
152.7
133.2
93.5
49.0
9.8
1.2

77.7
1.5
54.0
147.6
173.8
151.5
101.7
54.2
13.7
0.6

57.3
0.4
5.7
29.5
61.9
123.3
129.4
62.9
15.6
1.1

0.6
-0.2
-6.8
-18.5
-15.3
1.9
8.1
5.5
1.6
0.3

8.5
-0.9
-16.1
-32.0
0.9
30.6
27.4
15.8
2.0
0.8

10.6
0.2
3.7
10.2
2.7
15.0
9.3
6.1
2.6
0.2

8.3
0.1
-5.3
-5.3
6.3
16.7
28.2
10.7
3.2
0.4

Asian**

* Births are from calendar year 2009; the number of women is as of April 1, 2010
** Including Pacific Islanders
Sources: U.S. Department of Commerce, Census Bureau, decennial censuses (number of women) and
Arizona Department of Health Services, Bureau of Public Health Statistics (number of births).

	
  

9

Though accurate fertility rates cannot be calculated annually, the annual time series of the
number of births provides insight into when during the decade the decrease in fertility rates
occurred. The annual percentage change in the number of births by racial/ethnic group is shown
in the top half of Table 3. The fluctuations in the overall percentage change primarily are related
to the economic cycle. The number of births in 2001 into 2002 rose little due to the 2001
recession. The deep recession that began in late 2007 and ran through 2009 is largely responsible
for decreases in the number of births across racial/ethnic groups from 2008 through 2010.
From 2003 through 2008, the number of births to Hispanics was greater than the number to nonHispanic whites. Hispanics accounted for approximately 44 percent of the total during this
period. In most years, even before 2008, the percentage increase in the number of births to blacks
and Asians was greater than that of Hispanics. However, these racial groups are relatively small
in Arizona. In 2010, Blacks accounted for only 5 percent, and Asians less than 4 percent, of the
total population.
In the bottom half of Table 3, the percentages by race/ethnicity are expressed relative to the
overall percent change. The number of births to Hispanic mothers rose roughly equal to or faster
TABLE 3
NUMBER OF BIRTHS IN ARIZONA BY RACE/ETHNICITY, PERCENT CHANGE

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010

Total

Hispanic

5.6%
0.3
2.5
3.9
2.9
2.6
6.5
0.6
-3.4
-6.7
-6.0

7.5%
3.6
2.8
8.3
4.4
3.3
6.4
1.9
-6.8
-10.0
-10.5
1.9
3.3
0.3
4.4
1.5
0.7
-0.1
1.3
-3.4
-3.3
-4.5

Non-Hispanic
Native
White
Black
American
Percentage Change
3.5%
2.3%
5.8%
-2.3
-0.9
-4.7
-0.3
0.6
3.8
0.1
9.8
6.1
1.1
6.4
4.1
1.0
7.3
2.8
8.5
12.0
1.1
-1.9
7.8
0.7
-0.7
3.3
-0.7
-5.1
1.8
-3.0
-2.5
-1.1
-5.7
Relative to the Total
-2.1
-3.3
0.2
-2.6
-1.2
-5.0
-2.8
-1.9
1.3
-3.8
5.9
2.2
-1.8
3.5
1.2
-1.6
4.7
0.2
2.0
5.5
-5.4
-2.5
7.2
0.1
2.7
6.7
2.7
1.6
8.5
3.7
3.5
4.9
0.3

Source: Arizona Department of Health Services, Bureau of Public Health Statistics.

	
  

10

Asian
8.9%
3.1
8.4
10.7
3.6
7.3
11.8
8.8
0.4
-0.8
-3.0
3.3
2.8
5.9
6.8
0.7
4.7
5.3
8.2
3.8
5.9
3.0

than the overall figure in each year from 2000 through 2007, with the faster growth especially
apparent in 2001 and 2003. Beginning in 2008, however, the percent change in the number of
Hispanic births was significantly less than the overall figure. Part of this differentially large
decrease almost certainly results from a reduction in the number of Hispanic women of childbearing age due to the passage of the employer sanctions law.
The large decreases in the number of births to Hispanic women after 2007 also are due to
declines in fertility rates. It appears that the disproportionate decline in the fertility rate of
Hispanics between 2000 and 2010 did not begin until 2007. Annual data from the ACS were
examined to provide insight into the decrease in fertility rates, but the results are subject to
considerable sampling error. The analysis compared Hispanics to non-Hispanics, with a
distinction made by the ability to speak English. Between 2006 and 2009, the fertility rate of
non-Hispanics was unchanged at 2.0. The fertility rate of Hispanics who speak English well
decreased slightly from 2.2 to 2.1. In contrast, the fertility rate of Hispanics who do not speak
English well dropped substantially, from 3.8 to 3.1. This suggests that a decrease in fertility
occurred among undocumented immigrants who chose to remain in the state after the employer
sanctions law went into effect.
Deaths
As with births, deaths are reported by race/ethnicity and age and are believed to be highly
accurate. Annual data are available for several decades. Mortality rates change only gradually
over time and are largely unaffected by economic conditions. Since mortality rates are extremely
related to age, an increase or decrease in the number of deaths largely reflects a change in the
size of the elderly population, which may be independent from the change in the younger
population.
Mortality rates of Hispanics are only a little higher than those of non-Hispanic whites and are
lower than those of Blacks and Native Americans in most age groups. Since relatively few
Hispanics in Arizona are elderly, Hispanics accounted for only 11-to-12 percent of the total
number of deaths over the last decade — compared to 39-to-44 percent of the births.
Annual percentage increases in the number of Hispanic deaths were lower in the second half of
the last decade than in earlier years. This slowing also occurred among non-Hispanic whites and
suggests lesser increases in the number of very elderly living in Arizona.
Components of Hispanic Population Change
Net Natural Increase and Total Net Migration
Between the 1980 and 1990 censuses, the number of Hispanics living in Arizona rose by about
248,000 (see Chart 2). With approximately 114,000 more births than deaths (net natural
increase), net migration (domestic migration plus legal immigration plus undocumented
immigration) was around 134,000. Net migration was 54 percent of the total population change.
The numeric growth of Hispanics in Arizona soared to about 607,000 between 1990 and 2000.
Net natural increase was up to 215,000 (89 percent higher than in the preceding decade) and net

	
  

11

CHART 2
POPULATION CHANGE OF HISPANICS IN ARIZONA
(In Thousands)

	
  
Source: U.S. Department of Commerce, Census Bureau, decennial censuses (total change) and Arizona
Department of Health Services, Bureau of Public Health Statistics (net natural increase).

migration soared 193 percent to 392,000. Net migration accounted for 65 percent of the total
population change.
The population change of Hispanics in Arizona between 2000 and 2010 barely slipped to around
600,000. Net natural increase continued to rise to about 349,000, but net migration fell 36
percent to approximately 251,000 — only 42 percent of the total population change.
The net migration of Hispanics to Arizona between 2000 and 2010 consisted largely of children
and young adults: 58 percent of all migrants were between 10 and 29 years old in 2010. On
average, at the time the migration occurred, this group was between 5 and 24 years of age.
Among non-Hispanics, only 28 percent of the migrants were in this age group.
Domestic Migration
The decennial census through 2000 and now the ACS is the only source of information regarding
the migration of Hispanics within the United States. The decennial census asked for place of
residence five years earlier. Since the ACS version of the question asks for residence one year
earlier, the census and ACS results are not directly comparable.
In 1990, only 9 percent of Hispanics had moved to Arizona from another U.S. state in the
preceding five years, a proportion less than half that of non-Hispanics (21 percent). In 2000, the
Hispanic proportion moving from another state rose to 10 percent while the non-Hispanic share
dropped to 19 percent. The same general relationship is seen in the 2005-09 ACS data, with 2.7
percent of Hispanics and 4.9 percent of non-Hispanics moving to Arizona from another U.S.
state in the prior 12 months.
	
  

12

Hispanics were divided into three citizenship categories in the analyses made from the decennial
census and ACS data: born in the United States, naturalized citizens, and noncitizens. Those
Hispanics born in the United States had a slightly higher rate of migration from another state
than all Hispanics. Noncitizens had a slightly lower rate.
Immigration
Immigration, particularly undocumented, has been a significant factor in the growth of the
Hispanic population in Arizona (and in the nation). Information on those legally admitted to the
United States is available from the U.S. Department of Homeland Security’s Office of
Immigration Statistics (OIS). However, only an estimate of the number of undocumented
immigrants is available from the OIS; other groups also produce estimates, as seen in Table 4.
None of these estimates of undocumented immigration distinguish between Hispanics and nonHispanics. However, immigrants from Mexico are estimated to make up around 60 percent of the
undocumented immigrants to the United States (compared to about 15 percent of legal
immigrants). Since the combined share of undocumented immigrants from Mexico and other
Latin American countries is estimated to be greater than 75 percent, it can be assumed that a
similar share of undocumented immigrants are Hispanic.
The numbers of immigrants in Table 4 are expressed as an annual or annual average change. The
number of legal immigrants includes both new arrivals and those already in the country who
received an adjustment from a visa to permanent residency. These adjustments account for about
TABLE 4
ESTIMATED NUMBER OF IMMIGRANTS TO ARIZONA AND THE UNITED STATES
(In Thousands)
Legal (OIS)

OIS

Undocumented
Pew

PPIC

ARIZONA
Average, 1992-2000
Average, 2000-05
Average, 2006-07
2008
2009
2010

9
17
20
21
21

29
30
25
30
-100
10

21
30
25
-25
-100
25

21
22
15
1

UNITED STATES
Average, 1992-2000
Average, 2000-05
Average, 2006-07
2008
2009
2010

777
980
1,159
1,107
1,130

574
406
645
-180
-850
40

490
540
450
-400
-500
100

510
470
180
-58

Sources: OIS: U.S. Department of Homeland Security, Office of Immigration Statistics, Yearbook of
Immigration Statistics (various years).
Pew: Pew Hispanic Center (various reports).
PPIC: Public Policy Institute of California, Unauthorized Immigrants in California: Estimates for Counties
(Technical Appendix).

	
  

13

60 percent of the total, so the year represents the year that individuals received permanent
residency instead of the year in which they first entered the United States. Nearly two-thirds of
the legal immigrants are either immediate relatives of U.S. citizens or are sponsored by family
members.
The numbers of undocumented immigrants in Table 4 are derived from the estimates of the total
number of undocumented immigrants living in a state by year. In 2010, undocumented
immigrants were estimated to account for 3.5 percent of the national population. The share was
twice as high in Arizona, California, and Texas, and even higher in Nevada. The change over
time in the number of undocumented includes those moving from one U.S. state to another as
well as those moving to or from the United States. Notable differences are seen in the estimates
across the three sources.
The mobility question that was used in the decennial census and the comparable question in the
ACS does not distinguish between immigrants and U.S. citizens who returned to the United
States after living in another country. Without differentiating between legal and illegal
immigrants, it is possible to estimate the amount of immigration that occurred during the period
by cross-tabulating the mobility results by citizenship. In 1990, around 30,000 Hispanic
noncitizens had moved to Arizona from another country in the preceding five years. This figure
more than tripled to nearly 100,000 in 2000. These figures are understated since some
individuals entered and left the state during the five years. The ACS results are not comparable.

	
  

14

IMMIGRATION
This section focuses on immigration for several reasons:
• the unusual importance of immigration in explaining Arizona’s overall population growth
and the growth of the Hispanic population in particular
• the actions taken and discussions still taking place to limit undocumented immigration,
both in Arizona and nationally
• the potential volatility of immigration of Hispanics to Arizona relative to the other
components of population change
Several topics are addressed, including the factors that cause individuals to immigrate,
immigration law, the estimated effect of Arizona’s employer sanctions law, and an outlook for
immigration apart from considering changes to immigration law. Immigration from Mexico is
emphasized due to the large proportion of Arizona’s immigrants who moved from Mexico.
Federal Immigration Law
Federal immigration policies have changed many times over the last century; only a few of those
changes are mentioned here. Between 1924 and 1965, a quota system was used, with specific
limits on the number of immigrants by country. This was replaced by a preference system in
which the skills of potential immigrants and their relationship to family living in the United
States took precedence. This resulted in a shift in the origins of immigrants from Europe to Asia,
Central America, and South America.
In response to the growing number of undocumented workers living in the United States, the
U.S. Congress passed the Immigration Reform and Control Act (IRCA) in 1986. The main
provisions of this act
• required employers to attest to their employees’ immigration status
• made it illegal to knowingly hire or recruit unauthorized immigrants
• granted amnesty to certain seasonal agricultural workers
• granted amnesty to illegal immigrants who had resided in the United States continuously
since before January 1, 1982
• increased enforcement at the border
According to the Department of Homeland Security, around 2.7 million people took advantage
of the amnesty provisions of IRCA and became legal residents of the United States, with nearly
all of these individuals gaining this status between 1989 and 1992.
Since 1990, the United States has limited the number of legal immigrants to 675,000 per year. (If
the number admitted is less than that in one year, the limit is increased for the following year.)
Immediate relatives of U.S. citizens, the largest category of admission, are not subject to this
limit. Thus, the total number of people granted permanent residence in each year greatly exceeds
675,000; it was 1.1 million in 2009.
While the policy of the nation is to consider immigrant skills and employment in the United
States, the limit on the number of legal immigrants is not modified in response to workforce
needs in the United States. From the early 1990s through mid-2000s, the combination of strong
economic growth and relatively few U.S.-born youths entering the workforce resulted in worker
shortages. Without any change in the limit on the number of legal immigrants, the worker

	
  

15

shortages were solved by a large increase in the number of undocumented immigrants. Had the
legal immigration limit been adjusted to reflect workforce needs, many of those who entered
illegally could have entered legally.
Efforts to Limit Undocumented Immigration in Arizona
Of the various immigrant-related laws that have been passed by the Arizona Legislature, two in
particular have had an impact on the number of undocumented immigrants living in Arizona. By
far, the greatest impact came from the Legal Arizona Workers Act (House Bill 2779 passed in
the first regular session of 2007), also known as the “employer sanctions law.” Passed in July
2007, this legislation went into effect at the beginning of 2008. The constitutionality of this law
was affirmed by the U.S. Supreme Court in 2011. This legislation has significantly reduced the
number of undocumented immigrants living in Arizona (discussed in the next subsection) and
likely will result in significantly less undocumented immigration to the state in the future than in
the past.
The other notable legislation is the “Support Our Law Enforcement and Safe Neighborhoods
Act” (Senate Bill 1070) passed in April 2010. Its goal was “attrition through enforcement.”
While it likely contributed to some additional undocumented immigrants leaving the state, its
effect was limited by a preliminary injunction by a federal judge that has kept some parts of the
legislation from being implemented. A number of legal challenges are pending.
Though additional bills to restrict the number of undocumented immigrants living in the state
introduced in the 2011 regular session were not passed, the possibility remains that additional
legislation will be passed by the Arizona Legislature in coming years.
An Estimate of the Effect of Arizona’s Employer Sanctions Law
The Public Policy Institute of California (PPIC) released a report in March 2011, Lessons From
the 2007 Legal Arizona Workers Act. The PPIC did a sophisticated analysis of conditions in
Arizona and in other states in an effort to estimate the number of Arizonans who left the state
after the passage of the Act. The methodology was designed to differentiate between the effects
of the recession and the effects of the Act.
While the methodology was sophisticated, the quality of the available data (primarily from the
American Community Survey and the Current Population Survey) leaves much to be desired.
Thus, the results of the study can provide only an estimate of the likely effect of the employer
sanctions law. The PPIC estimated that 92,000 undocumented immigrants left the state in 2008
and 2009 due to the employer sanctions law. This represents between 16 and 18 percent of the
estimated number of undocumented immigrants living in Arizona. The total number leaving was
greater than 100,000, with the balance of those leaving doing so because of the
disproportionately severe recession in Arizona.
Factors in Immigration Decisions
Most of the decisions made by people to migrate — across international boundaries or within the
same country — are based on “push” and “pull” factors. Among people of working age, a lack of
economic opportunity in the community in which an individual lives — for example, the loss of
major employers that has resulted in a chronically high unemployment rate — is a primary factor

	
  

16

that pushes an individual to consider a move. Any number of other factors — for example, an
undesirable climate — also can push an individual (regardless of workforce status) to consider a
long-distance move.
Job opportunity is a primary pull factor, one that has been particularly significant in Arizona,
which typically is among the job creation leaders in the nation. Noneconomic pull factors include
climate, environmental conditions, political atmosphere, presence of family and friends, etc.
Most of the immigration of Hispanics to the United States consists of young adults (and their
children), with most immigrants coming from Mexico. Economic factors, both push and pull,
have been the predominant causes of immigration to Arizona. Family reunification also is an
important factor.
In any country, whenever the size of a generation is substantially different from that of previous
generations, the supply and demand relationship for jobs is affected. For example, the babyboom generation (those born from 1946 through 1964) in the United States was substantially
more numerous than previous generations (see Chart 3). From 1952 through 1964, annual
average births were 66 percent higher than those from 1928 through 1941 (the Great Depression
occurred during this period). As baby boomers began to reach employment age in large numbers
around 1970, they greatly outnumbered those retiring and those moving up from entry-level jobs.
Thus, far more young people sought jobs than the number available. As a result, the
unemployment rate rose and the average wage stagnated. During this period (from about 1970
into the early 1980s), the economic pull factor of the United States to potential immigrants was
muted.
CHART 3
ANNUAL AVERAGE NUMBER OF BIRTHS IN THE UNITED STATES

Source: Centers for Disease Control and Prevention, National Center for Health Statistics.

	
  

17

In contrast, the “baby-bust” generation (those born from 1965 into the 1980s) was less numerous
than the baby-boom generation. The average number of births from 1965 through 1987 was 12
percent (475,000 per year) less than during the 1946-to-1964 baby boom and the average from
1972 through 1978 was 22 percent less than from 1952 through 1964. As the number of
American-born youths aging into workforce age began to decline significantly in the early 1990s,
a labor shortage developed in certain occupations, particularly entry-level positions and jobs not
deemed attractive (such as working in a poultry-processing facility). At the same time as the
demographic situation was changing, the U.S. economy embarked on the longest and strongest
economic expansion in history, creating many new jobs. Due to the worker shortages, the
economic pull factor of the United States was particularly strong from the early 1990s through
the mid-2000s.
Immigration of Mexicans
Of all Hispanics living in Arizona, 90 percent reported Mexican origin in the 2005-09 ACS. The
percentage is higher among noncitizens at more than 93. Mexicans do not account for a very
large share of legal immigrants, but appear to make up 60 percent of the undocumented
immigrants. Presumably, a sizable portion of the immigration from Mexico has always been
undocumented.
During the 20th century, only one generation in the United States was substantially larger than its
predecessors, but in Mexico each generation was significantly more numerous than the prior
generations. The increasingly larger generations were a result both of high fertility rates and a
large number of women of child-bearing age.
Under these conditions, it was difficult for the Mexican economy to create an adequate number
of jobs for young people entering the workforce and the competition for the jobs held wages in
check. Thus, the economic push factors from Mexico were significant throughout the 20th
century, though the imbalance in generation sizes peaked from the 1970s into the 1990s. In the
short term, peso devaluations and economic downturns temporarily caused even larger numbers
of Mexicans to seek work in other countries.
The first sizable influx of Mexican immigrants appears to have begun in the 1970s. A major
driver was the worsening conditions in Mexico related to young people finding work. In 1960,
the ratio of the number of Mexicans between 15 and 19 years old (a proxy for the number
entering the workforce) to the number between 60 and 64 years old (a proxy for the number
retiring from the workforce) was 4.75. It rose to 5.5 in 1970 and peaked at 6.9 in 1980. Many of
the Mexicans who illegally immigrated to the United States during this period became permanent
residents through IRCA.
Between the passage of IRCA and the early 1990s, the Arizona economy was in a long slump,
limiting opportunities for immigrants. Beginning in 1993 and lasting through 2006, the Arizona
economy created many jobs, though a lull occurred during the 2001 recession and the slow
recovery that followed. The pool of U.S.-born young adults from which Arizona draws much of
its workforce declined in size during this period of strong job growth, corresponding to the lower
number of births during the baby bust. The lure of jobs in Arizona not being filled by Americans

	
  

18

(a pull factor) caused a surge in undocumented immigration to the state. A peso devaluation in
1994 (a push factor) added to the flow of unauthorized immigrants from Mexico.
This increase in the number of immigrants during the 1990s created opportunities for immigrant
entrepreneurs to create businesses that catered largely to the immigrant population. This
provided the opportunity for an even larger number of immigrants to come to the United States
than might be expected from the worker shortages. The annual increase in the number of
undocumented immigrants living in Arizona doubled between the early 1990s and 2000.
The ratio of the number of Mexicans between 15 and 19 years old to the number between 60 and
64 years old still was very high in 1990 at 6.0, but dropped back to 4.6 in 2000, reducing the
push factors from Mexico. The number of undocumented immigrants entering Arizona slipped
after 2000, but remained high through 2006.
The period of strong job growth ended in 2007, as the Arizona economy entered a long and deep
recession late in the year. The number of legal immigrants to the state was about the same in
2008 and 2009 as in 2004 through 2006. However, the number of undocumented immigrants
living in Arizona quickly shifted from increases in every year to sizable decreases in 2008 and
2009. This abrupt shift was in part due to the weak economy, but the passage of the employer
sanctions law had a greater effect.
Also contributing to the decrease in the number of undocumented immigrants is the greater
number of Americans entering the workforce and competing for jobs that had been filled by
immigrants, due to an increase in U.S. births during the 1980s. The generational imbalance in
Mexico also continues to shrink. Thus, even without considering Arizona’s anti-immigrant
legislation, a slowdown in undocumented immigration would have occurred during the past few
years.
U.S. Immigration Outlook
The Census Bureau is the source of population projections for the United States. While the
Census Bureau also produces projections for Mexico, the Mexico projections cited in this section
were made by CONAPO — Consejo Nacional de Poblacion, Mexico’s National Population
Council. Relative to CONAPO, the Census Bureau assumes higher birth rates, lower death rates,
and lesser net out-migration from Mexico.
Since the late 1980s, the number of births in the United States has not fluctuated much and has
averaged just more than 4 million per year, nearly as many as in the peak years of the baby
boom. Thus, the number of young adults entering the workforce will be roughly equal to the
number of baby boomers over the next 15 years. (The ratio of the number 15-to-19 years old to
the number of baby boomers in the peak five-year age group is expected to reach 1 in the next
several years, compared to 0.8 in 1990 and 0.9 in 2000.) With the number of births in the United
States projected to rise, the ratio of the number 15-to-19 years old to the number 60-to-64 years
old will rise from about 1.1 in the next several years to 1.3 in coming decades.
Under these conditions, opportunities for immigrants (the economic pull of the United States)
will be considerably less than from the mid-1990s through mid-2000s. Jobs likely will be

	
  

19

available for relatively uneducated immigrants in certain occupations in which American-born
youths show limited interest, such as hotel maids, but otherwise job opportunities for
undocumented immigrants will be tight.
In Mexico, the age demographics are rapidly changing. Fertility rates have fallen considerably,
from 6.7 in 1970 and 2.7 in 2000 to less than 2.1 in 2010, resulting in 500,000 fewer births in
2010 than in 2000. The fertility rate in Mexico now is identical to that in the United States. The
number of births in Mexico is expected to continue to drop, with the fertility rate dropping below
1.9. A stable population size (not considering immigration) is associated with a fertility rate of
2.1.
The decrease in the number of births in Mexico leads to a situation within the next several years
in which the number aging into the workforce will begin to drop. After figures as high as 6 as
recently as 1990, the ratio of those 15-to-19 years old to those 60-to-64 years old will drop to 2
in 2020 and about 1 in 2040. The age distribution will resemble that of the United States by
2030. Thus, the economic push from Mexico should lessen, with supply and demand for jobs in
Mexico falling into balance, followed by increases in wage rates. (The primary risk for Mexico is
that the drug cartels and the government’s efforts to limit the cartels will so destabilize the
country that the Mexican economy is adversely affected.)
Therefore, even if Arizona had not passed anti-immigrant laws, push and pull factors should
combine to keep undocumented immigration lower than it was from the mid-1990s through mid2000s. If the status quo remains unchanged — federal immigration policy is unchanged or
modified only slightly, Arizona’s anti-immigrant legislation remains in place, and a significant
number of other states do not adopt similar measures — then it is likely that Arizona will see a
greatly reduced flow of undocumented immigrants compared to any historical period prior to
2008.

	
  

20

DEMOGRAPHIC PROFILE OF HISPANICS IN ARIZONA
This profile is based on the results of the 2010 census compared to the 1990 and 2000 censuses
whenever possible. Since the 2010 census was limited to only a few questions, the 2005-09 ACS
is used for the most up-to-date data for several indicators. Only percentages of the total
population are presented.
Hispanics are compared to non-Hispanics in this section. The Hispanic population was
subdivided into three citizenship categories: those born in the United States, those who have
become naturalized citizens, and those who are not citizens. Since citizenship was not included
in the 2010 census, the latest data for Hispanics by citizenship come from the ACS.
All people born in the United States or to U.S. citizens are citizens. Those born in another
country to parents who are not citizens are classified either as naturalized citizens or noncitizens.
No distinction can be made between those who are legally living in the United States and those
who are undocumented, but a sizable share of the Hispanics who are not citizens likely are
undocumented immigrants.
The figures reported in this section were derived from the Public Use Microdata Sample (PUMS)
for the 1990 and 2000 censuses and for the 2005-09 ACS. (A PUMS file is not available for
2010.) Accessing the PUMS is the only way to obtain a variety of detailed census data. Standard
Census Bureau tabulations do not provide many cross-tabulations of two or more indicators
(such as educational attainment by age), do not provide any data on Hispanics by citizenship
status, and in some cases do not even provide data for Hispanics as a whole. The primary
concern with using the PUMS is that the sample size is small, causing sampling error to be a
significant concern.
Race/Ethnicity and Origin
Hispanics are considered to be an ethnic group. Individual Hispanics can be of any race. Data on
race and ethnicity are available from the 2010 census.
Prior to the 2000 census, respondents had to select a single race. Since then, respondents have
the option to report two or more races. The race data since 2000 are tallied in two ways: (1)
“alone” — those of more than one race are reported in a separate category and the sum of the
categories equals the population count; and (2) “combined” — each race of a person of more
than one race is included in the race category’s total; the sum of the races sums to more than the
number of people counted. Regardless of how the data are tallied, the 2000 census and
subsequent race data are not directly comparable to the earlier data.
In 2010 in Arizona, 82 percent of non-Hispanics identified their race as white; somewhat less
than in 2000 (85 percent) and 1990 (88 percent). Very few non-Hispanics indicated a race other
than those specified (white, black, American Indian, Asian, and Pacific Islander). In contrast,
about half of the Hispanics in 2010 identified their race as white and nearly as many as “other.”
These percentages have not changed much over time. Based on the 1990 and 2000 censuses, the
racial percentages do not vary much across the three Hispanic citizenship categories. (The 200509 ACS results for Hispanic races are inconsistent with those of the 2010 and earlier censuses.)

	
  

21

The origin of Hispanics (the birthplace of an individual not born in the United States or for those
born in the United States the birthplace of their ancestors) is not available from the 2010 census.
Hispanics living in Arizona are overwhelmingly of Mexican origin, accounting for 90 percent of
the total in 2005-09, barely less than in 1990.
Place of Birth and Citizenship
Place of birth and citizenship are not available from the 2010 census. The place of birth of an
individual has been classified into one of four categories: Arizona, elsewhere in the United
States, Mexico, or other foreign country. Among non-Hispanics, only 32 percent were born in
Arizona, according to the 2005-09 ACS. A little more than 60 percent were born elsewhere in the
United States and 7 percent were born in another country. A gradual shift has occurred since
1990, with the shares of non-Hispanics born in Arizona and in other countries increasing.
In contrast, a higher proportion of Hispanics were born in Arizona (47 percent) and relatively
few (17 percent) were born elsewhere in the United States. The proportion born in another
country was much higher at 36 percent. The foreign-born proportion has increased over time
while the Arizona-born share has decreased. More than 90 percent of the Hispanics who were
born in another country came from Mexico. Among Hispanics who were naturalized citizens, 86
percent were born in Mexico. The percentage was higher at 94 percent among Hispanic
noncitizens.
A surge in immigration during the 1990s lowered the percentage of the Hispanics living in
Arizona who were born in the United States from 74 in 1990 to 64 in 2000. The Hispanic
noncitizen proportion rose from 18-to-28 percent during this decade. The change in the
percentages reversed marginally after 2000, based on the 2005-09 ACS (see Chart 4).
CHART 4
HISPANICS IN ARIZONA BY CITIZENSHIP, SHARE OF TOTAL

Source: U.S. Department of Commerce, Census Bureau, decennial censuses (1990 and 2000) and
American Community Survey (2005-09).

	
  

22

Citizenship of Hispanic Children
Nearly all of Arizona’s Hispanics under the age of 5 in 2005-09 were born in the United States,
as were more than 80 percent of those of school age (5 to 17). The U.S.-born percentage dropped
during the 1990s with the surge in immigration, but returned to the 1990 level in 2005-09 (see
Table 5).
A more in-depth look at the citizenship of Hispanic children relative to their parents was made
through a special analysis of PUMS data that was limited to Hispanic children under the age of
21 who were identified as a child of the householder. This “own-child” designation includes
adopted children and step-children. The children were divided into four categories:
• all children in the family were born in the United States
• all children in the family were naturalized citizens
• all children in the family were not citizens
• the citizenship of the children in a family was mixed
For families with just one parent present, the parent was placed into one of the three citizenship
categories. For families in which two parents are present, the citizenship of the parents was
placed into one of nine categories, as specified in Table 6.
As can be seen at the top of Table 6, this analysis verifies that a high percentage of Hispanic
children were born in the United States, and that all of their siblings also were born in the United
States. Less than 10 percent of the children in 2005-09 lived in families in which all of the
children were not citizens, and less than 4 percent of the children lived in families in which the
children had differing citizenship statuses.
The cross-tabulation of the citizenship of children and citizenship of parents largely yields
predictable results. In 90 percent of the families in which all of the children are not citizens,
either the sole parent or both parents were not citizens. In families in which the children were of
mixed citizenship, 75 percent of the children lived with a single parent or both parents who were
not citizens. In families in which all of the children were U.S. born, 47 percent of the children
lived with a single parent or with both parents who were born in the United States. However, 27
percent lived with a single parent or both parents who were not citizens.
TABLE 5
HISPANICS LESS THAN 21 YEARS OF AGE IN ARIZONA
BY CITIZENSHIP, SHARE OF TOTAL
1990
Age
<5
5-13
14-17
18-20
Total

94.6%
88.6
81.5
74.6
87.1

U.S. Born
2000
2005-09
93.2%
83.4
71.9
62.6
81.2

97.1%
88.1
80.4
69.1
87.1

Naturalized Citizen
1990
2000
2005-09
0.6%
1.4
3.1
4.6
1.9

0.4%
1.2
2.8
3.2
1.5

0.1%
0.6
1.7
2.9
0.9

1990
4.9%
10.0
15.4
20.9
11.0

Not a Citizen
2000
2005-09
6.4%
15.3
25.3
34.3
17.3

2.8%
11.3
15.7
28.0
12.0

Source: U.S. Department of Commerce, Census Bureau, decennial censuses (1990 and 2000) and
American Community Survey (2005-09).

	
  

23

TABLE 6
HISPANIC PARENTS AND CHILDREN IN ARIZONA BY CITIZENSHIP,
SHARE OF TOTAL
1990

2000

2005-09

Citizenship of Children:
All U.S. Born
All Naturalized Citizens
All Not Citizens
Mixed

87.7%
1.4
7.7
3.1

82.5%
1.1
11.7
4.7

86.8%
0.7
9.0
3.5

Citizenship of Parents:
One Parent, U.S. Born
One Parent, Naturalized Citizen
One Parent, Not a Citizen
Both Parents U.S. Born
Father U.S. Born, Mother Naturalized
Father U.S. Born, Mother Not a Citizen
Father Naturalized, Mother U.S. Born
Both Parents Naturalized Citizens
Father Naturalized, Mother Not a Citizen
Father Not a Citizen, Mother U.S. Born
Father Not a Citizen, Mother Naturalized
Both Parents Not a Citizen

18.6
2.9
5.7
38.2
3.1
3.8
3.3
2.8
2.8
4.3
1.8
12.7

17.0
2.9
8.8
26.4
2.3
3.4
2.5
4.3
4.9
3.1
2.6
21.8

18.8
3.3
11.9
22.0
2.5
4.8
1.7
3.4
4.5
2.8
1.7
22.7

Source: U.S. Department of Commerce, Census Bureau, decennial censuses (1990 and 2000) and
American Community Survey (2005-09).

In the second part of Table 6, it can be seen that three-fourths of the children in 2005-09 lived in
four of the 12 categories of parental citizenship: one parent U.S. born, both parents U.S. born,
one parent not a citizen, and both parents not a citizen. A significant change in the percentages
occurred between 1990 and 2000, with further change between 2000 and 2005-09 in three
categories: the percentage of children living in families in which each parent was U.S. born has
dropped, while the percentage living in families in which both parents or the only parent is not a
citizen increased.
Putting it all together, while a high share of Hispanic children are U.S. citizens by birth, an
increasing proportion of these children have been born to parents who are not citizens. Because
of the increase in immigration in the 1990s, a higher proportion of Hispanics are noncitizens than
in 1990. It appears that the fertility rate of Hispanic noncitizens is higher than that of Hispanics
born in the United States.
Language
Information on the language spoken is not available from the 2010 census. Earlier censuses and
the ACS report the language spoken at home both for households and for persons. About 90
percent of non-Hispanics spoke English at home in 2005-09, with about 2 percent speaking
Spanish. These percentages have not changed over time. In contrast, more than 70 percent of
Hispanic households spoke Spanish at home. This percentage ranges from 56 percent among
U.S.-born Hispanics to 97 percent of Hispanic noncitizens. The proportion of U.S.-born
Hispanics speaking Spanish at home has decreased over time.
	
  

24

If a language other than English is spoken, then the ability of the person to speak English is
classified by the ACS and earlier decennial censuses into one of four categories: very well, well,
not well, or not at all. Among non-Hispanics speaking a language other than English at home,
more than 70 percent spoke English very well and nearly 90 percent spoke English at least well
in 2005-09. These percentages have increased over time.
Similarly, more than 80 percent of U.S.-born Hispanics who speak another language at home
spoke English very well and more than 90 percent spoke English at least well in 2005-09. In
contrast, only 20 percent of Hispanic noncitizens spoke English very well. The percentage of
Hispanic noncitizens not speaking English at all has increased since 1990 to 29 percent (see
Table 7); another 32 percent did not speak English well in 2005-09.
If none of the individuals older than 14 living in a household can speak English very well, then
the household is considered to be linguistically isolated. Less than 2 percent of non-Hispanic
households were linguistically isolated and only 5 percent of households headed by a Hispanic
born in the United States were isolated. In contrast, more than half of the Hispanic noncitizen
households were linguistically isolated in 2005-09, a percentage that has increased since 1990
(see Table 8).
TABLE 7
ABILITY TO SPEAK ENGLISH OF ARIZONANS NOT SPEAKING ENGLISH AT
HOME BY ETHNICITY AND CITIZENSHIP, SHARE OF TOTAL

Very Well:
1990
2000
2005-09
Not at All:
1990
2000
2005-09

Not Hispanic

Total

63.1%
67.6
71.9

59.2%
51.0
52.0

3.0
2.0
2.4

Hispanic
Naturalized
U.S. Born
Citizen

7.4
12.0
13.8

75.3%
75.1
81.5

44.9%
41.9
45.8

0.9
1.4
1.3

8.5
7.4
9.1

Not a Citizen
26.6%
23.1
20.3
22.2
26.7
29.1

TABLE 8
LINGUISTIC ISOLATION OF ARIZONA HOUSEHOLDS BY ETHNICITY AND
CITIZENSHIP, SHARE OF TOTAL

1990
2000
2005-09

Not Hispanic
1.9%
1.8
1.5

Hispanic
Naturalized
U.S. Born
Citizen
7.7%
27.9%
7.0
27.9
5.0
25.5

Total
18.2%
24.0
24.8

Not a Citizen
44.2%
50.2
54.7

Source (Tables 7 and 8): U.S. Department of Commerce, Census Bureau, decennial censuses (1990 and
2000) and American Community Survey (2005-09).

	
  

25

Gender and Age
According to the 2010 census, males make up slightly less than half of the non-Hispanic
population but slightly more than half of the Hispanics living in Arizona. The differential largely
is due to the larger share of older people in the non-Hispanic population and the earlier age at
which men die. Using the 2005-09 ACS, males accounted for less than half of the Hispanic
naturalized citizens, who are disproportionately older than the rest of the Hispanic population. In
contrast, slightly more than half of the U.S.-born Hispanics were males. Among Hispanic
noncitizens, nearly 56 percent were men, a percentage that has increased over time (see Table 9).
Though the age distribution is somewhat different for males and females, the differences are
primarily among the elderly; the following discussion of age is based on males and females
combined. The age distribution in 2010, in five-year age groups, is compared for Hispanics and
non-Hispanics in Chart 5. The distribution is sharply different, with Hispanics much younger
than non-Hispanics. The higher percentages of the baby-boom generation (those from 46 to 64
years old in 2010) than other age groups are easily seen among non-Hispanics.
The aging of the baby-boom generation dominates the changes in the age distribution of nonHispanics between 1990 and 2000 and from 2000 to 2010. In contrast, large changes in the age
distribution of Hispanics did not occur between 1990 and 2000. Between 2000 and 2010, the
share of Hispanics between the ages of 45 and 64 increased, offset by a decline in the share of
young adults (especially those 20-to-29 years old) and children less than 10 years of age.
In 1990, the age distribution of Hispanics in Arizona was much younger than non-Hispanics, but
was not as young as that of Mexico. This relationship began to change during the 1990s due to
fertility rates in Mexico falling more than among Hispanics in Arizona and because of the
increase in the immigration of young adults from Mexico to the United States. In 2010, the age
distribution of Arizona’s Hispanics was younger than that of Mexico. A higher proportion of
TABLE 9
GENDER OF ARIZONA RESIDENTS BY ETHNICITY AND CITIZENSHIP, MALES AS
A SHARE OF THE TOTAL

1990
2000
2005-09
2010
Change:
1990 to 2000
2000 to 2005-09
2000 to 2010

Not Hispanic
49.1%
49.2
49.3
49.5
0.1
0.1
0.3

Hispanic
Naturalized
U.S. Born
Citizen
50.1%
51.5%
50.4
48.9
50.8
48.2
na
na

Total
50.5%
51.5
52.0
50.2
1.0
0.5
-1.3

0.3
0.4
na

-2.6
-0.7
na

Not a Citizen
51.9%
54.6
55.9
na
2.8
1.3
na

na: not available
Source: U.S. Department of Commerce, Census Bureau, decennial censuses and American Community
Survey (2005-09).

	
  

26

CHART 5
AGE DISTRIBUTION OF ARIZONA RESIDENTS IN 2010 BY ETHNICITY

Source: U.S. Department of Commerce, Census Bureau, 2010 census.

Arizona’s Hispanics were less than 10 years old and a lesser proportion were at least 60 years
old. (The data for Mexico come from their decennial census, which is conducted within several
weeks of the U.S. decennial census.)
In order to examine the age distribution of the three Hispanic citizenship categories, the 2005-09
ACS must be used. The age distribution differs notably across the three categories (see Table
10). Relative to the other categories, the U.S.-born group is disproportionately less than 20 years
old, with relatively few 30 or older. Hispanic naturalized citizens are much older, with relatively
many 35 or older and few less than 25. Hispanic noncitizens are especially numerous between
the ages of 20 and 39, with few 55 or older.
Several factors should be considered in interpreting the changes in age distribution over the last
20 years across the three Hispanic categories:
• Because IRCA legalized long-time Hispanic residents between 1989 and 1992 and since
some went on to become naturalized citizens in the years following, the naturalized
population became older than it had previously been.
• Children born in the United States to noncitizens are included in the U.S.-born category
rather than the noncitizen category.
• A large increase in immigration occurred during the 1990s, shifting the age distribution
of the noncitizen category.
Changes over time in the age distribution also have varied considerably across the three Hispanic
categories. The U.S.-born category has experienced a large rise in the share of children despite a
decreasing share in the 25-to-44 age group (largely due to births in the United States to
noncitizens). The noncitizen category has experienced a significant increase in share between the

	
  

27

TABLE 10
AGE DISTRIBUTION OF ARIZONA RESIDENTS BY AGE, ETHNICITY AND
CITIZENSHIP, SHARE OF TOTAL

Not Hispanic
Age 0-19:
1990
2000
2005-09
2010
Age 20-39:
1990
2000
2005-09
2010
Age 40-59:
1990
2000
2005-09
2010
Age 60 and Older:
1990
2000
2005-09
2010

Hispanic
Naturalized
U.S. Born
Citizen

Total

Not a Citizen

26.8%
25.2
24.1
23.3

42.7%
42.0
40.9
41.0

50.0%
53.8
55.4
na

10.2%
7.8
4.5
na

24.6%
24.3
16.5
na

32.3
27.3
25.8
25.0

34.2
35.2
33.6
31.3

29.9
26.8
24.8
na

45.4
37.9
33.2
na

48.0
53.7
54.8
na

21.4
26.8
27.9
27.7

15.7
16.7
18.7
20.0

13.6
13.5
13.9
na

28.4
37.9
42.4
na

19.5
18.3
23.6
na

19.5
20.7
22.3
24.2

7.4
6.1
6.7
7.7

6.5
5.9
5.9
na

16.0
16.4
19.9
na

7.9
3.7
5.1
na

na: not available
Source: U.S. Department of Commerce, Census Bureau, decennial censuses and American Community
Survey (2005-09).

ages of 25 and 59, with large decreases in share among those less than 25. The share of
naturalized citizens between the ages of 40 and 74 has climbed, offset by a large decrease in
share among those less than 35, particularly among those 20 to 34 years old.
Household Type and Relationship
Approximately 98 percent of Arizonans live in households, a percentage that varies little
between non-Hispanics and Hispanics or between the three Hispanic categories. The remainder
of the population lives in group quarters, such as a prison or a nursing home.
In households, one person is identified as the householder; the number of householders equals
the number of occupied housing units. The relationship of others living in a household is
expressed relative to the householder; broad categories include spouse, child, other relative, and
not related. If more than one related person is living in a household, then the household is
considered to be a family. The number of people living in households is divided by the number
of households to get the average household size.

	
  

28

Data on household type and relationship are available from the 2010 census for Hispanics and
non-Hispanics. Household type is expressed as the number of households. Family households
include married couples and single-parent families. Nonfamily households include people living
alone and households of more than one person in which none are related. The percentage of
households headed by a married couple has been shrinking, dropping below half of all
households between 2000 and 2010. Married couples account for a slightly greater share of
Hispanic than non-Hispanic households. Married-couple households are more common among
Hispanic immigrants than among Hispanics born in the United States. A higher percentage of
non-Hispanic than Hispanic households are nonfamilies, but this results in part from
disproportionately more elderly non-Hispanics than Hispanics; those living alone are not
classified as a family. Single-parent families are more common among Hispanics.
Household relationship refers to people. In part because of the higher proportion of Hispanic
households with a young-to-middle-age adult householder and in part due to higher fertility rates
among Hispanics, a higher proportion of Hispanics than non-Hispanics are children. A greater
proportion of Hispanics are related to the householder in a way other than spouse or child; the
percentage is particularly high among Hispanic noncitizens. A slightly higher percentage of
Hispanics are not related to the householder among Hispanic noncitizens.
Economic Well-Being
In order to get a sense of the economic well-being of the Hispanic population, median household
income and the poverty rate were analyzed. The latest figures come from the 2005-09 ACS;
these topics were not included in the 2010 census. The poverty rate of Hispanics has been more
than twice that of non-Hispanics over the last two decades (see Table 11). Among Hispanics,
noncitizens have a much higher poverty rate than citizens. Hispanics born in the United States
have not experienced any narrowing of the gap with non-Hispanics over the last 20 years.
However, the gap has narrowed among immigrants — both naturalized citizens and noncitizens.
Household income is highly correlated to the age of the householder. It is lowest among those
with a household head less than 25 years old or 75 or older, and highest among householders
between the ages of 45 and 54. In every age group, the median household income of Hispanics in
2005-09 was less than that of non-Hispanics. Of the three Hispanic citizenship categories, U.S.born Hispanics have the highest incomes, and noncitizens the lowest figures, in every age group.
TABLE 11
POVERTY RATE IN ARIZONA BY ETHNICITY AND CITIZENSHIP

1990 Census
2000 Census
2005-09 ACS

Not Hispanic
12.8%
10.3
11.1

Hispanic
Naturalized
U.S. Born
Citizen
25.0%
28.1%
20.9
19.1
22.1
17.1

Total
28.4%
24.1
23.6

Not a Citizen
42.7%
32.9
28.9

Source: U.S. Department of Commerce, Census Bureau, decennial censuses (1990 and 2000) and
American Community Survey (2005-09).

	
  

29

The differential in median household income between non-Hispanics and Hispanics is least
among households with young householders (see Chart 6). The differential between U.S.-born
Hispanics and non-Hispanics is 10 percent among householders less than 25 but approximately
20 percent among householders between the ages of 35 and 74.
During the 1990s, median income rose considerably after adjusting for inflation, with Hispanics
experiencing gains much larger than non-Hispanics; among Hispanics, the increase was greater
for immigrants than for those born in the United States. In contrast, between 2000 and 2005-09,
median household income dropped slightly after adjustment for inflation, with declines similar
among Hispanics and non-Hispanics, and among U.S.-born Hispanics and immigrants.
Workforce Indicators
The latest data reported in this subsection come from the 2005-09 ACS. The percentage of
Hispanics 16 or older who are employed was greater than that of non-Hispanics, but this is a
result of the differences in the age distribution — a higher share of non-Hispanics is of
retirement age. Among Hispanics, there is little difference across the three citizenship categories.
The percentage unemployed was higher among Hispanics than non-Hispanics, with little
difference in the unemployment rate between U.S.-born Hispanics and noncitizens.
Variations in the age distribution also explain some of the differences in the number of workers
in family households. Relative to non-Hispanics, a higher share in all three Hispanic citizenship
categories have at least one person in the family in the workforce.
CHART 6
MEDIAN HOUSEHOLD INCOME OF HISPANICS AS A PERCENTAGE OF NONHISPANICS IN ARIZONA, 2005-09, BY AGE OF HOUSEHOLDER AND CITIZENSHIP

Source: U.S. Department of Commerce, Census Bureau, American Community Survey.

	
  

30

Hispanics and non-Hispanics vary considerably in the occupations in which they work. A higher
share of U.S.-born Hispanics than non-Hispanics work in the blue-collar occupations of
agriculture, mining, construction, and production. The shares in these occupations among
Hispanic immigrants, especially noncitizens, are much higher than those of U.S.-born Hispanics.
Relative to non-Hispanics, a considerably lower proportion of U.S.-born Hispanics work in
professional occupations. Very few Hispanic noncitizens work in professional occupations. The
differential between Hispanics and non-Hispanics has increased over time. In service
occupations, the proportion of U.S.-born Hispanics was not much different from non-Hispanics
in 1990, but since then, the non-Hispanic share has fallen while the U.S.-born Hispanic share has
increased a little. After being lower in 1990, the proportion of Hispanic noncitizen immigrants
working in service occupations has become equal to that of non-Hispanics.
Compared to non-Hispanics, the share of U.S.-born Hispanics who are self employed is only half
as much, offset by somewhat higher proportions working as wage and salary employees in both
the private and public sectors. Very few Hispanic noncitizens work in the public sector; the
proportion self employed is between that of non-Hispanics and U.S.-born Hispanics.
In the past, working Hispanic noncitizens averaged fewer hours worked per year than other
Hispanics and non-Hispanics. According to the ACS, however, this differential has disappeared,
with Hispanic noncitizens working more hours per year than others (see Table 12).
Educational Attainment
Educational attainment is not available from the 2010 census. A question on past decennial
censuses and on the ACS asks for the “highest degree or level of school completed” of each
TABLE 12
SELECTED CHARACTERISTICS OF ARIZONA WORKERS,
2005-09, BY ETHNICITY AND CITIZENSHIP

Occupation:
Professional
Services
Blue Collar
Other
Class of Worker:
Public Employee
Private Employee
Self Employed
Hours Worked:
2,000 or More
1,000 to 1,999
Less Than 1,000

Hispanic
Naturalized
U.S. Born
Citizen

Not Hispanic

Total

36.9%
45.2
9.2
8.7

15.5%
49.8
23.5
11.2

21.5%
53.4
14.0
11.1

19.2%
47.5
22.0
11.3

5.8%
45.1
37.7
11.4

15.9
72.2
11.1

12.0
79.2
7.5

17.5
75.7
5.4

14.5
75.0
9.8

3.4
85.4
9.8

58.6
22.2
19.2

60.1
22.0
17.9

57.8
21.9
20.3

66.1
20.9
13.0

61.6
22.4
16.0

Source: U.S. Department of Commerce, Census Bureau, American Community Survey.

	
  

31

Not a Citizen

person. The responses from the 2005-09 ACS have been summarized in Table 13. The standard
tabulation produced by the Census Bureau includes all persons 25 or older; age 25 is used since a
high proportion of the younger population still is attending school. However, because of a very
strong relationship between educational attainment and age, and because of the differing age
distributions between non-Hispanics and the various Hispanic citizenship categories, it is
necessary to look at educational attainment by age group.
Looking first at those between 25 and 34 years old — the youngest age group in which a sizable
proportion of the individuals have completed their formal education — a very large difference in
educational attainment is seen between non-Hispanics and Hispanics. A far higher percentage of
Hispanics have not graduated from high school and a much lesser percentage have earned at least
a bachelor’s degree.
Wide differentials in educational attainment are seen across the three Hispanic categories, with
educational attainment much higher among the U.S.-born population than among those who are
not citizens. However, even U.S.-born Hispanics have a much lesser attainment than nonHispanics — more than twice the share have not graduated from high school and less than half
the percentage have earned at least a bachelor’s degree.
The relationships are similar among the older age groups, with the differential between nonHispanics and Hispanics even larger than in the 25-to-34 age group. In contrast, the differentials
are smaller among those 15-to-24 years old. However, since many individuals in this age group
are still attending school, the observed differentials are likely to increase.
The changes in educational attainment between 2000 and 2005-09 are shown in Table 14. The
educational attainment of Hispanics living in Arizona has improved over time, with the
percentage without a high school diploma falling and the percentage with at least some college
increasing. However, the gap between Hispanics and non-Hispanics has not changed
significantly: while the improvement in the percentage of high school graduates was greater
TABLE 13
EDUCATIONAL ATTAINMENT IN ARIZONA, 2005-09

Age Group
Less Than High School
Graduate:
15-24
25-34
35-44
45-54
Bachelor’s Degree or Higher:
15-24
25-34
35-44
45-54

Not
Hispanic

Total

Hispanic
Naturalized
U.S. Born
Citizen

40.5%
7.5
7.5
6.9

53.3%
35.6
37.0
39.0

51.3%
17.5
17.3
20.3

35.1%
28.0
34.9
41.5

5.9
32.4
33.7
30.8

2.1
9.2
10.3
10.6

2.2
13.4
15.1
12.7

3.6
12.8
12.2
11.7

Source: U.S. Department of Commerce, Census Bureau, American Community Survey.

	
  

32

Not a
Citizen
59.8%
54.5
56.5
66.5
1.4
4.5
5.2
6.7

TABLE 14
PERCENTAGE POINT CHANGES IN EDUCATIONAL ATTAINMENT IN ARIZONA
BETWEEN 2000 AND 2005-09

Age Group
Less Than High School
Graduate:
15-24
25-34
35-44
45-54
Bachelor’s Degree or Higher:
15-24
25-34
35-44
45-54

Not
Hispanic

Total

Hispanic
Naturalized
U.S. Born
Citizen

Not a
Citizen

-2.7
-2.4
-2.1
-0.8

-8.9
-8.2
-7.1
-7.3

-5.0
-6.5
-6.6
-6.5

-23.3
-17.6
-11.2
-8.3

-12.7
-9.8
-14.1
-11.6

0.7
2.4
5.5
-1.3

0.7
0.7
1.3
1.2

0.6
0.7
3.5
0.5

2.1
5.9
2.1
1.6

0.5
0.1
0.1
2.5

Source: U.S. Department of Commerce, Census Bureau, decennial census (2000) and American
Community Survey (2005-09).

among Hispanics than non-Hispanics, Hispanics did not gain relative to non-Hispanics in terms
of the percentage who have earned at least a bachelor’s degree.
The improvement in the percentage graduating from high school has been greater for Hispanic
immigrants than for U.S.-born Hispanics, but the gap in this measure still is significant. Not
much difference between Hispanic immigrants and U.S.-born Hispanics is present in the
improvement in the percentage earning at least a bachelor’s degree.
In each age group, the educational attainment of Arizona’s Hispanics on average is greater than
that of residents of Mexico. However, the percentage of Hispanics in Arizona who have earned a
bachelor’s degree or more is less than the percentage in Mexico. Further, the educational
attainment gap is being narrowed as educational attainment in Mexico rises faster than among
Hispanics in Arizona. The differences between Arizona Hispanics and Mexicans are smallest in
the youngest (15-to-24) age group. Mexico still has a disproportionate share of its 15-to-24 yearold residents not going beyond a junior high education, but otherwise its deficiencies in
attainment relative to Arizona Hispanics are minor. (Educational attainment in Mexico comes
from Mexico’s decennial census. It is part of the questionnaire sent to all households and thus
sampling error is not present. However, since Mexico’s educational system differs from that of
the United States, the educational attainment results for Mexico are not fully consistent with
those of the United States and are not presented in this report.)
Comparing Arizona’s noncitizen Hispanics to residents of Mexico, differences in educational
attainment at the low end are not significant, as the percentage graduating from high school is
slightly higher among Arizonans. However, the percentage earning a bachelor’s degree or more
is far less among Arizona’s noncitizen Hispanics.

	
  

33

EDUCATIONAL INDICATORS
In the 2001 Morrison Institute report Five Shoes Waiting to Drop on Arizona’s Future, the
“Latino Education Dilemma” was discussed. The rationale for focusing on Latinos was based on
a combination of factors: the large size of the Latino school-age population, the fast growth in
this population, and the below-average educational achievement and attainment of this group.
Educational achievement, or performance, is based on measures such as test scores and
eligibility rates for admission to Arizona universities.
In the remainder of this section, all references to racial groups imply the non-Hispanic portion of
each group. Public school enrollment in 2010 was nearly as high among Hispanics as whites,
with each group accounting for between 42 and 43 percent of the total enrolled. While the rate of
increase in Latino enrollment over the last 10 years was substantial at 50 percent, it was nearly as
high among blacks (45 percent) and was much higher among Asians (80 percent). (In contrast,
the number of white students decreased slightly and enrollment of Native Americans fell more.)
The educational achievement of Hispanics on a variety of measures is considerably below that of
whites and Asians. However, Hispanics perform as well as blacks and somewhat better than
Native Americans.
Thus, the challenges the state faces regarding Latino education — underachievement and fast
growth in the number of students — is typical of that of other minority groups. What differs is
the large size of the Hispanic population. Looking forward, the percentage growth in the number
of Hispanic children may drop below blacks and Asians, based on reduced undocumented
immigration due to the state’s anti-immigration laws and the sharp drop in the Hispanic fertility
rate that already has occurred.
Factors Affecting Educational Achievement and Attainment
Two reports written by Paul Barton and Richard Coley for the Educational Testing Service
provide good summaries of the factors affecting educational achievement and attainment. These
factors apply to all racial/ethnic groups.
In The Family: America’s Smallest School (2007), Barton and Coley discuss several factors:
• Parent-Pupil Ratio. In particular, children in single-parent households do not perform as
well academically.
• Family Finances. Low incomes/high poverty, food insecurity, and unemployment of
parents all negatively affect educational achievement.
• Literacy Development. Before starting kindergarten, differences in academic abilities are
measured in children. Those with lesser abilities had not been read to by their parents as
much as others and more generally, did not hear as many words spoken.
• Quality of Child Care.
• The Home as an Educational Resource. The availability of books, magazines, newspapers
and the Internet, as well as a desk or table at which to work, all affect academic
development. The amount of time watching television is negatively associated with
educational achievement.
• Parent-School Relationship (parental participation).
Four factors — single parent household, parents reading to children, hours watching television,
and the frequency of school absences — were found to explain 68 percent of the differences

	
  

34

across states in eighth grade reading scores on the National Assessment of Educational Progress
(NAEP) test.
A synthesis of research studies done by Barton and Coley, Parsing the Achievement Gap II
(2009) indicated that 16 factors related to life experiences and conditions are correlated with
cognitive development and academic achievement. Differences in these factors across
racial/ethnic groups and across income groups mirror the gaps seen in educational achievement
across these groups:
• Curriculum Vigor. This includes participation in Advanced Placement classes, which is
lower among minorities and low-income students.
• Teacher Preparation. In particular, minorities and low-income children are more likely to
be taught math by teachers who did not major or minor in math.
• Teacher Experience. Minorities and low-income children are taught by individuals with
less average teaching experience.
• Teacher Absence and Turnover. These are higher in minority/low-income schools.
• Class Size. Minority schools have larger average sizes.
• Availability of instructional technology. Minorities and low-income children have less
access.
• Safety at School. Minorities are negatively affected by street gangs and fighting in
schools.
• Parent Participation. This is lower in minority groups.
• Frequent Changing of Schools. Minorities change schools more often.
• Low Birth Weight. Blacks have lower average birth weights.
• Environmental damage. Minorities and low-income children have more exposure to lead
and mercury.
• Hunger and Nutrition. These are larger issues among minorities and low-income children.
• Talking/Reading to Babies and Young Children. This is lower among minorities and lowincome children.
• Excessive Television. This is more common among minorities and low-income children.
• Parent-Pupil Ratio. This is lower in minority and low-income households.
• Summer Achievement. Academic progress over the summer is less among minorities and
low-income children.
Two studies published by the Pew Hispanic Center provide particular insight into Hispanic
educational issues. In How Far Behind in Math and Reading Are English Language Learners?
(2007), Rick Fry indicates that English language learners (ELL) score far below the rest of the
population on both math and reading on the NAEP tests. Hispanic ELL scores are lower than
those of other Hispanics.
In Latinos and Education: Exploring the Attainment Gap, Mark Lopez in 2009 reported the
results of a national survey of Hispanics. While nearly 90 percent of young Latinos (16-to-25
years old) — a higher percentage than non-Hispanics — understand that a college education is
important to success in life, less than half plan to get a college degree (a lower percentage than
non-Hispanics). The primary reason for the discrepancy in the percentages is the need to support
a family. Poor English skills also were frequently cited as a reason. As high a proportion of U.S.born Hispanics planned to get a college degree as non-Hispanics, but the figure for Hispanic

	
  

35

immigrants was less than half as high. Financial pressures for immigrants not only include
supporting a spouse and children, but also sending remittances to other family members in their
home country.
Another survey question asked why Hispanics do not do as well in school as others. The most
frequently cited reason was that Hispanic parents do not play an active role in helping their
children succeed academically. Poor English skills and a different cultural background from their
teachers also were commonly cited.
Thus, in addition to the factors affecting educational achievement and attainment that are
common to all children, Hispanic immigrants, particularly those entering the country illegally,
face additional educational challenges. Undocumented immigrants tend to come from countries
with a culture of generally low educational attainment and usually experienced low
socioeconomic status in their home country as well as in the United States. In addition,
immigrant children must learn English as a second language and need to make cultural
adjustments.
Achievement
The 2009 Minority Student Progress Report released by AMEPAC (Arizona Minority Education
Policy Analysis Center), part of the Arizona Commission for Postsecondary Education, provides
a number of measures of educational achievement by race/ethnicity. Results are consistent across
such measures as the AIMS test, enrollment and test scores in Advanced Placement classes, and
eligibility for admission to Arizona universities. Hispanics, blacks, and Native Americans lag
substantially behind whites and Asians on each of these measures. Each of the minority groups is
slowly narrowing the educational achievement gap with whites and Asians.
The differentials in achievement are not as great when looking only at those individuals in each
racial/ethnic group who are interested in pursuing a postsecondary education. Still, Hispanics,
blacks and Native Americans score lower on the SAT and ACT university admissions tests.
Postsecondary enrollment is relatively low among Hispanics and the attainment of college
degrees is low among Hispanics and Native Americans.
Attainment
The achievement measures presented by AMEPAC reflect those students attending K-12 schools
in Arizona. The decennial census/ACS educational attainment measure discussed in the previous
section is broader, including those who moved to the state after completing their education
elsewhere. Since employment-driven immigration begins among those in their mid-teens, even
the educational attainment of 18-year-old Hispanic Arizonans is affected by the low attainment
of many of the immigrants from Mexico and other Latin American countries. Generally,
however, the educational attainment figures are consistent with the achievement measures
presented in the AMEPAC report. Hispanics, blacks and Native Americans have both lower
educational achievement and lesser educational attainment than whites and Asians.
In order to provide more information on the educational attainment of young U.S.-born
Hispanics versus Hispanic immigrants, a special analysis was conducted using the 2005-09 ACS
PUMS file. Sampling error is a major concern when subsetting the population as much as was

	
  

36

done in this analysis. The analysis focused on Hispanics from 16-to-24 years old, with results
produced by single year of age. In addition to educational attainment, school enrollment was
examined. These measures were analyzed by place of birth and (for immigrants) the length of
time in the United States.
Among Hispanics 16-to-24 years old, school enrollment is much higher among those who were
born in the United States than among those born in another country, particularly Mexico. No
difference is apparent between those born in Arizona and those born elsewhere in the United
States. Among those who had immigrated to the United States, school enrollment was positively
correlated with the number of years of residence in the United States. Educational attainment —
whether measured as the percentage with a high school diploma (or GED), the percentage
attending college, or the percentage with a bachelor’s degree — also was greater among those
born in the United States.
For those Hispanics 16-to-24 years old who were living with a parent, enrollment and attainment
also were examined based on the birthplace of the parent. Among those 16-to-21 years old born
in the United States, enrollment was higher among those with a U.S.-born parent than with an
immigrant. Among those 22-to-24 years old, the results were reversed. The results by age of
child are inconsistent for the educational attainment indicator. Sampling error may account for
the inconsistency in results by age; the data do not support any conclusion being reached as to
whether the birthplace of the parents has any influence on the American-born child.
One can conclude that school enrollment and educational attainment are lower among Hispanics
aged 16 to 24 who were born in another country. Thus, Hispanics who moved to the United
States as children are the largest challenge in reducing the Latino educational gap. Otherwise, the
Latino gap appears to be no different from that of other minority groups, or more generally, of
those with low socioeconomic status.

	
  

37

OUTLOOK FOR LATINO EDUCATION
The magnitude of the Latino education issue in Arizona in coming decades will depend largely
on the size of the Hispanic population relative to the state’s entire population. The expansion or
restriction of educational programs tailored to students learning English as a second language or
to Hispanic/minority/low-income students generally also would have an impact.
The fertility rate of Hispanics will play an important role in determining the number of Hispanics
living in Arizona, particularly in the number going to school. However, the single largest factor
that will influence Latino educational achievement and attainment in Arizona in coming years is
the number and characteristics of Latino immigrants who will live in the state. Since few changes
have occurred over the last 20 years in the characteristics of Hispanic immigrants and Arizona’s
Hispanic population as a whole, the significance of the Latino education issue in coming years
will largely be determined by the number of immigrants, particularly undocumented immigrants.
Immigration
Immigration to Arizona in the future is dependent largely on policy decisions made in Arizona,
though certain changes in federal immigration law would have an effect. Economic and
demographic conditions, typically the most important factors influencing immigration —
particularly undocumented immigration — may also play a role, but the presence of Arizona’s
employer sanctions law likely will relegate them to secondary importance.
Court decisions regarding the legality of portions of Senate Bill 1070 could have an impact on
the number of undocumented immigrants living in the state. If the currently blocked portions of
the law are allowed to take effect, some current residents of the state could leave and another
barrier would be put in place to discourage future undocumented immigrants. However, these
effects likely would be much smaller than those of the state’s employer sanctions law.
Policy Scenarios
Arizona. Currently, Arizona has the most restrictive immigration laws among the states, making
it unlikely that many new undocumented immigrants will move to the state. Assuming that
employers obey the employer sanctions law, other than day labor or another source of income in
the underground economy, the only way that an undocumented immigrant could earn money in
Arizona is by obtaining a job through ID theft.
Additional legislation relating to undocumented immigrants could be passed in Arizona and
existing laws could be strengthened or weakened in the future. Probably the most significant
effect on undocumented immigration to Arizona, however, will be the degree of enforcement of
the employer sanctions law.
Given the weak economic conditions that have been in place ever since the employer sanctions
law took effect, its full impact has not yet been tested. As the economy recovers and employers
add jobs, manpower shortages in certain occupations and industries likely will develop that will
not be resolved by individuals moving to Arizona from elsewhere in the United States to fill the
jobs. Some employers then would have a motivation to ignore the employer sanctions law. If the
law is not strongly enforced, some employers may hire undocumented workers, increasing
undocumented immigration to Arizona and worsening the Latino education issue.

	
  

38

If the law is vigorously enforced, the penalties seem likely to keep most employers in
compliance. The most likely scenario is that the employer sanctions law will be adequately
enforced, motivating employers to follow the law and discouraging undocumented immigrants
from moving to Arizona.
Federal. The likelihood of substantive changes being made to existing federal immigration law
anytime soon seems low, based on the unwillingness of Congress to address the issue over the
last decade and the remaining dichotomy between those desiring to restrict immigration further
and those wanting a policy that is more in tune with the workforce needs of the American
economy.
Current federal restrictions on undocumented immigration could be strengthened by the federal
government by enforcing the existing provision that it is illegal to knowingly hire or recruit
unauthorized immigrants. This could be accomplished by requiring all employers nationally to
use E-Verify. Though Arizona’s employer sanctions law would no longer be unique, passage of
such a national law would not likely have much effect on the number of undocumented
immigrants living in the state.
Another means of restricting undocumented immigration is to tighten border security even
further. The effectiveness and practicality of this strategy is in doubt, since it would take
enormous resources to truly control the border. Further, since many undocumented immigrants
entered the United States legally on a student or tourist visa, this initiative would need to include
a way to find and deport those who overstay their visa (and/or to greatly limit the number of such
visas). With state laws already discouraging undocumented immigrants from permanently
locating in Arizona, this federal strategy likely would have limited effect on the number of
immigrants living in Arizona.
A very different direction for immigration reform would be to adjust the overall legal
immigration quota over time to match changes in workforce conditions in the United States. This
could be done by using easily predicted long-term changes in the age distribution and making
short-term projections of the strength of the national economy. In this scenario, Arizona likely
would receive a larger number of legal immigrants, but the characteristics of these immigrants
probably would not be different from those of the state’s existing undocumented immigrants.
This would aggravate the Latino education gap.
Rather than adjust the number of legal immigrants permanently admitted, guest worker programs
could be expanded. Currently, the H-1B visa program allows aliens to work in the United States
for a limited number of years. This program requires an employer-employee relationship and is
designed for workers in specialty occupations, particularly those requiring substantial
educational attainment in fields such as mathematics and science.
A similar program could be created for the labor shortages that likely will redevelop in some of
the lowest-paid occupations or in physically demanding and difficult jobs. Undocumented
immigrants, largely from Mexico and Latin America, have filled these openings in the past.
Whether such a guest worker program is feasible is doubtful. Many of the nation’s
undocumented residents originally entered the country legally. Moreover, since most of the labor

	
  

39

shortages of this nature are likely to persist for many years, temporary workers are not a good fit.
A guest worker program likely would result in an increase in the number of Hispanics living in
Arizona, adding to the Latino education challenge.
Other States. If the federal government does not enact significant changes to existing
immigration law, then more states may pass laws similar to those enacted in Arizona. If many
states, especially populous and proximate states, pass state legislation similar to that in Arizona,
the state no longer would stand in isolation. However, with Arizona’s employer sanctions law in
place, actions taken in other states would be unlikely to cause a significant change in the number
of undocumented immigrants living in Arizona.
Economic Scenarios
If immigration laws in the United States were not an issue and if Arizona had not passed its antiimmigration legislation, then economic and demographic conditions in the United States and in
the countries from which immigrants are moving would be the most important factors affecting
immigration to Arizona in the future. As discussed earlier, changes in the age distribution in both
the United States and Mexico almost certainly will result in less undocumented immigration in
coming decades. With the age distribution of the working-age population essentially fixed for the
next 20 years, forecast scenarios will not vary much based on demographics.
Other economic factors in the United States and Mexico (and other Latin American countries)
could cause variations in the amount of immigration. Though unlikely, the possibility of
substantial social and economic instability exists in Mexico, due to the power of the drug cartels
and the government’s efforts to fight the cartels. If instability were to occur, emigration from
Mexico almost certainly would increase.
The rate and nature of economic growth in the United States also will affect future
undocumented immigration. Very fast growth — especially in occupations that are not attractive
to American-born workers — would lead to larger labor shortages and more opportunities for
undocumented immigrants.
The relevance of the following discussion of economic growth in Arizona and the number of
immigrants moving to the state is questionable. The existence of the employer sanctions law
likely will prevent much of a response from undocumented immigrants to labor shortages in the
state.
The course of the Arizona economy hardly is certain, particularly as one looks ahead for
decades. In the near term, it is highly likely that the Arizona economy will gradually recover and
that job growth will accelerate. Within a few years, the unemployment rate is likely to lower
substantially, creating a need for individuals to migrate to the state, as in the past, to fill the jobs
created. Whether economic growth will be as fast as in prior economic expansions is unclear.
Similarly, the nature of the growth is uncertain. Of particular interest are those industries that
historically have been heavily dependent on immigrants for labor, including agriculture,
construction and tourism. In a baseline scenario, the rate and nature of economic growth during
the next expansion is expected to be similar to that experienced in Arizona in recent decades.

	
  

40

Uncertainty increases the further into the future one looks. At some point, economic growth will
slow in Arizona. The norm for any area is to experience rapid growth for a period of time,
followed by a longer period of much slower growth. The primary reason for this life cycle is that
as major urban areas become more populous, the negative effects of growth become increasingly
apparent while the advantages of growth dissipate.
The Phoenix area already has a population of more than 4 million and is one of the most
populous metropolitan areas in the nation. Like any other large and growing urban area, the
Phoenix area already faces a variety of issues, including air quality, traffic congestion, and
housing being built many miles from job centers in order to improve housing affordability. If the
Phoenix area resumes fast growth, these issues will increase in significance.
The actions of the Arizona Legislature over the last two decades to sharply reduce state
government revenues, which caused major budget reductions in the last few years, could affect
both the quantity and quality of the state’s future economic growth. Even before the recent
budget cuts, the state’s K-12 educational outcomes already were among the lowest in the nation,
with per pupil spending nearly the lowest in the nation. Significant reductions in investments for
higher education also have occurred in recent years. At some point, employers offering highquality jobs likely will refuse to relocate to, or expand in, Arizona due to concerns regarding the
quality of education and the quality of the workforce.
Future investments in other types of infrastructure also are a concern. An environment in which
taxes continue to be reduced does not bode well for the ability of the state to provide the public
infrastructure and other public services required by quality employers. For example, an
inadequate transportation system is more than an inconvenience to commuters — it is a cost to
employers.
A slowing of economic growth in Arizona would result in reduced immigration and domestic
migration to the state. Fewer opportunities would exist for immigrants, and some U.S.-born
Hispanics might choose to leave the state. Whether this would be accompanied by a change in
the Hispanic educational gap is unclear. A lack of higher-paying jobs would be a disincentive to
some youths, regardless of race/ethnicity, to concentrate on their education.
Fertility
Hispanic fertility rates in Arizona fell considerably between 2000 and 2010, by much more than
the decrease in other racial/ethnic groups. Though still substantially higher than the rates of nonHispanic whites and Asians, Hispanic fertility rates are comparable to those of blacks and less
than those of American Indians. Further, the fertility rate of U.S.-born Hispanics is not much
higher than that of non-Hispanic whites.
There are three possibilities regarding Hispanic fertility rates in Arizona in the future:
• Hispanic fertility rates will continue to fall more than those of other racial/ethnic groups,
akin to the large declines in fertility rates in Mexico. This would reduce the significance
of the Latino education issue.

	
  

41

•
•

The disproportionately large decrease in the Hispanic fertility rate in the last decade was
a one-time phenomenon. If so, Hispanic fertility rates likely will change in the future at
the same pace as those of other racial/ethnic groups.
The large decrease in fertility rate in the last decade was an aberration due to the
economic recession combined with the uncertainty caused by Arizona’s anti-immigrant
legislation. In this scenario, Hispanic fertility rates could increase more than those of
other racial/ethnic groups once the economy improves and the justice system determines
the fate of S.B. 1070. This would elevate the significance of the Latino education issue.

The probabilities of each of these three scenarios depend on many factors. Regardless of the
scenario, the number of Hispanic immigrants moving to Arizona in the future likely will be a
significant factor.
Hispanic Education
In the most likely scenario, the educational achievement and attainment of Hispanics will
continue to rise at a slightly greater rate than that of non-Hispanic whites. However, given the
magnitude of the existing gap, it will be many years before Hispanics begin to approach nonHispanic whites in educational achievement and attainment. To date, most of the improvement in
attainment has occurred at the low end of the scale — relatively more Hispanics are graduating
from high school. Little change has occurred relative to non-Hispanics in the percentage of
college graduates.
The slow improvement in educational achievement and attainment likely is associated with the
slight progress in economic well-being that has occurred for Hispanics relative to non-Hispanics.
In addition, the educational gains that have taken place in Mexico have helped to raise the
achievement and attainment of immigrants.
In an optimistic scenario, widespread adoption of programs to improve the educational outcomes
of Latinos, such as the Center for the Future of Arizona’s “Beat the Odds” program, would close
the gap more quickly. In contrast, if existing efforts are scaled back, such as funding for students
learning English as a second language, the gap could widen in a pessimistic scenario. The
probabilities of each scenario will depend in part on the number of immigrants entering the state.

	
  

42