University of Southern California Dissertations and Theses

The economic security of an aging minority population: a profile of Latino baby boomers to inform future retirees

(USC Thesis Other)

The economic security of an aging minority population: a profile of Latino baby boomers to inform future retirees

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

THE ECONOMIC SECURITY OF AN AGING MINORITY POPULATION:

A PROFILE OF LATINO BABY BOOMERS TO INFORM FUTURE RETIREES

by

Zachary Demetrius Gassoumis

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(GERONTOLOGY)

May 2014

Copyright 2014 Zachary Demetrius Gassoumis

ii

ACKNOWLEDGMENTS

This dissertation was completed with financial support from the U.S. Social
Security Administration, via the Dissertation Fellowship Program administered by the
Center for Retirement Research at Boston College. Early stages of the research contained
in the dissertation were conducted with funding from the Ford Foundation (grant numbers
2007-2505 and 1070-1233-1) and from the USC Edward R. Roybal Institute on Aging. I
am grateful to these organizations and their staff for the support they provided throughout
my dissertation process.
The contributions over the past eight years of the Latinos & Economic Security
research team—primarily Fernando Torres-Gil, Chon Noriega, and Max Benavidez—
have been numerous and absolutely foundational. They have provided immense and
invaluable guidance and direction on all topics Latino, as well as the most enjoyable
research team I expect I’ll ever have the pleasure of working with. Further support on
this and other projects has also been fundamental to my doctoral education, both from
faculty and student collaborators. I am grateful to Adria Navarro, Mike Nichol, Kathryn
Thomas, Susan Enguidanos, Natalie Leland, Marti DeLiema, Gretchen Alkema, Diana
Homeier, Jeanine Yonashiro-Cho, Yongjie Yon, and many others over the years for
wonderful professional relationships as well as great personal friendships. I am indebted
as well to all of the faculty and staff, both current and past, at the USC Davis School of
Gerontology for their many years of support and encouragement.
My family—both immediate and extended, both biological and chosen—has been
a source of inspiration, encouragement, motivation, and occasionally much-needed
iii

distraction. I especially acknowledge my parents, for giving me the tools to succeed in
my educational pursuits: my father, for passing to me his structured and logical approach
to task he takes on, as well as an unwitting but early exposure to the complex world of
aging; and my mother, for nurturing in me an unfettered curiosity and leading by example
with an unparalleled amount of energy and drive. Among my chosen family, the
steadfastness and perseverance of Athan and Natalie, no matter what circumstances
crossed our paths, proved to be instrumental in getting me across the finish line with my
sanity mostly intact. Finally, to Kate Wilber (and family)... for years of patience,
guidance, leadership, occasional sternness, immense freedom, and persistent support.
You have not only made me a better scholar, but a better person. I owe you a debt of
gratitude for this dissertation, and for so much more.

iv

TABLE OF CONTENTS

Acknowledgments ii

List of Tables vi

Abstract vii

Chapter 1: Introduction 1
Characteristics of Latinos 2
Financial Strain Among Latinos 4
The Hidden Population: Latino Baby Boomers 5
The Demographic Imperative 5
The Historical Context: Latin American Migration into the U.S. 7
Review of Theories 11
Structure of Dissertation 12

Chapter 2: Characteristics of Latino Baby Boomers, by Ethnicity and Citizenship
Status 14
Introduction 14
Characteristics of Latinos 16
Financial Strain Among Latinos 18
Latino Baby Boomers 19
Methods 20
Measures 21
Results 25
Population Size & Citizenship Status 25
Characteristics in 2000 28
Comparing Historical Characteristics 32
Discussion 36
Policy Implications 38
Conclusion and Future Directions 40

Chapter 3: Income and Wealth Disparities: Structural Inequalities for Baby Boomers
and Across Cohorts 41
Background 41
The Structural Inequality of Race/Ethnicity 44
Cumulative Inequality 44
Mechanisms of Human Agency: Education and Naturalization 45
Hypotheses 46
Methods 47
Measures 49
Analyses 50
Results 51
Individual Income and Household Wealth, by Race/Ethnicity 51
v

Sociodemographic Characteristics 53
Regression Results: Income 56
Regression Results: Wealth 60
Marginal Effects 63
Discussion 65
Limitations 67
Implications 68

Chapter 4: Naturalization’s Effect on Income Growth: Presence and Persistence
Across the Lifespan 71
Introduction 71
Naturalization and Segmented Assimilation 73
Age Differences in Lifespan Income Trajectories 74
Hypotheses 75
Methods 75
Measures 76
Analyses 78
Results 80
Growth Curve Results 84
Marginal Effects 90
Discussion 91
Limitations 95
Implications 97

Chapter 5: Conclusion 98
Implications for Theory 99
Implications for Research and Policy 101
Concluding Remarks 103

References 105
vi

LIST OF TABLES AND FIGURES

Figure 1.1: Key Points in the History of Immigration between Latin America
and the U.S. 7

Table 2.1: Analysis variables and descriptions 23

Table 2.2: Citizenship status of baby boomers, by ethnicity and race, 1980-2008 26

Table 2.3: Baby boomer demographic and economic characteristics, by ethnicity
and citizenship status, 2000 29

Table 2.4: Select characteristics of Latino baby boomers, by citizenship status,
1980-2008 33

Table 3.1: Income and wealth, by cohort and race/ethnicity 52

Table 3.2: Demographic and economic characteristics, by analysis dataset/cohort 54

Table 3.3: Linear regression on income (IHS-transformed), by cohort 58

Table 3.4: Linear regression on wealth (IHS-transformed), by cohort 61

Table 3.5: Marginal effects of race/ethnicity on income and wealth, by cohort 64

Table 4.1: Demographic and economic characteristics 81

Table 4.2: Demographic and economic characteristics of younger (age 25-39)
immigrants, by citizenship status 82

Table 4.3: Demographic and economic characteristics of older (age 40-64)
immigrants, by citizenship status 85

Table 4.4: Linear growth curve models, full sample and by age group 88

Table 4.5: Marginal effects of citizenship status on income (intercept and slope),
full sample and by age group 91

vii

ABSTRACT

The United States is facing dramatic demographic changes due to the aging of the
Baby Boom Generation and increasing diversity, including rapid growth of the Latino
population. Questions have been raised regarding the economic security of the aging
baby boomers’ generational cohort once they retire, which are of particular relevance to
minority and Latino members of the cohort. Latinos tend to have lower levels of
financial security than their white, non-Latino counterparts, but there is little research that
examines individuals who fall into the intersection of these two groups: the Latino baby
boomers. Because Latino boomers are a largely hidden population, their economic status
and prospects are difficult to estimate. After laying out the historical and theoretical
contexts in an overarching introduction, this dissertation integrates three empirical
chapters to advance knowledge in this area, by: 1) laying out selected sociodemographic,
economic, and health characteristics of the cohort and drawing implications for national
social insurance policies; 2) assessing the degree of economic disparity between
racial/ethnic groups that is due to membership in that racial/ethnic group as opposed to
disparities in sociodemographic status, and how these dynamics have changed between
generational cohorts; and 3) addressing the degree to which these economic disparities
are diminished due to the acquisition of citizenship.
The first empirical chapter (Chapter 2) looks at the characteristics of the Baby
Boomer population living in the 50 U.S. states and the District of Columbia, broken
down by Latino ethnicity and citizenship status. Drawing from several U.S. Census
Bureau data sources, it revealed three key findings: 1) there were 80 million baby
viii

boomers in the U.S. in 2000—more than previously reported—of which 8.0 million
(10%) were Latinos; 2) U.S.-born Latino boomers were more similar to non-Latino
boomers in terms of demographic characteristics, whereas foreign-born citizens and non-
citizens scored less well on key demographic indicators; and 3) compared to non-Latino
baby boomers, U.S.-born Latino baby boomers had somewhat less favorable economic
characteristics.
The second empirical chapter identifies the magnitude of racial/ethnic structural
disadvantage for income and wealth in the years preceding retirement for the Baby Boom
Generation, then compares their structural disadvantage with that of members of the
Silent Generation cohort when they were the same age. After adjusting for
sociodemographic variables (age, gender, citizenship status, education, marital status, and
labor force participation), the structural effects of race/ethnicity on income—using the
American Community Survey—and wealth—using the Health and Retirement Study—
were considerably reduced, confirming two of the chapter’s four hypotheses; however,
the expected reduction in structural effects from the Silent Generation to the Baby Boom
Generation was seen for wealth but not for income, confirming only one of the remaining
two hypotheses. This reduction of structural disparities in wealth from the Silent
Generation to the Baby Boom Generation follows the expectation that these disparities
would be reduced over time, which signals good news for the younger members of the
Baby Boom Generation, Generation X, and future generational cohorts. But large gaps
still exist between racial/ethnic groups, even after sociodemographic adjustment; future
reduction in those structural inequalities can help decrease those gaps, an especially
important consideration for low-income racial/ethnic minority groups.
ix

The third empirical chapter takes an initial step toward disaggregating by age the
effect of naturalization on income growth. Using linear growth curve modeling on data
from the Survey of Income and Program Participation’s 2004 panel, it attempts to
replicate past findings across the entire lifespan, but fails to detect an effect of
naturalization on income growth; only non-citizens had a significantly higher level of
income growth during the study period than U.S.-born citizens. In subsetting the analysis
for older and younger working-age groups, an effect of naturalization was not detected
for either group, and the positive effect for non-citizens was seen only for the younger
age group. The predictor variables on the whole had minimal relationships with slope in
the model, with less than 1% of variance explained in each model. Although a stronger
effect of the predictor variables, including an effect of naturalization, may have appeared
were more years of data available, it was not detected over the 4-year study period. Two
unexpected findings were: 1) individuals in the younger sample who had naturalized
before the study had higher intercepts than U.S.-born citizens but no such difference
emerged in the older sample; and 2) in a bivariate context, those who naturalized during
the study represented a socioeconomic midpoint of sorts—on racial/ethnic composition,
education, and income—between non-citizens and those who had naturalized prior to the
study.
In sum, these chapters shed light on the Baby Boom cohort’s characteristics and
dynamics in the period leading up to their retirement age. This dissertation provides
insights into the characteristics, demographic history, and socioeconomic patterns of the
upcoming cohort of retirees. Implications of these findings have the potential to inform
and to modify practice and policy for the next cohort: Generation X. The findings
x

underscore the importance of reducing disparities in education and, to a degree,
citizenship as a mechanism for countering the persistent effects of structural inequality on
income. These insights have implications for both theory and policy and lay a foundation
for a wide range of future research, which is discussed in the final chapter.
1

CHAPTER 1: INTRODUCTION

The United States is facing dramatic demographic changes due to the aging of the
Baby Boom Generation and increasing diversity, including rapid growth of the Latino
population. These changes have occurred in the context of the most severe economic
downturn in decades, raising many questions regarding the economic security of the
aging baby boomers’ generational cohort once they retire. Such questions are of
particular relevance to minority and Latino members of the cohort, since they tend to
have lower levels of financial security than their white, non-Latino counterparts
(Gassoumis, Wilber, Baker, & Torres-Gil, 2010).
As part of its efforts to provide greater long-term stability to the economy, the
Obama administration has put the reform of social insurance programs, primarily Social
Security and Medicare, on the policy agenda. Without reform, Social Security is
expected to have exhausted reserves by 2033, after which benefits would need to be paid
at 77% of their promised value (Board of Trustees, 2013). At the same time, in the face
of eroding pensions and investments, Social Security is likely to provide an increasingly
crucial source of support for the retiring baby boom generation. Compared to the U.S.
population as a whole, Latinos are more likely to rely solely on Social Security for
retirement income (Halliwell, Gassoumis, & Wilber, 2007), making this debate especially
important for this population.
Although population aging, driven by the large baby boom generation, and
increasing ethnic diversity are significant and historic trends, there is little research that
examines individuals who fall into both categories: the Latino baby boomers. Because
2

Latino boomers are a largely hidden population, issues such as their economic status and
its effect on financial security and retirement prospects, the economic challenges faced by
immigrants, and the policy implications related to public benefits are difficult to estimate.
Building knowledge about this generational cohort is crucial, not only for supporting
their economic security as they retire, but also for informing the worklife economic
security of subsequent generations of Latinos.

Characteristics of Latinos
Latino refers to individuals of “Mexican, Puerto Rican, Cuban, South or Central
American, or other Spanish culture or origin, regardless of race” (Office of Management
and Budget, 1997). Over half of Latinos are of Mexican heritage, with the remaining
members spread among Puerto Rican, Cuban, Dominican, and other national origins
(Guzmán, 2001); 53% of Latinos in the United States aged 65 and older in 2006 were
born in the U.S. (Pew Hispanic Center, 2008). In 2005, it was estimated that 67% of the
foreign-born Latino population was in the U.S. on a legal, permanent basis, and that only
52% of this population had become naturalized citizens (Passel, 2007).
The primary indicators of socioeconomic status tend to be lower for the Latino
population than the U.S. population as a whole. For example, in 2006, 59% of Latinos
over age 25 had a high school degree or higher compared to 81% of non-Latino blacks
and 91% of non-Latino whites (Snyder, Dillow, & Hoffman, 2007). This is a vast
improvement from the 1980s, however, when the same measure generated scores of 45%
for Latinos, 51% for non-Latino blacks, and 72% for non-Latino whites (Snyder et al.,
2007). Lower education levels are associated with lower incomes. A recent report
3

showed median Latino personal earnings as roughly $20,000 compared with $23,000 for
non-Latino blacks, $30,000 for non-Latino whites, and $31,000 for non-Latino Asians
(Pew Hispanic Center, 2008).
In contrast, Latinos tend to fare better than the U.S. population as a whole on a
crucial indicator: life expectancy. In 2011, Latino men aged 65 could expect to live to
85, three years longer than the average 65 year old male in the U.S. Similarly, 65 year
old Latino females had a life expectancy of 89 years, compared to 85 years for the
average 65 year old female (U.S. Social Security Administration, 2013). The presence of
this mortality advantage in spite of performing worse on most sociodemographic
indicators, referred to as the “Hispanic paradox” (Markides & Eschbach, 2005), does not
apply to morbidity. Latinos tend to have a higher prevalence of disability (e.g., Rudkin,
Markides, & Espino, 1997) and functional limitations (Dunlop, Song, Manheim,
Daviglus, & Chang, 2007) than non-Latino whites. In addition, recent studies have
demonstrated that the Hispanic paradox does not apply to all subgroups of the Latino
population, suggesting that more work is needed to determine the underlying causes of
the mortality differential. Specifically, the paradox seems to be primarily applicable to
foreign-born individuals who are not from Cuba or Puerto Rico (Palloni & Arias, 2004).
Contrary to the positive characteristics previously associated with native-born
populations, it appears that U.S.-born older Latinos of Mexican descent may in fact be
worse off than the non-Latino population (Crimmins, Kim, Alley, Karlamangla, &
Seeman, 2007).

4

Financial Strain Among Latinos
Increased life expectancy coupled with higher levels of disability place additional
financial requirements on Latino older adults and their families. Compared to the U.S.
population as a whole, Latinos are less likely to have savings accounts and retirement
accounts, and those who have saved have likely accumulated balances that are inadequate
for their retirement needs (Ibarra & Rodriguez, 2006; Orszag & Rodriguez, 2005).
Between 1996 and 2002, 26% to 29% of Latino households exhibited zero or negative net
worth, compared to 11% to 13% for non-Latino white households (Kochhar, 2004). This
leaves them highly vulnerable to economic fluctuations and unforeseen expenses and
foreshadows dependence on public programs for income maintenance in retirement. In
light of the current economic downturn, savings and wealth accumulation is likely to be
further diminished.
Latinos’ low income levels and lack of savings from their working years lead to
aggravated economic problems in old age. In 2004, 43% of Latinos aged 65 and over
were classified as being either near-poor or poor, with incomes below 150% of the
poverty threshold. This rate was nearly twice that of the general U.S. population aged 65
and over, of which a quarter fell below 150% of the poverty threshold (Beedon & Wu,
2004). With the high level of hardship, a disproportionate number of older Latinos have
come to depend on public programs such as Social Security for most of their monthly
income. According to the Social Security Administration, 61% of unmarred older
Latinos and 44% of married older Latino couples received at least 90% of their income
from their earned Social Security benefits in 2011 (U.S. Social Security Administration,
2013), underscoring the importance of Social Security to this segment of the population.
5

As stated above, the U.S. Latino population includes many non-citizens, a
characteristic that may be linked to greater levels of financial strain. Much of the large
literature that focuses on economic disparities by race and ethnicity reports stark
variations based on citizenship status. While few studies have directly linked citizenship
and citizenship acquisition to financial security, existing research suggests that a positive
relationship exists between citizenship acquisition and income (e.g., Bratsberg, Ragan, &
Nasir, 2002; Mazzolari, 2009).

The Hidden Population: Latino Baby Boomers
The next 20 years will be marked by the steady retirement of members of the
baby boom generation. Born between 1946 and 1964, “boomers” represent the largest
generational cohort in U.S. history. Distinct in its high level of racial and ethnic
diversity, the aging of the baby boomers will lead to a larger number of Latino retirees
than has previously been seen in the U.S. As a hidden population and one that is growing
in both size and influence, more information is needed about the Latino baby boomer
population, its strengths, and its anticipated needs in retirement. Information about the
economic security of Latino baby boomers will provide more insights into the financial
landscape that can be expected for the emerging Latino population.

The Demographic Imperative
Numbering over 52 million people in 2012, the Latino population comprises the
largest racial/ethnic minority group in the U.S. (17% of the total population). One
striking feature is that the Latino population is considerably younger than the U.S.
6

population in general. While 14.9% of the working age population (aged 25-64) in 2012
was Latino, only 6.9% of the population aged 65 and older in 2012 was Latino (U.S.
Census Bureau, 2013).
The number of Latinos in the U.S. is increasing with remarkable speed, projected
to grow to 28% of the total U.S. population by 2050. During this period, the number of
working-age Latinos is expected to increase by over 120%, from 24.4 million in 2012 to
a projected 54.4 million in 2050. This compares to just a 2% projected increase in the
non-Latino workforce during this period, from a current 139 million to a projected 142
million in 2050 (U.S. Census Bureau, 2012b, 2013).
Based on these projections, the growth of the U.S. job force and economy is
expected to be driven over the next 40 years almost exclusively by the expanding Latino
population. And due to the funding structure of social insurance programs for retired
Americans, the economic wellbeing of the 65 and older population in the U.S. will be
increasingly tied to the economic success of Latino workers.
One of the most compelling U.S. policy areas related to the Latino population is
immigration—particularly documentation and naturalization (i.e., attainment of
citizenship) among the immigrant population. The topics of immigration and how to
address immigrants in the U.S. are fiercely debated, most notably in recent years with the
DREAM Act and Arizona’s SB 1070 (the Support Our Law Enforcement and Safe
Neighborhoods Act). As a large proportion of the current Latino workforce and retiree
population were born outside the U.S., documentation and naturalization are important to
consider when exploring socioeconomic status and economic security among the Latino
population.
7

The Historical Context: Latin American Migration into the U.S.
To fully understand the dynamics of aging and Latinization in the United States, it
is necessary to first understand the relationship between Latinos and the U.S. The Latino
population in the U.S. is a largely immigrant group; as such, it is important to highlight
the recent history of migration between the U.S. and Latin America (see overview in
Figure 1.1).
Although Latin American migration has always been an important consideration
in U.S. immigration policy, Latinos have often been subjected to different laws than
immigrants from other areas. Variations in immigration policies reflect the economic
dependency of the U.S.—particularly the agriculture industry—on migrant workers from
Latin America, largely Mexico (Massey, Durand, & Malone, 2002). From the 1848
acquisition of the American southwest with the Treaty of Hidalgo through the 1924
Immigration Act, migration across the newly established border had few restrictions for

Figure 1.1. Key Points in the History of Immigration between Latin America and the U.S.
8

Mexican migrants. Although costs and exclusions were put in place on those wishing to
obtain lawful permanent status, Mexican nationals were generally free to move across the
border as temporary laborers (Cruz & Carpenter, 2011), though a parallel system of
indentured labor was also prevalent during this period (Massey et al., 2002).
This open immigration policy with Mexico changed dramatically, however,
beginning with the 1924 Immigration Act. The Border Patrol was created, and visas,
migrant inspections, and deportations were instituted. Enforcement was often
indiscriminate and even targeted U.S.-born citizens. The policies and negative climate in
this period led to voluntary exits on a large scale, but labor demand in the U.S. was strong
enough for the flow of migration to soon resume in spite of the new policies (Cruz &
Carpenter, 2011; Massey et al., 2002). After the onset of the Great Depression, however,
joblessness was high, markets were flooded with an abundance of laborers, and migrant
populations were scapegoated. Massive deportations were instituted, which led to a
period with almost no net immigration from Mexico. The policy climate that had
previously encouraged the migration of Mexican laborers suddenly changed, and these
migrant populations were vilified (Massey et al., 2002).
With the end of the depression, the ample domestic labor supply available in the
U.S. quickly dried up, a state that was exacerbated by the U.S. entry into World War II.
This led to the Bracero Program, a system of temporary contract workers, primarily for
farm and agricultural work (“growers”). But the restricted supply of Braceros—or
farmhands—allowed for by the U.S. government was not enough to meet demand by
growers, which led to a dramatic growth in the undocumented immigrant population
9

(Massey et al., 2002). Following the Korean War, undocumented immigration became a
prominent political issue, despite widespread approval from the growers. In response to
growing public opposition over the influx in undocumented workers, the Immigration and
Naturalization Service (INS) formed “Operation Wetback” in 1954, in which
undocumented workers were deported to a Bracero processing site and immediately
processed as new, documented Braceros (Massey et al., 2002). Under Operation
Wetback, the U.S. government was effectively handing out work permits to
undocumented immigrants, but doing so in a way that kept the nativist public elements
under the slightly misled perception that deportations abounded.
With the rise of the civil rights movement in the early 1960s, the Bracero Program
began to be phased out. Coupled with marked changes to the U.S. immigration system
(e.g., the removal of Asian exclusion), the Bracero Program was ended in 1965 and a new
age of immigration began. Immigrant quotas were put in place, limiting the number of
legal migrants who could enter the U.S. from Mexico; however, the Bracero Program had
left the agriculture sector heavily dependent on foreign workers. This resulted in a
context of migration and workforce demand in which migrants, who were eager to cross
the border and fill the jobs left after the dissolution of the Bracero Program, began to use
undocumented immigration as a primary form of migration into the U.S. (Massey et al.,
2002). During this period, the border crossings operated as revolving doors, at which
only about one third of attempted crossings were unsuccessful (Massey & Pren, 2012).
A new age of migration was entered once again with the passage of the
Immigration Reform and Control Act (IRCA) or 1986. IRCA disincentivized future
undocumented migration by doubling the border patrol’s budget, imposing sanctions
10

against employers who knowingly hired undocumented workers, and providing the
Department of Labor with the resources to enforce the employer sanctions. At the same
time, it provided a one-time amnesty opportunity for undocumented residents. As a
result of these provisions, IRCA had the effect of changing the migrant population to the
U.S. from a pattern of temporary workers seeking employment to a system of permanent
residents seeking to settle (Massey et al., 2002).
The years since IRCA have seen a transformation in the policy approach to
immigrant populations. There have been gradual legal and policy moves to increase the
militarization of the U.S.-Mexico border and provide fewer opportunities to both
undocumented and legal residents, primarily via the Personal Responsibility and Work
Opportunity Act (PRWORA) of 1996. Since the Illegal Immigration Reform and
Immigrant Responsibility Act of 1996, deportations have also been steadily increasing
(Massey & Pren, 2012), reaching their highest levels in 2012 under President Obama’s
administration (U.S. Immigration and Customs Enforcement, 2013; Wadhia, 2013).
Perhaps due to this hostile immigrant environment, the “Dreamer” movement emerged
over the first decade of the 21
st
century, with a goal of rectifying the status for young
undocumented immigrants, most of whom were brought to the U.S. as children. Despite
increased public awareness, this movement was not able to secure legislative action; but
with an Executive Order in 2012, the Obama administration enacted the Deferred Action
for Childhood Arrivals (DACA) program, setting in place a partial remedy to the
undocumented status of many young adults (Wadhia, 2013).
Across the history of the U.S. southwest, the approach to immigration has been
largely varied. Through a system of contradictory approaches and policies, a culture of
11

undocumented status has emerged and prevailed. Current immigrant groups, with a focus
on settling, have taken steps to trying to attain documentation status and put themselves
on a road to naturalization. While this dissertation does not address the question of
documentation status, it characteristics the immigrant population in general—with a
focus on Latino immigrants—and identifies characteristics related to economic security
among this population.

Review of Theories
Several theories will be used in this dissertation to frame the issues of age,
race/ethnicity, citizenship status, and economic security. These theories are drawn from
and inform a range of social sciences, but the primary disciplines of relevance to these
issues as they will be framed are sociology and gerontology.
The first theory that has a broad reach across the dissertation and, indeed, across
most lifespan treatments of economic security is the theory of cumulative inequality.
Born from the parent theory of cumulative advantage/disadvantage (Dannefer, 1987), the
theory of cumulative inequality states that sociodemographic, economic, and other factors
experienced early in life will have ongoing and cumulative effects throughout the life
course (Ferraro & Shippee, 2009; Ferraro, Shippee, & Schafer, 2009). A crucial
component of the theory is that a life-course trajectory is shaped by the accumulation of
risk and resources, but that an individual’s trajectory can be modified through human
agency (e.g., education or behavior change; Ferraro & Shippee, 2009).
A second theory that is of great importance to any study of racial/ethnic economic
dynamics is the structural inequality of race/ethnicity. This theory outlines the social
12

constraints that exist within a society, and the attainment of individual success within the
context of those constraints. Important drivers of structural inequality can include split
labor market dynamics and their resultant race/ethnicity-based occupational pigeonholing
and discrimination (Bonacich, 1972; Restifo, Roscigno, & Qian, 2013), citizenship-based
employment constraints for immigrant groups (e.g., Bratsberg et al., 2002), and broad-
scale discriminatory workplace dynamics (Green, 2003).
The final theory that is central to the aims of this dissertation is the theory of
segmented assimilation. This theory has its roots in the theory of straight-line
assimilation, which denotes a single upward assimilational trajectory toward a society’s
majority norm. Portes and Zhou (1993), however, saw a need for non-normative
trajectories of assimilation and developed the segmented assimilation model. In it, there
are three primary strategies for assimilation into a host society: 1) assimilating toward the
majority norm, similar to the traditional model of straight-line assimilation; 2) selective
assimilation, in which upward mobility is achieved through embracing one’s cultural
values and solidarity; and 3) “downward assimilation” into a niche within the majority
society that reinforces poverty, such as gang membership (Bloemraad, Korteweg, &
Yurdakul, 2008; Portes & Rumbaut, 2001; Portes & Zhou, 1993). The theory was
originally developed to apply to the immigrant second generation (Portes & Zhou, 1993),
but the typology can be applied directly to the assimilation of immigrants.

Structure of Dissertation
This dissertation is structured using a three-paper format, which includes three
stand-alone publishable empirical papers framed by an introduction and a conclusion.
13

Following this introductory chapter, the first paper addresses the question of what
disparities exist between Latino citizenship groups in the pre-retirement years. This is
accomplished by laying out selected sociodemographic, economic, and health
characteristics of the cohort of primary interest in this dissertation—the Baby Boom
Generation—and drawing implications for national social insurance policies. The next
paper is in response to the question of what degree of the economic disparity between
racial/ethnic groups is due to membership in that racial/ethnic group as opposed to
disparities in sociodemographic status, and how these dynamics have changed between
generational cohorts. This aspect of economic security is addressed by presenting linear
regression models of both income and wealth for the baby boomers, focusing particularly
on the effects of education and citizenship status, and comparing those to comparable
models for the preceding generational cohort. The final paper in the three-paper series
addresses the degree to which these economic disparities are diminished due to the
acquisition of citizenship with the application of longitudinal modeling. This paper uses
linear growth curve modeling to analyze the growth of income over time, focusing
specifically on the role of naturalization and the difference between age groups. The
dissertation concludes with a discussion of the specific findings and their implications for
theory, research, and policy.

14

CHAPTER 2: CHARACTERISTICS OF LATINO BABY BOOMERS,
BY ETHNICITY AND CITIZENSHIP STATUS

Introduction
The United States is in the midst of dramatic demographic change. Simultaneous
and transformational shifts resulting from the aging of the baby boomers and increasing
diversity, including rapid growth of the Latino population, have profound implications
for U.S. policies over the coming years. These changes are occurring in the context of a
third dynamic that portends even more dramatic change: the most severe economic
downturn in decades. As part of its efforts to provide greater long-term stability to the
economy, the administration has put the reform of entitlement programs, including Social
Security and Medicare, on the policy agenda. The outlays for the Social Security
program (the Old-Age, Survivors, and Disability Insurance [OASDI] program) currently
exceed the program’s non-interest income, and without reform its reserves are expected
to be exhausted by 2033 (Board of Trustees, 2013). At the same time, in the face of
eroding pensions and investments, Social Security is likely to provide an increasingly
crucial source of support for the retiring Baby Boom Generation.
The policy debate on entitlement programs will be shaped by the competing needs
of various groups, including the growing Latino population. The federal deficit—which
peaked at $1.55 trillion in 2009 as a response to the recession but has gradually dropped
to an anticipated $1.00 trillion in 2013 (Office of Management and Budget, 2013)—will
have an impact on the ability and willingness of Congress to provide new programs and
financing that may alleviate the vulnerabilities facing Latinos in general. Because
15

Latinos are younger than the U.S. population as a whole, the relative contribution of
Latino workers to the current pay-as-you-go entitlement system is increasing. Moreover,
contributions by undocumented immigrants are made with little hope of remuneration,
bolstering the programs' balance sheets. Even some legal immigrants may be ineligible
for public benefits, depending on the length of time since immigration and their state’s
policies (Hero & Preuhs, 2007). At the same time, compared to the U.S. population as a
whole, Latinos are more likely to rely solely on Social Security for retirement income
(Halliwell et al., 2007). Thus, Latinos may benefit from reforms that strengthen the
Social Security program and expand health coverage.
Although population aging, driven by the large Baby Boom Generation, and
increasing ethnic diversity are significant and historic trends (Torres-Gil & Treas, 2008),
there is little research that examines individuals who fall into both categories: the Latino
baby boomers. Because Latino boomers are a largely hidden population, issues such as
their economic status and its effect on financial security and retirement prospects, the
effect of immigration, and the policy implications related to public benefits are difficult
to project. The purpose of this study is to begin to address these gaps and potential
misperceptions by examining the demographic and economic characteristics of Latino
baby boomers and comparing Latino boomers based on citizenship status. In doing so,
the paper seeks to draw attention to the confluence of aging and diversity, a phenomenon
that is progressively altering the demographic profile of the U.S. Despite widespread
acknowledgement of these two separate demographic shifts, a review of the literature
indicated that there has been little notice of the Latino baby boomers who stand at their
nexus. Thus, this dissertation focuses on the hidden Latino boomer population as a
16

barometer of these twin forces: the aging of a largely white English speaking population,
and a largely young Latino population that is enjoying increased longevity. It begins
with a brief discussion of what we know about the U.S. Latino population before
focusing on the characteristics of Latino baby boomers.

Characteristics of Latinos
Latino in this paper refers to individuals of “Mexican, Puerto Rican, Cuban, South
or Central American, or other Spanish culture or origin, regardless of race” (Office of
Management and Budget, 1997). Existing research indicates that over half of Latinos are
of Mexican heritage, with the remaining members spread among Puerto Rican, Cuban,
Dominican, and other national origins (Guzmán, 2001); 53% of Latinos in the United
States aged 65 and older in 2006 were born in the U.S. (Pew Hispanic Center, 2008).
As a whole, the Latino population scores poorly on the primary indicators of
socioeconomic status. For example, in 2006, 59% of Latinos over age 25 had a high
school degree or higher compared to 81% of non-Latino blacks and 91% of non-Latino
whites. This is a vast improvement from the 1980s, however, when the same measure
generated scores of 45% for Latinos, 51% for non-Latino blacks, and 72% for non-Latino
whites (Snyder et al., 2007).
Low education levels are reflected in low incomes. A recent report showed
median Latino personal earnings as roughly $20,000 compared with $23,000 for non-
Latino blacks, $30,000 for non-Latino whites, and $31,000 for non-Latino Asians (Pew
Hispanic Center, 2008). Low-paying jobs often provide minimal pension and healthcare
benefits or no benefits at all. Latinos also tend to have lower rates of health insurance
17

coverage. For example, an analysis of U.S. Census data by Reschovsky, Hadley, and
Nichols (2007) found a 22-point disparity in the proportion of uninsured between the
Latino and non-Latino white populations. Between 1987 and 2004, the proportion of
Latinos who were uninsured ranged from 30% to 35%, whereas the proportion was 10%
to 12% for non-Latino whites and 19% to 22% for Non-Latino blacks. Rates of insurance
are believed to be tied to income and education; Baughman (2005) reported that nearly
half of poor full-time workers have no insurance, a rate that far exceeds the population
average of approximately 15% (Reschovsky et al., 2007).
Research suggests that early opportunity, advantage, and disadvantage are often
cumulative, leading to increased disparity between individuals on indicators of health,
disability, and financial security over the life course (Dupre, 2007; Gregoire et al., 2002).
“Cumulative disadvantage” refers to the ongoing and compounding effect of multiple risk
factors, such as low income, lack of education, and poor access to health care, on people
as they age (Dannefer, 1987). Although Latinos have long been known to have a higher
prevalence of disability than non-Latino whites (e.g., Rudkin et al., 1997), a recent study
illustrates the role of cumulative disadvantage by suggesting that these high disability
rates are largely a result of differences in health and socioeconomic factors (Dunlop et al.,
2007). Cumulative disadvantage has been linked to less health coverage (Ponce,
Cochran, Mays, Chia, & Brown, 2008), poorer health outcomes (Shuey & Willson,
2008), and fewer economic resources (Gregoire et al., 2002).
Despite their cumulative disadvantage, Latinos have a longer life expectancy than
the U.S. population as a whole. In 2011, Latino men aged 65 could expect to live to 85,
three years longer than the average 65 year old male in the U.S. Similarly, 65 year old
18

Latino females had a life expectancy of 89 years, compared to 85 years for the average 65
year old female (U.S. Social Security Administration, 2013). Because this mortality
advantage, referred to as the “Hispanic paradox” (Markides & Eschbach, 2005),
contradicts the conventional association between low socioeconomic status and negative
health outcomes, researchers have proposed various explanations, including data
insufficiency, “salmon bias,”—that less healthy Latinos return home to die—healthy
migrant effects, and low-risk lifestyles (Franzini, Ribble, & Keddie, 2001). Recent
studies, demonstrating that the Hispanic paradox does not apply to all subgroups of the
Latino population, suggest that more work is needed to determine the underlying causes
of the mortality differential. Specifically, the paradox seems to be primarily applicable to
foreign-born individuals who are not from Cuba or Puerto Rico (Palloni & Arias, 2004).
Contrary to the positive characteristics often associated with native-born populations, it
appears that U.S.-born older Latinos may in fact be worse off than the non-Latino
population (Crimmins et al., 2007).

Financial Strain Among Latinos
Increased life expectancy coupled with higher levels of disability place additional
financial requirements on Latino older adults and their families. Not surprisingly,
compared to the U.S. population as a whole, Latinos are less likely to have savings
accounts and retirement accounts, and those who have saved have likely accumulated
balances that are inadequate for their retirement needs (Ibarra & Rodriguez, 2006; Orszag
& Rodriguez, 2005). Between 1996 and 2002, 26% to 29% of Latino households
exhibited zero or negative net worth (Kochhar, 2004), leaving them highly vulnerable to
19

economic fluctuations and unforeseen expenses and foreshadowing dependence on public
programs for income maintenance in retirement. In light of the current economic
downturn, savings and wealth accumulation is likely to be further diminished.
Latinos’ low income levels and lack of savings lead to aggravated economic
problems in old age. In 2004, 43% of Latinos aged 65 and over were classified as being
either near-poor or poor, with incomes below 150% of the poverty threshold. This rate
was nearly twice that of the general U.S. population aged 65 and over, of which a quarter
fell below 150% of the poverty threshold (Beedon & Wu, 2004). With the high level of
hardship, a disproportionate number of older Latinos have come to depend on public
programs such as Social Security for most of their monthly income. According to the
Social Security Administration, 61% of unmarred older Latinos and 44% of married older
Latino couples received at least 90% of their income from their earned Social Security
benefits in 2011 (U.S. Social Security Administration, 2013), underscoring the
importance of Social Security to this segment of the population.

Latino Baby Boomers
This paper looks at the characteristics of the Baby Boomer population living in
the 50 U.S. states and the District of Columbia. The population was broken down by
race/ethnicity for the presentation of citizenship data, and information on the
characteristics of Latino Baby Boomers was presented stratified by citizenship status.
Data from several U.S. Census Bureau sources were used, with each data source
addressing a different aspect of the Boomer generational cohort: 1) population sizes were
extracted from the full 2000 decennial census sample, the most recent decennial census
20

available as of the commencement of this study; 2) detailed demographic and economic
characteristics were derived from a 5% sample of the 2000 census respondents; and 3)
comparisons of some of these characteristics across time were drawn from historical 1980
and 1990 decennial censuses and the 2008 American Community Survey (ACS), the
most recent ACS available as of the commencement of this study.
1

Methods
To fully assess the demographic and economic conditions of the Latino baby
boomer population, several U.S. Census Bureau data sources were used. Historical
decennial data and 2008 ACS data were taken from the University of Minnesota’s
Integrated Public Use Microdata Series (IPUMS; Ruggles, Alexander, Genadek, Goeken,
Schroeder, & Sobek, 2010); data from only the 2008 ACS were used to minimize
fluctuations due to economic turmoil during the 2006-2008 period. IPUMS sample sizes
for 2000 vary slightly from those reported for the Census Bureau’s PUMS data due to
differences in weighting, but the component data are otherwise identical.
Macro-level statistics from the 2000 decennial census were garnered from the
U.S. Census Bureau’s Summary File 1 (SF1), which provides summary tables based on
100% data. This source was used to establish accurate population counts for baby
boomers, overall and by ethnicity. More detailed demographic information was pulled
from the U.S. Census 5% Public Use Microdata Sample (PUMS). Use of these data
allowed for bivariate analyses that are unavailable using the SF1. Weights were applied
to this dataset’s 3,992,849 baby boomer records to more closely approximate the national

1
This chapter builds on analyses previously published in the Journal of Aging & Social Policy (Gassoumis,
Wilber, Baker, & Torres-Gil, 2010) and as a book chapter (Torres-Gil, Gassoumis, & Wilber, 2012).
21

population, so that the numbers presented are comparable to those that would be expected
from 100% data. The population percentages generated from the PUMS data were
applied to the population counts established from SF1 to determine the size of population
sub-groups.
The Baby Boom Generation is comprised of all individuals born between 1946
and 1964. Unfortunately, public Census datasets do not include a birth year variable,
only a variable that provides age as of April 1, 2000. Thus, the Baby Boom Generation
was operationalized as all those born between April 2, 1946 and April 1, 1965, who were
between ages 35 and 53 as of April 1, 2000. Since 1980, Latino status in the U.S. census
has been based on self-reported identification as Spanish, Hispanic, or Latino.
Only data for people living in the 50 U.S. states and the District of Columbia were
used in these analyses. To enhance clarity, those born in a U.S. territory (see below)
were dropped from the historical comparisons. In classifying data by country of origin,
countries with the largest numbers of Latino boomers are presented separately: Mexico,
Puerto Rico, Cuba, Dominican Republic, El Salvador, Colombia, and Guatemala. All
other Latinos were aggregated into an Other category, which includes Spaniards and
other Hispanics and Latinos. Characteristics by country of origin are presented for
citizenship status, education, individual and household income, and interest income as a
proxy for wealth.

Measures
Variables that assess demographic and economic characteristics were included in
the analysis (see Table 2.1). Gender, education (less than high school, high school
22

degree, some college, college degree, graduate/professional degree), and marital status
(married, widowed, divorced/separated, and never married) provide basic demographic
information. The English ability of the individual is reported, with the highest level
indicating that English is either the only language spoken in the home or that the
individual speaks English very well despite another language being spoken in the home.
The other three categories—speaks English well, speaks English not well, and speaks
English not at all—are only relevant to those who speak a language other than English at
home. Household linguistic isolation means that no one in the household who is over 13
years old either speaks English very well or only speaks English at home.
Several demographic variables reported have relevance to economic status.
Housing tenure details homeownership status, which is broken into homeowner with and
without a mortgage, renter, and occupying a residence without payment. Living alone is
also reported, meaning the household size was one person. Those listed as employed
under labor force participation include people in the armed forces and those who are
employed but not currently at work, compared to people who are unemployed and not in
the labor force (e.g., homemakers, retirees, institutionalized individuals).
Disability is reported as a dichotomous variable, based on responses to the 2000 Census’
six “disability” domains: sensory disability, physical disability, mental disability, self-
care disability, going-outside-home disability, and employment disability (Wang, 2005).
The presence of at least one disability was used as a marker for some type of broad-
spectrum disability.
Material hardship is a dichotomous variable, calculated from responses to
questions about five types of hardship: plumbing facilities, kitchen facilities, telephone
23

Table 2.1. Analysis variables and descriptions.
Variable Description/Values

Female Sex: male; female

Education Less than high school; high school degree; some college; college
degree; graduate/professional degree

English Ability English is either the only language spoken in the home or is spoken
well despite another language being spoken in the home; English is
spoken well; English is spoken not well; English is not spoken at all
(The last three values are only relevant to those who speak a language
other than English at home)

Linguistically Isolated No one in the household over the age of 13 either speaks English very
well or only speaks English at home

Marital Status Married; widowed; divorced/separated; never married

Lives Alone Living arrangement: lived alone; lived with others

Housing Tenure Homeowner with a mortgage; homeowner without a mortgage; renter;
occupying a residence without payment

Material Hardship Whether or not the person had any one of five types of hardship as
measured by problems or lack of: plumbing facilities, kitchen facilities,
telephone service, fuel used for heating, and availability of vehicles

Disability Indicated by a yes response to any of the 2000 Census’ six “disability”
domains: sensory disability, physical disability, mental disability, self-
care disability, going-outside-home disability, and employment disability

Labor Force
Participation
All people in the labor force, including those in the armed forces and
those who were employed but not currently at work; all people not in the
labor force (e.g., homemakers, retirees, institutionalized individuals)

Individual Income Combines all income categories for an individual: wage/salary, net self
employment, interest, Social Security, Supplemental Security Income
(SSI), public assistance, retirement income, and any other income

Household Income Aggregated incomes for all members of a household

Poverty Status Household income in relation to the federal poverty line: ≤ 100%; 100%
to ≤ 150%; 150% to ≤ 200%; > 200%

Household Income
Devoted to Housing
Ratio of selected housing expenses (e.g., mortgage/rent, utilities) to
total household income, for both owned and rented households: ≤ 30%;
30% to ≤ 50%; > 50%

Has Interest Income The individual had any interest income (a proxy for investment); no
interest income

24

service, fuel used for heating, and availability of vehicles. While no universal index or
measure of material hardship exists (Ouellette, Burstein, Long, & Beecroft, 2004), these
five questions assess various domains of material hardship often used in research (U.S.
Census Bureau, 2005).
Variables that more strictly represent economic status are also presented.
Individual income combines all income categories for an individual: wage/salary, net self
employment, interest, Social Security, Supplemental Security Income (SSI), public
assistance, retirement income, and any other income. Household income is the
aggregated incomes of all members of a household. Poverty status is based on where the
household’s income falls in relation to the federal poverty line: 100% or below, greater
than 100% up to 150%, greater than 150% up to 200%, and greater than 200%. Housing
costs as a percentage of income is the ratio of selected housing expenses (e.g.,
mortgage/rent, utilities) to total household income, for both owned and rented
households. This variable is grouped based on commonly used cutoff points: 30% or
below, greater than 30% up to 50%, and greater than 50%.
Although the long form questionnaire collected detailed information about
income, it did not ask about wealth. However, the income data collected include a
question on interest income. Previous studies have used this variable as a continuous
marker for level of wealth (Gabriel & Painter, 2008; Gentry & Hubbard, 2004). Interest
income reflects wealth, but it is not an accurate proxy for level of wealth, as different
savings vehicles offer highly varied interest rates. Therefore, a dummy variable was used
for the presence of interest income as a proxy for the existence of wealth.
25

The analysis includes stratification by citizenship status, primarily for the Latino
population. This variable delineates four categories of citizenship: 1) those who were
born in the U.S. (50 states plus D.C.) or abroad to a U.S. parent; 2) those who were born
in a U.S. territory (e.g., Puerto Rico, Guam, U.S. Virgin Islands); 3) those who are
naturalized U.S. citizens; and 4) those who are not citizens of the U.S. As the
demographic and economic differences between those born abroad to a U.S. parent (1.7%
of category) and those born in the U.S. (98.3% of category) are minimal, this paper treats
them all as U.S.-born citizens. Year of entry and age upon entry to the U.S. was also
compared for the three groups that were not U.S.-born.

Results
Population Size & Citizenship Status
There were roughly 80.0 million baby boomers—those born between 1946 and
1964—in the United States in 2000; 8.0 million (10.0%) of these were Latinos.
2
Of the
80 million boomers, 15.0% (12.0 million) were born outside of the United States.
Among the 12 million foreign born baby boomers, Latinos accounted for over 5 million
(43%). In 2000, over one-third of Latino boomers were U.S.-born citizen (37%); a
similar proportion were non-citizen (36%); about one in five were naturalized citizens
(21%) and citizens born in U.S. territories represented 6% (Table 2.2). Among citizens
born in U.S. territories, 95% identified Puerto Rico as their country of origin. The mean

2
This figure deviates from the 78 million estimate often attributed to the 2000 Census data. As public
Census datasets do not include a birth year variable—only a variable that provides age as of April 1,
2000—we treat all those born between April 1946 and March 1965, who were between ages 35 and 53 as
of April 1, 2000, as the Baby Boom Generation. Using 36 to 54 as the age constraints for the boomer
population results in an estimation of 78.3 million people who were born between April 1945 and March
1964, a less accurate representation of the Baby Boom Generation.
26

Table 2.2. Citizenship status of baby boomers, by ethnicity and race, 1980-2008.
1980 1990 2000¹ 2008
n = 3,819,553 n = 3,936,922 n = 4,039,402 n = 772,837
Latino Citizens, U.S.-born
60.5 44.1 37.0 36.7

Citizens, born in
U.S. territory
6.9 6.7 6.1 5.5

Naturalized
citizens
7.9 12.8 20.9 24.7
Non-citizens 24.7 36.4 36.1 33.1
N 275,397 356,802 411,216 80,651
Non-Latino
White Citizens, U.S.-born 97.6 96.8 96.0 95.5

Citizens, born in
U.S. territory
0.0 0.0 0.0 0.0

Naturalized
citizens
1.1 1.4 2.0 2.8
Non-citizens 1.3 1.7 1.9 1.6
N 2,974,660 2,964,006 2,915,062 557,790
Black Citizens, U.S.-born
96.5 93.4 91.2 89.3

Citizens, born in
U.S. territory
0.1 0.1 0.2 0.1

Naturalized
citizens
1.2 2.0 4.6 6.5
Non-citizens 2.2 4.5 4.1 4.0
N 472,534 457,869 462,382 86,828
Asian Citizens, U.S.-born
33.7 19.2 13.2 12.8

Citizens, born in
U.S. territory
0.7 0.6 0.6 0.7

Naturalized
citizens
16.2 32.5 49.3 59.4
Non-citizens 49.4 47.8 36.9 27.2
n 65,773 125,751 158,629 33,644
Other Citizens, U.S.-born
92.3 94.0 76.6 88.0

Citizens, born in
U.S. territory
0.1 0.2 0.2 0.4

Naturalized
citizens
2.1 1.8 11.9 7.6
Non-citizens 5.6 3.9 11.3 4.1
n 31,189 32,494 92,113 13,924

Note: All units are percentages, by row. "Other" includes American Indian, Alaska Native, and
Native Hawaiian; from 2000, "Other" also includes people who identify with more than one race.
¹Population sizes are: Latino = 8,034,496; Non-Lati no = 71,974,423; White = 58,040,447; Black
= 9,166,431; Asian = 3,003,516; Other = 1,764,029.

27

age of entry into the United States was 20 years for both citizens born in U.S. territories
and naturalized citizens, and 28 years for non-citizen Latino boomers.
The citizenship status in 2000 of non-Latino white and black baby boomers
reflected populations made up of few immigrants: in both groups, fewer than one in ten
individuals was not a U.S.-born citizen. The largely U.S.-born nature of the non-Latino
white and black populations is reflected at all four time points (1980, 1990, 2000, and
2008), though it appears that there is slightly more immigration among non-Latino blacks
than there is among non-Latino whites, relative to their total population sizes. Non-
Latino Asian boomers, on the other hand, were composed of a greater proportion of
immigrants than were Latinos at all four time points. And despite a steadily growing
population, much of the increase was in naturalized citizens as opposed to non-citizens:
across the four time points, the proportion of Asian boomers who were non-citizens
decreased considerably while the proportion who were naturalized citizens increased
sharply. In contrast, the proportion of Latino boomers who were non-citizens grew from
1980 to 1990, after which it leveled out, while the proportion who were naturalized
citizens grew at a relatively modest pace. Based on population sizes for a cohort that
roughly approximates the baby boomers, the real number of naturalized Latino citizens
increased by approximately 1.2 million from 1980 to 2000 whereas non-citizen Latinos
increased by approximately 1.5 million across the same time period. This compares to an
increase of approximately 1.3 million naturalized Asian citizens and 400,000 non-citizen
Asians from 1980 to 2000.

28

Characteristics in 2000
Table 2.3, which compares Latino baby boomers, as a whole and by citizenship
status, to non-Latino baby boomers, indicates that Latinos were at a greater disadvantage
than non-Latinos across a wide range of demographic and economic variables. Latinos
had lower levels of education, especially those Latinos who were not born in the United
States. For example, while over 60% of non-Latino boomers had at least some college
education, less than half of each of the four Latino groups—U.S.-born citizens (47%),
citizens born in a U.S. territory (34%), naturalized citizens (36%), and non-citizens
(18%)—had some college. Latino boomers were less likely to be proficient in English
than their non-Latino counterparts. This was especially true for those who were not born
in the United States: only 56% of citizens born in a U.S. territory, 44% of naturalized
citizens, and 21% of non-citizen Latinos either spoke only English in the home or spoke
English very well, compared to 88% of U.S.-born Latinos and 96% of non-Latinos.
Consequently, nearly two-fifths of non-citizen Latino boomers and approximately one-
fifth of both naturalized Latino citizens and those born in U.S. territories were
linguistically isolated, with no one in the household speaking English very well.
Latino boomers were less likely than their non-Latino counterparts to live alone
(11% for non-Latinos vs. 5% for all Latinos); the rates for those born in the United States
or its territories, however, were much closer to that of non-Latinos, at 8% and 9%
respectively. While the homeownership rate for U.S.-born (67%) and naturalized (65%)
Latino citizens was close to that of non-Latinos (77%), non-citizen Latinos (42%) and
those born in U.S. territories (43%) were considerably less likely to own their homes. Of
those U.S.-born Latinos who owned their homes, more were free of a mortgage (19%)

29
Table 2.3. Baby boomer demographic and economic characteristics, by ethnicity and citizenship status, 2000.

All
Non-Latino
All Latino
Latino, by Citizenship Status
Citizens,
U.S.-born
Citizens, born
in U.S. Territory
Naturalized
Citizens
Non-
citizens

n = 3,586,372 n = 406,477 n = 150,289 n = 24,649 n = 84,972 n = 146,567
Female 50.8 49.4 51.2 52.4 50.7 46.3
Education
< High School 11.1 44.6 24.1 39.7 43.4 67.1
High School degree 28.4 22.1 29.4 25.9 20.5 15.0
Some college 31.6 21.6 31.9 21.8 22.8 10.5
College degree 18.2 7.0 9.5 8.0 8.0 3.8
Graduate/Professional degree 10.7 4.6 5.2 4.7 5.4 3.5
English Ability
Primary language/Speaks very well 96.1 52.7 87.9 55.7 44.2 21.1
Speaks well 2.4 19.0 9.2 24.4 32.0 20.7
Does not speak well 1.4 18.4 2.4 15.0 19.6 34.7
Does not speak English 0.2 9.8 0.4 4.8 4.2 23.5
Linguistically Isolated 1.7 20.8 4.4 17.4 20.0 38.8
Marital Status
Married 68.2 67.6 62.4 56.7 73.8 71.3
Widowed 1.5 1.8 1.7 2.5 1.7 1.7
Divorced/Separated 17.7 17.2 21.0 23.0 15.5 13.5
Never married 12.7 13.4 14.9 17.9 9.0 13.5
Lives Alone 10.9 5.4 8.2 9.3 3.9 2.8
Housing Tenure
Owned with mortgage 64.1 46.3 53.9 38.8 55.9 34.4
Owned free & clear 12.8 9.6 12.8 4.6 9.4 7.5
Rented 21.6 42.6 31.8 55.1 33.4 56.6
Occupied without payment 1.4 1.4 1.5 1.5 1.2 1.4

30
Table 2.3 (cont'd). Baby boomer demographic and economic characteristics, by ethnicity and citizenship status, 2000.

All
Non-Latino
All Latino
Latino, by Citizenship Status
Citizens,
U.S.-born
Citizens, born
in U.S. Territory
Naturalized
Citizens
Non-citizens

n = 3,586,372 n = 406,477 n = 150,289 n = 24,649 n = 84,972 n = 146,567
Material Hardship 7.6 17.6 11.8 30.6 13.9 23.5
Disability 18.6 26.8 24.2 35.1 26.5 28.2
Labor Force Participation
Employed 78.2 64.2 69.7 56.3 66.9 58.2
Unemployed 2.9 5.0 4.3 4.9 4.2 6.2
Not in labor force 18.9 30.9 26.0 38.8 28.9 35.7
Individual Income (mean) $38,004 $23,558 $28,134 $22,971 $27,123 $16,897
Individual Income (median) $28,000 $17,000 $22,000 $16,000 $20,000 $12,300
Household Income (mean) $75,124 $55,271 $60,470 $49,411 $60,741 $47,883
Household Income (median) $60,000 $44,000 $50,000 $39,000 $48,000 $37,000
Poverty Status
100% of poverty line or less 7.6 17.3 12.8 22.6 12.6 23.8
>100% to 150% 5.1 12.9 8.7 10.9 11.8 18.1
>150% to 200% 5.9 12.1 8.9 9.6 12.2 15.7
>200% 81.3 57.6 69.6 56.8 63.4 42.5
Household Income Devoted to Housing
30% or less 80.2 70.8 76.7 67.3 69.2 66.3
>30% to 50% 12.6 17.4 14.2 17.0 19.0 19.7
>50% 7.2 11.8 9.1 15.7 11.8 14.0
Has Interest Income 22.9 8.1 10.8 7.5 10.1 4.2
Year of Immigration (mean) - - - 1975 1976 1985
Age at Immigration (mean) - - - 20.0 20.3 28.0

Note: Unless otherwise indicated, all units are percentages, by column.
31

than the other Latino groups and even non-Latinos. Latino boomers, especially non-
citizens (23%) and those born in U.S. territories (31%), were more likely to exhibit at
least one marker of material hardship than non-Latinos (8%), often due to a lack of a
vehicle or household utilities.
Disability rates were higher across the board for Latino boomers (27%) than for
non-Latino boomers (19%). These high rates of disability were reflected in the relatively
low employment rates of Latino boomers: 78% of non-Latino boomers were employed,
compared with rates between 56% and 70% for Latinos, depending on citizenship status.
Similarly, individual earnings for Latino baby boomers (mean = $23,558) were
considerably lower than for non-Latino boomers (mean = $38,004). Household income
was also considerably lower for Latino households (mean = $55,271) than for non-Latino
households (mean = $75,124). In terms of citizenship status, however, U.S.-born and
naturalized Latino citizens’ incomes were much closer to the incomes of non-Latino
boomers, both individual and household. Also, household income was approximately
three times larger than individual income for non-citizen Latinos, a much larger spread
than for the other Latino groups and the non-Latino boomers. This likely reflects the
lower likelihood of living alone in this population and indicates a greater frequency of
multiple incomes within a household.
In terms of the federal poverty level, Latino baby boomers fared much worse than
non-Latinos. This was especially true for non-citizen Latinos and Latinos born in U.S.
territories, of whom nearly one-quarter fell at or below 100% of the poverty level. This
pronounced disparity was reduced somewhat, however, when income sufficiency was
defined by actual cost-of-living rather than the raw, standardized poverty level. When
32

looking at the percentage of household income that was devoted to housing costs, only
14% of non-citizen Latino boomers and 16% of Latino boomers born in U.S. territories
fell into the least favorable category, in which over half of household income was spent
toward housing. This compared with 7% of non-Latinos, roughly equivalent to the
proportion of non-Latinos at or below 100% of the poverty level (8%).
Finally, the receipt of interest income was less prevalent among Latinos,
indicating that fewer of them had non-real estate assets than did non-Latino boomers.
Once again, U.S.-born (11%) and naturalized Latino citizens (10%) were comparable on
this measure. Their rates were less than half of that seen for non-Latinos (23%) but
markedly above the rates for non-citizen Latinos (4%) and Latino boomers born in U.S.
territories (7%).

Comparing Historical Characteristics
Three demographic and economic characteristics are especially useful for comparing
Latino citizenship groups across time: education, English ability, and income. Table 2.4
presents these and other characteristics for Latino boomers who were either U.S.-born
citizens, naturalized citizens, or non-citizens in 1980, 1990, 2000, and 2008. The
characteristics for the samples of U.S.-born citizens are reflective of a relatively stable
population, with small levels of attrition due to emigration and death. In contrast, the
naturalized citizen and non-citizen groups have had a very fluid composition over time
due to movement from non-citizen to naturalized citizen status as well as increases due to
immigration among both groups.

33
Table 2.4. Select characteristics of Latino baby boomers, by citizenship status, 1980-2008.
1980 1990
Citizens,
U.S.-born
Naturalized
Citizens
Non-citizens
Citizens,
U.S.-born
Naturalized
Citizens
Non-citizens

n = 166,672 n = 21,657 n = 68,117 n = 146,801 n = 42,087 n = 119,235
Female 50.1 49.3 47.0 50.4 48.3 45.2
Education
< High school 47.9 46.0 66.3 25.8 43.6 63.2
High school degree 25.9 20.8 15.4 30.9 19.6 16.3
Some college 21.2 24.2 14.3 31.0 23.7 13.9
College degree 3.0 4.9 1.8 8.9 8.7 4.1
Graduate/professional degree 2.0 4.1 2.2 3.4 4.5 2.5
English Ability
Primary language/speaks very well 81.6 56.3 28.9 86.3 53.1 26.4
Speaks well 14.7 23.5 24.7 10.4 24.2 24.3
Does not speak well 3.1 13.5 27.2 2.8 17.1 31.8
Does not speak English 0.6 6.7 19.3 0.5 5.6 17.5
Marital Status
Married 37.8 53.0 50.2 58.8 70.2 66.1
Widowed 0.3 0.3 0.3 0.7 0.8 0.8
Divorced/separated 7.2 7.3 4.8 16.4 12.0 9.8
Never married 54.7 39.4 44.7 24.2 17.0 23.3
Individual Income (mean)¹ $12,200 $15,044 $11,322 $22,552 $22,617 $14,912
Individual Income (median)¹ $6,896 $11,486 $7,757 $18,816 $17,472 $12,096
Year of Immigration (mean) - 1969 1974 - 1974 1980

Note: Unless otherwise indicated, all units are percentages, by column. ¹Income across all years is ex pressed in 1999 dollars.

34

Table 2.4 (cont'd). Select characteristics of Latino baby boomers, by citizenship status, 1980-2008.
2000 2008
Citizens,
U.S.-born
Naturalized
Citizens
Non-citizens
Citizens,
U.S.-born
Naturalized
Citizens
Non-citizens

n = 143,325 n = 80,497 n = 138,919 n = 28,495 n = 18,372 n = 20,968
Female 51.2 50.7 46.3 51.2 51.1 46.8
Education
< High school 24.1 43.4 67.1 19.7 35.5 62.5
High school degree 29.4 20.5 15.0 30.4 24.5 18.4
Some college 31.9 22.8 10.5 33.1 23.4 10.3
College degree 9.5 8.0 3.8 10.7 10.5 6.0
Graduate/professional degree 5.2 5.4 3.5 6.2 6.1 2.9
English Ability

Primary language/speaks very well 87.9 44.2 21.1 88.9 42.9 14.1
Speaks well 9.2 32.0 20.7 7.6 30.8 18.9
Does not speak well 2.4 19.6 34.7 2.8 22.0 39.0
Does not speak English 0.4 4.2 23.5 0.7 4.4 28.0
Marital Status
Married 62.4 73.8 71.3 58.1 69.9 66.8
Widowed 1.7 1.7 1.7 3.2 3.0 3.4
Divorced/separated 21.0 15.5 13.5 24.8 18.7 16.0
Never married 14.9 9.0 13.5 14.0 8.4 13.8
Individual Income (mean)¹ $28,134 $27,123 $16,897 $29,017 $26,303 $15,739
Individual Income (median)¹ $22,000 $20,000 $12,300 $20,898 $19,350 $11,610
Year of Immigration (mean) - 1976 1985 - 1978 1988

Note: Unless otherwise indicated, all units are percentages, by column. ¹Income across all years is ex pressed in 1999 dollars.
35

Educational attainment among U.S.-born Latino citizens steadily improved from
1980, when the youngest boomers were still of high school age, to 2008, when 50% had
at least some college education and 17% had at least a college degree. The educational
attainment of naturalized citizens increases as well, from 33% having at least some
college education in 1980 to 40% in 2008. This increase is likely due to both increasing
education of existing naturalized citizens and higher levels of education among newly
naturalized citizens. The educational attainment of non-citizen Latinos remained fairly
static across the time span, although there was a steady increase in college graduates.
In line with education, U.S.-born Latino citizens attained consistently higher
levels of English proficiency from 1980 to 2008. The English proficiency of naturalized
and non-citizen Latinos, however, became steadily worse. In 1980, 56% of naturalized
Latino citizens rated themselves as fluent English speakers, whereas this proportion had
dropped to 43% by 2008. Similarly, non-citizen Latinos dropped from 29% fluency in
1980 to 14% in 2008. Both of these reductions in group proficiency are likely due to a
lower level of English ability among new immigrant groups who migrated late in life
versus those who immigrated at a relatively young age prior to 1980.
As with the patterns seen in the above in-depth analysis of year 2000 data,
individual income levels for naturalized Latino citizens are similar to those for U.S.-born
Latino citizens across the 1980-2008 time span. The income levels in Table 2.4,
presented in 1999 dollars for all four years, increase steadily for both citizen groups
through 2000, after which there is a slight drop in 2008 that is likely attributable to the
economic recession. Non-citizen Latinos, on the other hand, experience increased
income from 1980 to 1990 but no significant increase in real dollars past 1990.
36

Discussion
A discussion of Latino baby boomers must begin by correcting the record: the
largest generation in the nation’s history is even larger than commonly recognized. The
original baby boomers included roughly 76 million individuals born in the U.S. between
1946 and 1964 (U.S. Public Health Service, 1948-1966). Even though it has been over
forty years since the birth of the last boomer, the figure of 76 million is still frequently
cited; however, due to immigration, emigration, and death, the number of boomers in the
U.S. in 2000 was roughly 80 million. Of these, 12 million were born outside the U.S.,
consisting of 64% of Latino boomers and 9% of non-Latino boomers; this means that
approximately 8 million of the original 76 million-person Baby Boom Generation had
either emigrated or died. In addition to their sheer numbers, immigrants have changed
the composition of the Baby Boom Generation. For example, nearly half of the 12
million boomers born outside the U.S. were non-citizens in 2000, comprised of roughly
equal proportions of Latinos & non-Latinos. In the years since the 2000 Census, these
numbers have shifted slightly, with the number of baby boomers dropping to 78 million
in the 2010 Census; but the demographic profile and the relatively large number of
Latinos has persisted.
This paper compared four categories of Latino baby boomers, based on their
citizenship status: citizens born in the U.S. (including those born abroad to a U.S. parent),
citizens born in U.S. territories, naturalized citizens, and non-citizens. With little more
than a third of Latino boomers having been born in the U.S., much more of the
population is foreign-born (approximately 63%) than the current population of Latino
elders (47%; Pew Hispanic Center, 2008). The large number of non-citizens coupled
37

with striking—though not surprising—differences in demographic and economic
characteristics by citizenship status draws attention to questions of how well these diverse
groups will fare in retirement.
In general, U.S.-born Latino boomers were more similar to non-Latino boomers in
terms of demographic characteristics, whereas foreign-born citizens and non-citizens
scored less well on key demographic indicators. For example, the proportion of Latino
boomers born in the U.S. who spoke English very well was comparable to that of non-
Latino boomers. Few Latinos born in the U.S. were linguistically isolated, and while
their rates of education were lower than those of non-Latinos, more than three-quarters of
U.S.-born Latino boomers had a high school degree or higher and almost one-half had at
least some college education. Latino baby boomers who were not born in the U.S.
painted a bleaker educational and linguistic picture, especially among non-citizens and
those born in U.S. territories.
Economic characteristics of U.S.-born Latino baby boomers were somewhat less
favorable than those of non-Latino boomers. For example, few Latino boomers had high
levels of income and savings compared to non-Latino baby boomers. Nevertheless,
among those born in the U.S., two-thirds owned their own home, and their proportion of
income devoted to housing was similar to that of non-Latinos. With low levels of income
and high rates of poverty, Latino baby boomers who were not citizens or were born in a
U.S. territory fared poorly on most economic measures. Comparing Latino boomers in
general to current Latino elders suggests that boomers are better off in terms of income,
with 43.4% of older Latinos at 150% of poverty (Beedon & Wu, 2004) compared to
38

30.2% of Latino boomers; however, as two-thirds of Latinos boomers were employed,
the extent to which these indicators will change when they retire is not known.

Policy Implications
As the increasingly diverse generation of baby boomers enters retirement in an era
of economic uncertainty, policymakers will face a number of critical questions raised by
massive demographic shifts. Although specific policy solutions for Latino baby boomers
would be impractical, it is important for policymakers to understand the characteristics of
Latino boomers—10% of the large Baby Boom Generation—when addressing policy
reform. Well researched, data driven information will help national- and state-level
policymakers better anticipate the needs and available resources of this large segment of
the baby boom population.
In the face of rising federal deficits and long-term actuarial shortfalls, the
sustainability of social insurance programs has been a very present topic in U.S. political
discourse since the 1980s (Daguerre, 2011). This attention has continued into the current
administration, with President Obama pledging to focus on social insurance early in his
first term (“Social Security/Medicare,” 2009). Any discussion of entitlement reforms,
however, should consider the impending retirement of the baby boomers’ most
vulnerable members, a group which includes many Latinos. With high poverty levels
and a historical dependence on Social Security, Latinos in general are particularly
susceptible to entitlement reforms that aim to reduce benefits (Halliwell, Gassoumis, &
Wilber, 2007). Nonetheless, it is important to recognize that, as a result of the
immigration-driven transformations that have occurred in the boomer generation over the
39

last five decades, a sizable proportion of Latino baby boomers are not likely to be eligible
for these social insurance programs. In the face of the current economic situation, the
fiscal instability of aging entitlements should be addressed in a fashion that upholds their
core tenets.
Representing a hidden subset within an understudied demographic, non-citizen
Latino baby boomers are much more likely to be poor, linguistically isolated, and have
less than a high school education. These non-citizen Latino boomers, who represent over
one-third of all Latino boomers, are the cornerstone upon which two controversial policy
areas converge: policies that address entitlement reform and policies related to
immigration reform. Barriers to achieving citizenship among this group are not identified
in the data presented here, yet non-citizen Latino baby boomers had been in the U.S. an
average of 15 years in 2000. The citizenship status of this sizable proportion of Latino
baby boomers underscores gaps in policies relating to entitlement programs as well as
immigration and naturalization, gaps that should be addressed to improve the prospects of
the upcoming boomer retirees.
Understanding the intricacies of all Latino baby boomers will shed light on the
complex factors related to the nexus of aging and diversity in the U.S. Policies
concerning the Latino boomer population will serve as a barometer of the willingness of
U.S. politicians to address the emerging Latino population. Given that Latinos are a
relatively young group, the issues confronting Latino boomers will signal how
policymakers might prepare for the future aging of upcoming Latino generations.

40

Conclusion and Future Directions
This study offered a first look at key indicators that provide early insights into
Latino baby boomers’ economic status as they approach retirement age. With the U.S. on
the cusp of a burgeoning older adult population, it is important to build on this study to
further examine the financial well-being of Latino boomers. The analyses presented here
elucidate the importance of citizenship status when considering the demographic and
economic characteristics of Latino baby boomers. Future studies will be important in
determining to a greater degree the diversity in resources and benefit eligibility among
Latino baby boomers as they begin to enter retirement.
Additional steps to inform the policy debate should include more specific
information about savings rates and how much Latino baby boomers have in assets, as
well as how these assets are allocated (e.g., real estate, defined benefit pensions, defined
contribution plans such as IRAs or 401(k) accounts). New sources of data will shed light
on how job losses and other ramifications of the current economic downturn are affecting
Latino boomers. Given relatively high rates of home ownership, it will be important to
know how much mortgage debt has been accrued and how the recession and related
increases in foreclosures have affected this population. Such information will help
foreshadow the impact of baby boomers’ retirement on public benefits such as Social
Security, Medicare, and Medicaid.
41

CHAPTER 3: INCOME AND WEALTH DISPARITIES: STRUCTURAL
INEQUALITIES FOR BABY BOOMERS AND ACROSS COHORTS

Background
Retirement income in the United States has historically been based on the concept
of a 3-legged stool: comprised of equal parts pensions, savings, and Social Security. The
current employment landscape, however, has seen defined benefit pensions largely
disappear (Butrica, Iams, Smith, & Toder, 2009); their defined contribution counterparts
such as 401(k) plans, which can be considered part of the savings leg of the stool, are far
from pervasive, and savings rates nationwide are at low levels (Chen, 2013). As a result,
the traditional three-legged stool is shifting increasingly to a one-legged stool, held up
largely by Social Security. Social Security benefits, however, were never intended to
account for the entirety, or even the majority, of an individual’s retirement income
(Altman, 2005).
Overreliance on Social Security and related economic insufficiency in retirement
is more common among racial/ethnic minority retirees (Gassoumis, Lincoln, & Vega,
2011; Smith, 1997). In 2011, Latinos and black/African Americans aged 65 and older
had median income levels that were 1/3 less than non-Latino whites ($22,126 and
$22,738 vs. $34,113), whereas Asians/Pacific Islanders had median incomes slightly
($37,000) higher than non-Latino whites (Wu, 2013). These disparities are associated
with stark discrepancies in poverty status, with far more racial/ethnic minorities falling
below or near the federal poverty line when compared with non-Latino whites
(Gassoumis et al., 2011; Wu, 2013). Reliance on federal income sources (Social Security
42

and SSI [Supplemental Security Income], combined) for those below 200% of the
poverty line is similar across racial/ethnic lines, making up over 2/3 of total income;
among those above 200% of the poverty line, however, Latinos and black/African
Americans relied more heavily on these sources than non-Latino whites (Gassoumis et
al., 2011).
The reduction in pensions and inadequacy of Social Security as the sole source of
income in retirement means that there will necessarily be an increased emphasis on
savings for future retirees. But income sufficiency in the worklife is a necessary
condition for building savings, as a household with income that only covers the most
basic living expenses will lack the discretionary income necessary to build savings. As
such, understanding the factors that contribute to racial/ethnic income disparities during
the worklife is an important path toward enhanced economic security among all retirees.
Among several possible sources of income disparities between racial/ethnic
groups, the most widely recognized is education. Educational attainment, a metric that
has strong links with both income and wealth, varies considerably by race/ethnicity.
Asians have much higher rates of college graduation than the non-Latino white
population, whereas black/African Americans and Latinos fare worse than non-Latino
whites (U.S. Census Bureau, 2012a). Immigrant and citizenship status is the second most
widely recognized driver of income disparity that is particularly relevant to racial/ethnic
minority groups (e.g., Chiswick, 1978; Gassoumis et al., 2010). Although these
sociodemographic factors are strongly associated with economic disparities within the
U.S., racial/ethnic differences are also linked to inherent structural inequalities between
minority and majority groups (Sandefur & Pahari, 1989). The degree to which economic
43

disparities are due to structural forces in contrast to individual-level sociodemographic
characteristics, however, has received little attention from scholars studying the
economics of aging.
To better understand economic security in retirement, this study examines income
and wealth among Latinos in the years immediately preceding retirement. At over 52
million people, it is the largest racial/ethnic minority group in the U.S., making up 17%
of the total population (U.S. Census Bureau, 2013). It is also increasing with remarkable
speed, projected to grow to 28% of the total U.S. population by 2050 (U.S. Census
Bureau, 2012b). Additionally, Latinos offer a diverse demographic group to study, with
a broad range of educational attainment—typically well below the U.S. population
average—and large variations based on immigration and citizenship status (Gassoumis et
al., 2010).
The purpose of this study is twofold: to assess the degree to which racial/ethnic
structural economic disparities are due to the societal factor of ingrained race-ethnic
disadvantages versus individual-level factors, and to examine whether the interplay
between those two factors has changed across generational cohorts. These disparities are
tested immediately before retirement age—when they are expected to be largest due to
the effects of cumulative inequality (Ferraro & Shippee, 2009)—among two generational
cohorts: the Baby Boom Generation (born 1946-1964) and the Silent Generation (born
1925-1945). These analyses and interpretations focus primarily on Latinos, the largest
and fastest growing racial/ethnic group in the U.S. and one that is subject to particularly
high heterogeneity in sociodemographic factors due to high levels of immigration and
relatively low naturalization rates.
44

The Structural Inequality of Race/Ethnicity
Individual economic success, as with any societal factor, is attained within the
context of a broader society. As such, it is subject to the social constraints that exist
within that society (Alba & Nee, 2003). These social constraints on economic status can
emerge from various factors, including split labor market dynamics and their resultant
race/ethnicity-based occupational pigeonholing and discrimination (Bonacich, 1972;
Restifo et al., 2013), citizenship-based employment constraints for immigrant groups
(e.g., Bratsberg et al., 2002), and broad-scale discriminatory workplace dynamics (Green,
2003).
Social scientists have long observed these structural inequalities in economic
variables after adjusting for sociodemographic characteristics, both in income (e.g.,
Waters & Eschbach, 1995) and housing/housing wealth (e.g., Krivo & Kaufman, 2004).
Based on the work of classic social theorists, structural inequalities would be expected to
lessen over time, as society’s valuation of performance supersedes the importance of
racial/ethnic differences; indeed, Sandefur and Pahari (1989) found evidence that the
second half of the 20
th
century saw a closing of the gap in the earnings of racial/ethnic
minority groups. Although disparities due to structural inequalities still exist, they appear
to have been decreasing over the recent past.

Cumulative Inequality
Under the theory of cumulative inequality (Ferraro & Shippee, 2009; Ferraro et
al., 2009), sociodemographic, economic, and other factors experienced early in life are
expected to have ongoing and cumulative effects throughout the lifecourse. Cumulative
45

inequality and cumulative advantage/disadvantage (Dannefer, 1987)—the original theory
on which cumulative inequality was based—is most commonly applied to issues of
biomedical and health disparities (e.g., Shuey & Willson, 2008; Wickrama, Mancini,
Kwag, & Kwon, 2013). However, it has great relevance to income, wealth, and other
economic factors that are similarly shaped by sociodemographic characteristics and
exhibit cumulative properties over the lifecourse, and is frequently applied to studies of
these factors (e.g., Crystal & Shea, 1990; Gregoire et al., 2002).
Core to the cumulative inequality framework is that inequalities between
demographic groups are manifested and compounded over the lifecourse, with their
trajectories shaped by the accumulation of risk and resources. However, an individual’s
trajectory can crucially be modified by human agency (Ferraro & Shippee, 2009),
through means such as education and behavioral changes. The provision for human
agency within the cumulative inequality framework is crucial in studies involving
immigrant populations. The act of immigrating is, in and of itself, an example of that
agency, and the attainment of citizenship may induce a further alteration of one’s
trajectory.

Mechanisms of Human Agency: Education and Naturalization
Two primary mechanisms of human agency will be addressed in this study:
education and, for immigrants, naturalization. Education is largely accumulated early in
the lifecourse, but the degree of educational attainment is determined in part by human
agency. Although additional formal education can be sought later in life, this can be
more difficult for racial/ethnic minority immigrant populations, many of whom enter the
46

U.S. without any secondary education. When attained by immigrants, however, it comes
with higher earnings (Bratsberg & Ragan, 2002).
Almost all immigrants enter the U.S. as non-citizens; those who wish to attain
citizenship must do so through the process of naturalization. Naturalization comes with
an economic cost and is not open to everyone. There is currently no broad-scale
mechanism in place for individuals who entered the U.S. undocumented to obtain
documentation, let alone citizenship. But to those eligible who have the necessary
means, naturalization has the potential to make a downward trajectory of cumulative
inequality curve upward by removing barriers to occupational mobility. The theory of
segmented assimilation (Portes & Zhou, 1993) lays out three trajectories of assimilation:
toward the majority norm, in a position of upward mobility with limited assimilation to
the majority norms, or into a niche within the majority society that reinforces poverty.
Under segmented assimilation, naturalization can represent either a step down the path of
assimilation toward the majority norm or a move away from a poverty-reinforcing
position within society and toward an assimilational position that allows for economic
prosperity.

Hypotheses
This study first identifies the magnitude of racial/ethnic structural disadvantage in
the years preceding retirement for the Baby Boom Generation, the most recent
generational cohort to reach retirement age. Next, it compares their structural
disadvantage during this age range with that of members of the Silent Generation cohort
when they were the same age. These analyses are performed both for income and wealth.
47

Based on the assumption that structural inequalities will diminish over time, and taking
into account the findings of Sandefur and Pahari (1989), the analyses are undertaken to
test four specific hypotheses.

Hypothesis 1: Most of the disparity in income for racial/ethnic minority members of the
Silent Generation and Baby Boom Generation cohorts can be accounted for by
sociodemographic characteristics as opposed to the effects of structural disadvantage.
Hypothesis 2: Once adjusting for sociodemographic characteristics, the remaining
structural disadvantage in income for racial/ethnic minority individuals is less for
members of the Baby Boom Generation cohort than for members of the Silent Generation
cohort.
Hypothesis 3: Most of the disparity in wealth for racial/ethnic minority members of the
Silent Generation and Baby Boom Generation cohorts can be accounted for by
sociodemographic characteristics as opposed to the effects of structural disadvantage.
Hypothesis 4: Once adjusting for sociodemographic characteristics, the remaining
structural disadvantage in wealth for racial/ethnic minority individuals is less for
members of the Baby Boom Generation cohort than for members of the Silent Generation
cohort.

Methods
This study assessed the levels of income and wealth of two cohorts in the years
before they reached retirement age: a subset of the Baby Boom Generation (born 1946-
1953) and a subset of the Silent Generation (born 1938-1945). To capture these two
48

groups at similar points in their lives, 2008 data were used for the baby boomers, when
they were aged 54-62, and 2000 data were used for the Silent Generation cohort, when
they were at the same ages.
Data sources provided by the U.S. Census Bureau were used for income analyses,
to ensure national generalizability. Baby boomer income data from 2008 were taken
from the American Community Survey (ACS), the Census Bureau’s annual snapshot that
samples roughly 1% of the of the U.S. population. Although the ACS does not provide
birth year, it provides quarter of birth and age at the time of the survey. Since data for the
ACS are collected throughout the year, people born from January through June were
more likely to have had their birthday at the time of their data collection, whereas people
born from July through December were more likely not yet to have had their birthday.
To achieve a sample that best resembled those born 1946-1953, the analysis sample
included those aged 54 who were born in July through December, those aged 55-61, and
those aged 62 who were born in January through June.
Silent Generation income data were taken from the 2000 Decennial Census’ 5%
public-use microdata sample, which is a compilation of the roughly 6% of the U.S.
population that completed the long-form version of the 2000 Census. The decennial
census in the U.S. collects a snapshot as of April 1
st
of the calendar year in which the data
are collected. Year of birth information is not released, but age is provided. So that the
sample best captured those born 1938-1945, it included everyone aged 54-61 as of April
1, 2000.
Wealth data for both cohorts were drawn from the Health and Retirement Study
(HRS), a large-scale, longitudinal, nationally representative sample of the U.S.
49

population over the age of 50. Wealth and demographic data were taken from the 2008
HRS wave for the baby boom cohort and from the 2000 HRS wave for the Silent
Generation cohort.
ACS and Census data were extracted from the Integrated Public Use Microdata
Series (IPUMS) out of the University of Minnesota (Ruggles et al., 2010), and HRS data
were extracted from the RAND HRS Data File (RAND HRS Data, 2011) and the RAND
Enhanced Fat Files. For all data sources, individuals living in group quarters were
excluded and nationally representative sample weights were applied. There were
incomplete (missing) data for 104 cases (2.6% of total) from the 2000 HRS sample and
23 cases (0.7% of total) from the 2008 HRS sample; since less than 5% of the cases in
each sample had missing data, case-wise deletion was employed to deal with the
missingness (Allison, 2002). The final sample sizes were: 1,063,588 for the 2000 Census
sample, 326,157 for the 2008 ACS sample, 3,868 for the 2000 HRS sample, and 3,094
for the 2008 HRS sample.

Measures
The outcome variables were individual income and household wealth. When
constructing the individual income measure for the HRS samples, the two components
measured at the household level (income from assets, and other household income) were
divided by two for individuals whose households contained a spousal partner. All 2000
income and wealth data are presented in 2008 dollars, using inflationary weights from the
Consumer Price Index for All Urban Consumers (CPI-U).
50

Both income and wealth are constructs known to contain considerable skew, but
since zero and negative values are possible, logarithmic transformation is unwieldy.
Therefore, an alternative transformation was used—the inverse hyperbolic sine—that
transforms data in a manner similar to the logarithmic transformation for values above 1
and provides comparable transformations of all values less than 1 (Burbidge, Magee, &
Robb, 1988; Pence, 2006). Following the work of Burbidge and colleagues, the value of
0.0001 was used for θ (Gale & Pence, 2006).
Key predictor variables were age, gender (female), race/ethnicity (non-Latino
white, non-Latino black, Latino, non-Latino Asian/Pacific Islander, or non-Latino other),
citizenship status (born as a U.S. citizen, naturalized U.S. citizen, or not a U.S. citizen),
education (less than high school, high school degree, some college, or college degree or
higher), marital status (married, divorced/separated, widowed, or single/never married),
and labor force participation (employed, unemployed, or not in the labor force). The
HRS samples did not include sufficiently large numbers of non-Latino Asian/Pacific
Islanders to warrant separating them out, so they were included in the non-Latino other
category for the wealth analyses. In the HRS, citizenship status and year of naturalization
were asked in 2006, 2008 and 2010; data from all three of these time points were used to
construct the citizenship status variables for both of the HRS samples.

Analyses
Descriptive statistics are presented for all variables across the four samples;
individual income and household wealth are further broken down by race/ethnicity. Two
nested OLS regression models are presented for each sample. The first model regresses
51

age, gender, and race/ethnicity on income (for the Census & ACS samples) or wealth (for
the HRS samples). The second model adds four additional predictor variables:
citizenship status, educational attainment, marital status, and employment status.
Unstandardized and standardized parameter estimates are presented with their
corresponding t statistics and significance levels, and fit statistics (R
2
) are presented for
each model. The marginal effects at the conditional means for each racial/ethnic group
were calculated for each model, using a calculation method similar to that which would
be employed for a log-linear model. All analyses were run in SAS version 9.2 (SAS
Institute Inc., Cary, NC).

Results
Individual Income and Household Wealth, by Race/Ethnicity
Table 3.1 presents mean and median levels of individual income (from the 2000
Census and 2008 ACS) and household wealth (from the 2000 and 2008 waves of the
HRS) for each cohort, broken down by race/ethnicity. Median income for Latinos in the
Silent Generation sample was 49.3% ($16,617 vs. $33,728) that of non-Latino whites,
and 54.3% ($19,000 vs. $35,000) for the baby boom sample. Non-Latino blacks had
moderately higher median incomes: compared to non-Latino whites, 67.7% ($22,832) for
the Silent Generation and 65.7% ($23,000) for the Baby Boom Generation. Those in the
non-Latino other category were in a similar range (63.9% [$21,564] compared to non-
Latino whites for the Silent Generation, 70.3% [$24,600] for the baby boomers), as were
non-Latino Asians (75.2% [$25,369] for the Silent Generation, 71.4% [$25,000] for the
baby boomers).
52

Table 3.1. Income and wealth, by cohort and race/ethnicity.
Silent Generation: 2000¹ Baby Boomers: 2008¹
Mean Median Mean Median
Individual Income, in 2008 dollars
Race/Ethnicity
Latino $27,619 $16,617 $28,469 $19,000
Non-Latino
White $50,715 $33,728 $50,526 $35,000
Black $33,232 $22,832 $31,142 $23,000
Asian/P.I. $45,674 $25,369 $42,081 $25,000
Other $34,788 $21,564 $37,829 $24,600
Household Wealth, in 2008 dollars
Race/Ethnicity
Latino $136,685 $58,764 $205,734 $63,000
Non-Latino
White $570,849 $231,807 $546,159 $239,000
Black $128,029 $43,761 $159,297 $31,600
Other² $342,603 $107,902 $873,314 $97,025

¹Income values are calculated from U.S. Census Bure au data sources (2000
Census [n = 1,063,558] and 2008 American Community Survey [n = 326,157]);
wealth values are calculated from two waves of the Health and Retirement Study
(2000 [n = 3,868] and 2008 [n = 3,094]). ²Too few Asian/Pacific Islanders exist in
the Health and Retirement Study samples to justify reporting them separately; they
have been included in the "Other" category. Note: P.I. = Pacific Islander.

In contrast to individual income, household wealth data showed lower median
wealth for non-Latino blacks than for Latinos. Non-Latino blacks had median wealth that
was 18.9% ($43,761 vs. $231,807) that of non-Latino whites in the Silent Generation
cohort and 13.2% ($31,600 vs. $239,000) for those in the boomer cohort, whereas
Latinos in the Silent Generation had 25.4% ($58,764) of the levels of non-Latino white
wealth and Latino boomers had 26.4% ($63,000) the level of non-Latino white wealth.
The non-Latino other category, which for the HRS-based wealth analysis included
Asians/Pacific Islanders, had considerably higher wealth levels than Latinos and non-
53

Latino blacks but were still below the levels of non-Latino whites: 46.5% ($107,902) for
the Silent Generation and 40.6% ($97,025) for the Baby Boom Generation.
Across all of the racial/ethnic groups, members of the baby boom cohort had
median levels of individual income that were higher or approximately equivalent to that
of the Silent Generation cohort at the same ages. This was most notable among Latinos
and members of the non-Latino other category, both of which had median incomes 14%
higher than those of the same age eight years earlier. Different results were seen for
levels of household wealth. While median wealth for Latinos and non-Latino whites
were mildly higher among the baby boom cohort than they had been for the Silent
Generation cohort (7.2% and 3.1% higher, respectively), the level was considerably
lower for both non-Latino blacks and others (27.8% and 10.1% lower, respectively).

Sociodemographic Characteristics
Full characteristics for the four samples are presented in Table 3.2. Due to the
application of population weights to all samples, there is considerable similarity across
most sociodemographic variables. Apart from the Silent Generation sample’s HRS data
(which had a lower proportion female, at 48.7%), the samples ranged from 50.6% to
51.9% female. The proportion of Latinos in the samples ranged from 7.1% to 8.6%, and
the proportion of non-Latino blacks ranged from 9.5% to 11.1%. A larger degree of
variation existed for citizenship status, with the HRS samples less likely than the Census
and ACS samples to have naturalized citizens (4.8% vs. 7.1% for the Silent Generation;
5.7% vs. 8.1% for the boomers) and non-citizens (1.9% vs. 4.5% for the Silent
Generation; 3.4% vs. 4.7% for the boomers).
54

Table 3.2. Demographic and economic characteristics, by analysis dataset/cohort.
Income Samples
Silent Generation Baby Boomers
(2000 Census) (2008 ACS)
n = 1,063,558 n = 326,157
Age 57.3 (2.26) 57.8 (2.34)
Female 51.99 51.93
Race/Ethnicity
Latino 7.12 8.50
Non-Latino
White 78.22 75.61
Black 9.47 10.09
Asian/Pacific Islander¹ 3.31 4.14
Other¹ 1.88 1.65
Citizenship Status
Born U.S. citizens 88.38 87.17
Naturalized citizens 7.08 8.09
Non-citizens 4.54 4.74
Education
< High School 19.20 11.69
High School degree 30.62 27.54
Some college 25.77 30.16
College degree & beyond 24.40 30.61
Marital Status
Married 71.04 67.12
Divorced/Separated 17.57 19.90
Widowed 5.99 4.87
Never married 5.40 8.10
Labor Force Participation
Employed 63.75 67.60
Unemployed 2.19 2.92
Not in labor force 34.06 29.49
Individual Income, in 2008 dollars $46,949 ($65,662) $46,135 ($61,523)

Transformed income, in 2008
dollars²
17,231 (10,604) 17,552 (10,089)
Household Wealth, in 2008 dollars N/A N/A

Transformed wealth, in 2008
dollars²
N/A N/A
Note: Units are percentages, by column, or mean (standard deviation) for continuous
variables. ACS = American Community Survey; HRS = Health and Retirement Study. ¹Too
few Asian/Pacific Islanders exist in the HRS samples to justify reporting them separately;
they have been included in the "Other" category. ² Transformed using the inverse hyperbolic
sine.

55

Table 3.2 (cont'd). Demographic and economic characteristics, by analysis dataset/cohort.
Wealth Samples
Silent Generation Baby Boomers
(2000 HRS) (2008 HRS)
n = 3,868 n = 3,094
Age 57.9 (2.30) 57.9 (2.40)
Female 48.72 50.62
Race/Ethnicity
Latino 7.70 8.57
Non-Latino
White 79.11 76.85
Black 10.89 11.06
Asian/Pacific Islander¹ N/A N/A
Other¹ 2.29 3.51
Citizenship Status
Born U.S. citizens 92.64 91.68
Naturalized citizens 4.84 5.70
Non-citizens 1.93 3.42
Education
< High School 18.41 10.84
High School degree 34.81 30.30
Some college 23.29 29.26
College degree & beyond 23.49 29.59
Marital Status
Married 66.68 66.22
Divorced/Separated 20.56 21.02
Widowed 7.89 6.55
Never married 4.88 6.21
Labor Force Participation
Employed 66.32 65.65
Unemployed 2.46 4.45
Not in labor force 31.21 29.90
Individual Income, in 2008 dollars
$56,331
($145,741)
$51,273
($79,160)

Transformed income, in 2008
dollars²
19,054
(10,209)
18,550
(10,041)
Household Wealth, in 2008 dollars
$483,933
($1,694,548)
$485,655
($1,275,721)

Transformed wealth, in 2008
dollars²
33,221
(17,762)
31,609
(20,770)
Note: Units are percentages, by column, or mean (standard deviation) for continuous
variables. ACS = American Community Survey; HRS = Health and Retirement Study.
¹Too few Asian/Pacific Islanders exist in the HRS s amples to justify reporting them
separately; they have been included in the "Other" category. ²Transformed using the
inverse hyperbolic sine.

56

Educational attainment was similar between the Census Bureau and HRS
samples, with the boomers showing slightly higher levels of academic achievement than
their Silent Generation counterparts (e.g., roughly 30% of boomers held a college degree
or higher, compared with roughly 24% of the Silent Generation cohort). Marital status
varied slightly between samples, but all four followed the general pattern of roughly one-
fifth divorced or separated and two-thirds married. Finally, roughly two-thirds (64%-
68%) of all four samples were employed at the time of the survey.

Regression Results: Income
The first regression model for the Silent Generation cohort’s individual income
expands on the bivariate statistics reported in Table 3.2. Model 1 reveals that being older
(standardized β [st. β] = -0.068) and being female (st. β = -0.382) are associated with
lower levels of income, as is being a race ethnicity other than non-Latino white (see
Table 3.3). Model 2 demonstrates that when additional sociodemographic characteristics
are considered, the direct effect of age, gender, and race/ethnicity is dramatically
mitigated. The effect of age became negligibly positive, and the effect of being female
dropped by 19% (st. β = -0.308). Profoundly, the effect of being Latino fell by 77%
(from -0.124 to -0.029, compared to non-Latino whites), non-Latino black by 76% (from
-0.075 to -0.018, compared to non-Latino whites), and Asian by 26% (from -0.031 to -
0.023, compared to non-Latino whites). Of the variables added into Model 2, being a
non-citizen had a negative effect (st. β = -0.047), whereas there was a negligible effect of
having naturalized, both when compared to those who were born as U.S. citizens. As is
typical, higher educational attainment had a strongly positive relationship with income,
57

and having left the labor force had the strongest negative effect on income (st. β =
-0.402), compared to those who were employed. For the Silent Generation cohort, the
addition of sociodemographic characteristics beyond age, gender, and race/ethnicity
increased the model’s explanatory power from 17.4% of variance to 44.6% of variance.
Income results for the Baby Boom Generation cohort were similar to those for the Silent
Generation cohort. In Model 1, the boomer cohort also exhibited a negative effect of age,
being female, and falling into any racial/ethnic group other than non-Latino white. There
was less of an effect of gender (st. β = -0.300) and age (st. β = -0.045) than there was for
the Silent Generation, but with the exception of the non-Latino other group, the
racial/ethnic differences were more pronounced. Upon the addition of sociodemographic
variables in Model 2, the effect of being female dropped again by 21%, Latino by 72%,
non-Latino black by 67%, and Asian by 42%. Educational attainment again had a
considerably positive effect on income; the effect of being a non-citizen was even greater
(st. β = -0.066), and being a naturalized citizen was linked to lower individual income (st.
β = -0.013), both compared to those who were born as U.S. citizens. The negative effects
of being out of the labor force and being unemployed were also stronger: -0.464 and -
0.151, respectively, compared to those who were employed. The percentage of variance
explained by Model 2 for the baby boomer cohort was comparable to that for the Silent
Generation cohort (45.4%), but the core demographics of Model 1 explained
considerably less of the variance in individual income among the boomers (11.9%) than it
had among the Silent Generation.

58

Table 3.3. Linear regression on income (IHS-transformed), by cohort.
Silent Generation (2000 Census; n = 1,063,558)
Model 1 Model 2
b St. β t b St. β t
Age -318 -0.068 -76.75 *** 20 0.004 5.96 ***
Female -8,111 -0.382 -433.33 *** -6,533 -0.308 -412.89 ***
Race/Ethnicity
Latino -5,095 -0.124 -139.21 *** -1,207 -0.029 -35.20 ***
Non-Latino
White (ref.) -- -- -- -- -- --
Black -2,706 -0.075 -84.10 *** -651 -0.018 -24.13 ***
Asian -1,813 -0.031 -34.58 *** -1,366 -0.023 -27.63 ***
Other -3,315 -0.042 -48.02 *** -1,582 -0.020 -27.83 ***
Citizenship Status

Born U.S. citizens
(ref.) -- -- --
Naturalized citizens 154 0.004 4.46 ***
Non-citizens -2,419 -0.047 -58.15 ***
Education
< High School (ref.) -- -- --
High School degree 1,995 0.087 85.77 ***
Some college 4,172 0.172 172.30 ***

College degree &
beyond 8,439 0.342 338.95 ***
Marital Status
Married (ref.) -- -- --
Divorced/Separated 712 0.026 34.44 ***
Widowed 2,319 0.052 69.66 ***
Never married -317 -0.007 -9.19 ***
Labor Force Participation
Employed (ref.) -- -- --
Unemployed -6,040 -0.083 -114.38 ***
Not in labor force -9,001 -0.402 -528.52 ***

Adjusted R² 0.174 0.446

Note: IHS = inverse hyperbolic sine; ACS = American Community Survey; b = parameter
estimate; St. β = standardized parameter estimate. The analyses were run using a transformed
version of the dependent variable (using the IHS transformation). *p < .05; **p < .01; ***p <
.001.

59

Table 3.3 (cont’d). Linear regression on income (IHS-transformed), by cohort.
Baby Boomers (2008 ACS; n = 326,157)
Model 1 Model 2
b St. β t b St. β t
Age -196 -0.045 -27.56 *** 84 0.019 14.80 ***
Female -6,059 -0.300 -182.42 *** -4,799 -0.238 -179.96 ***
Race/Ethnicity
Latino -4,789 -0.132 -79.79 *** -1,337 -0.037 -24.25 ***
Non-Latino
White (ref.) -- -- -- --
Black -3,083 -0.092 -55.43 *** -1,018 -0.030 -22.70 ***
Asian -2,409 -0.048 -28.78 *** -1,436 -0.028 -18.41 ***
Other -2,485 -0.031 -19.05 *** -848 -0.011 -8.24 ***
Citizenship Status

Born U.S. citizens
(ref.) -- --
Naturalized citizens -481 -0.013 -8.37 ***
Non-citizens -3,157 -0.066 -44.68 ***
Education
< High School (ref.) -- --
High School degree 1,444 0.064 30.12 ***
Some college 3,464 0.158 72.26 ***

College degree &
beyond 7,639 0.349 157.59 ***
Marital Status
Married (ref.) -- --
Divorced/Separated 99 0.004 2.94 **
Widowed 1,333 0.028 21.40 ***
Never married -945 -0.026 -19.22 ***
Labor Force Participation
Employed (ref.) -- --
Unemployed -9,024 -0.151 -115.21 ***
Not in labor force -10,274 -0.464 -342.97 ***

Adjusted R² 0.119 0.454

Note: IHS = inverse hyperbolic sine; ACS = American Community Survey; b = parameter
estimate; St. β = standardized parameter estimate. The analyses were run using a transformed
version of the dependent variable (using the IHS transformation). *p < .05; **p < .01; ***p < .001.

60

Regression Results: Wealth
In contrast to income, household wealth in the Silent Generation cohort’s Model 1
was positively impacted by age (st. β = 0.030), as expected (see Table 3.4). Being female
had a negative association with wealth (st. β = -0.093), as did being of a race/ethnicity
other than non-Latino white. Adding the additional sociodemographic variables in Model
2 increased the positive impact of age on wealth (st. β = 0.044) and dropped the impact of
being female to 0. The effect of being Latino on household wealth decreased by 44%
(from -0.201 to -0.112, compared to non-Latino whites) and being non-Latino black
decreased by 31% (from -0.286 to -0.197, compared to non-Latino whites). Non-citizens
did not have significantly less household wealth than those born in the U.S., and
naturalized citizens had higher levels than those born in the U.S. (st. β = -0.038). The
standardized effect of educational attainment, compared to those who had not completed
high school, ranged from 0.198 for a high school degree to 0.402 for a college degree or
beyond. In sharp contrast to the modest effect of marital status on income, it was
strongly related to household wealth among the Silent Generation cohort. Never being
married (st. β = -0.106) and being widowed (st. β = -0.122) had modestly negative effects
on wealth, compared to those who were married, and being divorced or separated had the
largest negative effect in the model (st. β = -0.252). The amount of variance in household
wealth explained by the model more than doubled, from 12.2% in Model 1 to 29.0% after
the additional sociodemographic variables were added in Model 2.
For the baby boomer cohort, age was again positively related to household wealth
in Model 1 (st. β = 0.055), but gender was not a significant predictor (see Table 3.4).
Although being Latino or non-Latino black had less of an effect on wealth for the
61

Table 3.4. Linear regression on wealth (IHS-transformed), by cohort.
Silent Generation (2000 HRS; n = 3,868)
Model 1 Model 2
b St. β t B St. β t
Age 232 0.030 1.99 * 339 0.044 3.16 **
Female -3,293 -0.093 -6.12 *** -39 -0.001 -0.08
Race/Ethnicity
Latino -13,399 -0.201 -13.23 *** -7,473 -0.112 -7.23 ***
Non-Latino
White (ref.) -- -- -- -- -- --
Black -16,337 -0.286 -18.85 *** -11,249 -0.197 -14.01 ***
Other -4,134 -0.035 -2.30 * -4,827 -0.041 -2.95 **
Citizenship Status
Born U.S. citizens (ref.) -- -- --
Naturalized citizens 3,159 0.038 2.66 **
Non-citizens -3,060 -0.024 -1.63
Education
< High School (ref.) -- -- --
High School degree 7,381 0.198 10.05 ***
Some college 10,377 0.247 13.10 ***

College degree &
beyond 16,859 0.402 20.72 ***
Marital Status
Married (ref.) -- -- --
Divorced/Separated -11,085 -0.252 -17.76 ***
Widowed -8,046 -0.122 -8.57 ***
Never married -8,782 -0.106 -7.70 ***
Labor Force
Participation
Employed (ref.) -- -- --
Unemployed -4,498 -0.039 -2.87 **
Not in labor force -1,359 -0.035 -2.47 *

Adjusted R² 0.122 0.290

Note: IHS = inverse hyperbolic sine; HRS = Health and Retirement Study; b = parameter
estimate; St. β = standardized parameter estimate. The analyses were run using a transformed
version of the dependent variable (using the IHS transformation). *p < .05; **p < .01; ***p < .001.

62

Table 3.4 (cont'd). Linear regression on wealth (IHS-transformed), by cohort.
Baby Boomers (2008 HRS; n = 3,094)
Model 1 Model 2
b St. β t b St. β t
Age 480 0.055 3.2 ** 502 0.058 3.65 ***
Female -1,029 -0.025 -1.43 1,639 0.039 2.48 *
Race/Ethnicity
Latino -12,184 -0.164 -9.47 *** -5,647 -0.076 -4.03 ***
Non-Latino
White (ref.) -- -- -- -- -- --
Black -16,467 -0.249 -14.31 *** -9,419 -0.142 -8.79 ***
Other -7,929 -0.070 -4.06 *** -6,214 -0.055 -3.36 ***
Citizenship Status
Born U.S. citizens (ref.) -- -- --
Naturalized citizens 2,824 0.032 1.81
Non-citizens -3,516 -0.031 -1.8
Education
< High School (ref.) -- -- --
High School degree 6,186 0.137 5.06 ***
Some college 11,459 0.251 9.25 ***

College degree &
beyond 19,510 0.429 15.4 ***
Marital Status
Married (ref.) -- -- --
Divorced/Separated -12,941 -0.254 -15.68 ***
Widowed -9,862 -0.117 -7.29 ***
Never married -11,311 -0.131 -8.32 ***
Labor Force
Participation
Employed (ref.) -- -- --
Unemployed -4,864 -0.048 -3.08 **
Not in labor force -2,245 -0.049 -3.03 **

Adjusted R² 0.085 0.263

Note: IHS = inverse hyperbolic sine; HRS = Health and Retirement Study; b = parameter
estimate; St. β = standardized parameter estimate. The analyses were run using a transformed
version of the dependent variable (using the IHS transformation). *p < .05; **p < .01; ***p < .001.

63

boomers than it had for the Silent Generation cohort, they were still significant negative
predictors (st. β = -0.164 and -0.249, respectively, compared with non-Latino whites).
The additional sociodemographic characteristics in Model 2 reduced the effect of being
Latino by 54% (to -0.076) and the effect of being non-Latino black by 43% (to -0.142).
There was no significant effect of citizenship, but the effects of educational attainment
and marital status were similar to those seen for the Silent Generation: education ranged
from st. β = 0.137 for those with a high school degree to 0.429 for those with a college
education or beyond, compared to those with less than a high school degree; being
widowed or never married had standardized effects of -0.117 and -0.131, respectively,
and being divorced/separated had roughly double the effect (st. β = -0.254). The variance
explained increased from 11.9% in Model 1 to 45.4% in Model 2.

Marginal Effects
Among the Silent Generation cohort, the marginal effect of being Latino on
individual income decreased in magnitude from -$12,637 in Model 1 to -$3,378 (a 73.8%
reduction) after adjusting for the other sociodemographic variables in Model 2. By
contrast, the marginal effect of being Latino on individual income for the Baby Boom
Generation cohort decreased from -$12,480 n Model 1 to -$3,845 in Model 2 (a 69.2%
reduction). The dramatic reductions between sociodemographic-driven income
disparities in Model 1 and the adjusted structural income disparities in Model 2 for both
cohorts confirm hypothesis 1; however, the larger structural effect of income for the Baby
Boom Generation compared to that for the Silent Generation fails to confirm hypothesis
2. In the context of the adjusted mean income level for non-Latino whites, Latinos had a
64

Table 3.5. Marginal effects of race/ethnicity on income and wealth, by cohort.
Silent Generation: 2000¹ Baby Boomers: 2008¹
Unadjusted Model 1 Model 2 Unadjusted Model 1 Model 2
Individual Income, in 2008 dollars
Race/Ethnicity
Latino -$12,886 -$12,637 -$3,378 -$12,554 -$12,480 -$3,845
Non-Latino
White (ref.) -- -- -- -- -- --
Black -$8,331 -$7,426 -$1,865 -$9,339 -$8,647 -$2,970
Asian/P.I. -$5,473 -$5,196 -$3,794 -$7,460 -$6,963 -$4,113
Other -$8,938 -$8,867 -$4,354 -$7,667 -$7,158 -$2,495
Household Wealth, in 2008 dollars
Race/Ethnicity
Latino -$139,715 -$137,186 -$88,492 -$115,457 -$114,151 -$60,733
Non-Latino
White (ref.) -- -- -- -- -- --
Black -$151,527 -$149,718 -$113,630 -$132,156 -$131,074 -$85,964
Other² -$66,463 -$62,828 -$64,342 -$90,326 -$88,630 -$65,155

¹Income values are calculated from U.S. Census Bur eau data sources (2000 Census [n = 1,063,558]
and 2008 American Community Survey [n = 326,157]); wealth values are calculated from two waves of
the Health and Retirement Study (2000 [n = 3,868] and 2008 [n = 3,094]). ²Too few Asian/Pacific
Islanders exist in the Health and Retirement Study samples to justify reporting them separately; they
have been included in the "Other" category. Note: P.I. = Pacific Islander.

mean income that was lower than non-Latino whites by 12.2% in 2000 and 13.3% in
2008. Similar patterns were apparent for both non-Latino blacks and Asians (see Table
3.5), though the 27% reduction seen for Asians was far smaller than that for the other two
racial/ethnic groups.
As with income, marked reductions in the race/ethnicity effects on wealth were
seen after adjusting for sociodemographic characteristics. Among the Silent Generation
cohort, the marginal effect of being Latino on household wealth decreased from
-$137,186 in Model 1 to -$88,492 in Model 2 (a 35.5% reduction), and among the
boomer cohort, it decreased from -$114,151 in Model 1 to -$60,733 in Model 2 (a 46.8%
65

reduction). The reductions between wealth disparities in Model 1 and the adjusted
structural disparities in Model 2 for both cohorts confirm hypothesis 3. Likewise, the
smaller structural effect for the Baby Boom Generation compared to that for the Silent
Generation confirms hypothesis 4. In the context of the adjusted mean wealth level for
non-Latino whites, Latinos had a mean wealth that was lower than non-Latino whites by
52.8% in 2000 and 43.3% in 2008. Similar patterns were apparent for non-Latino blacks
(see Table 3.5).

Discussion
This paper used data for two generational cohorts—the Silent Generation and
Baby Boom Generation—to evaluate the degree to which structural disadvantages in
economic indicators exist for racial/ethnic groups in the pre-retirement years. After
adjusting for the effects of sociodemographic variables—age, gender, citizenship status,
education, marital status, and labor force participation—the structural effects of
race/ethnicity on income and wealth were considerably reduced, confirming hypotheses 1
and 3. When comparing the two cohorts, however, the expected reduction in structural
effects from the Silent Generation to the Baby Boom Generation was seen for wealth but
not for income, confirming hypothesis 4 but failing to confirm hypothesis 2.
Although it pales in comparison to the unadjusted racial/ethnic disparities,
primarily for income, the presence of a structural disadvantage is still very real for the
broad racial/ethnic groups identified in this study. The reduction of structural disparities
in wealth from the Silent Generation to the Baby Boom Generation follows the
expectation articulated by Sandefur and Pahari (1989) that these disparities would be
66

reduced over time. As such, a continued trajectory in the direction of reduced structural
disparities can be expected, which signals good news for the younger members of the
Baby Boom Generation, Generation X, and future generational cohorts. There currently
exist large gaps in wealth levels between racial/ethnic groups, even after
sociodemographic adjustment, and a future reduction in structural inequalities can help
decrease those gaps. This will be increasingly important as the proportion of the U.S.
population that is not non-Latino white continues to grow.
The lack of findings that support the expected reduction of structural income
disparities between the two generational cohorts may reflect a true absence of such a
reduction. If this is the case, it would signal that occupational discrimination of some
sort is still in place, but that similar discrimination has less bearing on the ability of
racial/ethnic minorities to convert their income into wealth. It is possible, however, that
the absence of a reduction in structural income discrimination is due to the years in which
this study’s data were collected. The data for baby boomers were collected in the heart of
the “Great Recession,” and may reflect unique income dynamics that were in play at the
time due to heightened levels of unemployment and income insufficiency (Myers,
Calnan, Jacobsen, & Wheeler, 2012). Further research that spans different economic
climates will be necessary to discern how structural disparities are shifting in the 21
st

century.
It must be noted that, despite the relatively modest structural disparities that exist,
real differences in income and wealth based on sociodemographic disparities are stark.
The Asian population serves as a good example of the compensatory power of human
agency-driven factors such as education and naturalization. Despite being subject to
67

structural racial/ethnic disparities in income that are more pronounced than those
identified for Latinos and non-Latino blacks, the attainment of high educational levels
and naturalization among most Asian immigrants elevates the overall income of the
Asian population to a level that is above that of the other minority groups. Attainment of
similar educational levels and naturalization rates can be mechanisms for other
racial/ethnic minorities to counteract the economic pressures of structural disparity.
Additionally, improvements in financial literacy and savings behaviors are crucial to
balancing the especially disparate wealth levels, and early interventions are important to
mitigate the forces of cumulative inequality.

Limitations
Although these analyses used large, nationally representative datasets, limitations
still exist. The 2008 economic climate could have led to results that are not
representative for income, as discussed above, or for wealth due to disparate effects of the
recession by race/ethnicity (Kochhar, Fry, & Taylor, 2011) as well as age (Gassoumis,
2012). This distinction could not be made in the analyses, since they were unable to
distinguish between period and cohort effects. The relatively short timeframe between
the cohorts may also have led to an inability to detect effects. Additionally, the analysis
approach did not integrate nonparametric approaches and therefore could not take into
account growing levels of inequality between the upper and lower ends of the income and
wealth distribution.
For the sake of parsimony, this study considers the sociodemographic variables—
age, gender, marital status, education, and citizenship status—to have a homogenous
68

effect across different racial/ethnic groups. This assumption has been shown to be
invalid in other samples for education (Sandefur & Pahari, 1989); future studies on larger
samples should allow for heterogenous effects across racial/ethnic groups wherever
possible. This is underscored by the likelihood that educational attainment, a key
component of income and income disparity, can have different ramifications for
immigrants who may have received their education abroad versus U.S.-born populations
who likely received a domestic education; the marginal effect of one year of education
may vary based on where and when that education was attained.
HRS serves as a rich dataset for studying microeconomic factors, but the
participants tend to have above average levels of economic security; therefore, the results
for wealth may reflect patterns that do not generalize to the U.S. population. Also, none
of the datasets used have information on documentation status of immigrants, which is a
key component in income and wealth building as well as the ability to naturalize. This as
well as other measures not included in the model—including, but not limited to,
generation status (e.g., 1, 1.5, 2, 3+) and time since immigration for the foreign-born—
may have impacts on income and wealth that could mitigate the findings of the above
analyses.

Implications
Findings reveal that the lion’s share of the racial/ethnic disparity in income is due
to individual factors, including education and citizenship status. While this shows
promise for the next generation of Latinos, black/African Americans, and Asians
69

approaching retirement, it also points to the necessity for these groups to emphasize
education and, among immigrants, naturalization.
Education is a particularly important sociodemographic characteristic in terms of
earnings potential. Yet considerable educational disparities persist among racial/ethnic
minority youth. An often cited barrier to educational attainment is the cost, especially for
many low-income, minority, and immigrant parents who may have an aversion to
borrowing money for education or simply be inexperienced with credit markets
(Burdman, 2005; McDonough & Calderone, 2006). Increasing access to grant funding
(Gross, Torres, & Zerquera, 2013) and financial aid (Kim, DesJardins, & McCall, 2009)
is crucial for improving the accessibility of postsecondary education—especially in a
climate of soaring higher education costs—and providing information about education
financing may make these options more attractive to those who are wary (McDonough &
Calderone, 2006). But for families who are unwilling to incur debt, approaches to saving
for educational expenses in advance may be the best option. Mechanisms such as the
Individual Development Account has been shown to be an effective route to encouraging
savings for postsecondary education (Sherraden, Schreiner, & Beverly, 2003; Zhan &
Schriener, 2005) and may increase the feasibility of attending college for low-income
minorities. Documentation status may provide an additional barrier for postsecondary
financing to many immigrants who would otherwise take advantage of credit markets as a
route to educational attainment (Gonzales, 2011), primarily among “dreamers” or the 1.5
generation (Rumbaut, 2004).
Even net of educational attainment, however, disparities in wealth are particularly
severe along racial/ethnic lines. While structural disparities seem to be decreasing, there
70

is still a long way to go before the gap between racial/ethnic minorities and non-Latino
whites is eradicated. Differences exist in savings patterns (Fisher & Hsu, 2012) and
investment decisions (Plath & Stevenson, 2005) that exacerbate this disparity. However,
culturally sensitive and culturally targeted financial literacy programs have proven
promising for improving the level of comfort with financial concepts and products (Forte,
2012; Spader, Ratcliffe, Montoya, & Skillern, 2009), and offer one approach to reducing
the sharp racial/ethnic wealth divides.

71

CHAPTER 4: NATURALIZATION’S EFFECT ON INCOME GROWTH:
PRESENCE AND PERSISTENCE ACROSS THE LIFESPAN

Introduction
Within immigration studies, the attainment of citizenship status is looked to as a
clear indicator of integration into a host country’s society. An immigrant’s decision to
acquire citizenship—through the process of naturalization—signals the intent to remain
in the host country and invest in becoming a fully vested member of the society’s
structure (Roberts, 1996). At the same time, the act of moving from a becoming a citizen
can effect or increase feelings of belonging in and attachment to the immigrant’s host
country (Bloemraad et al., 2008).
Attainment of citizenship in the United States is most commonly associated with
the right to vote, and voting rights are indeed the primary motivator for the pursuit of
naturalization (Passel, 2007). But with naturalization come several other tangible
benefits over those of legal permanent residents. In addition to voting rights, other civic
rights associated with citizenship include running for public office and serving on a
federal jury. Economic benefits afforded to citizens include: being eligible for certain
jobs, including a wide range of federal and other public sector positions, that are not
available to legal residents; being free from an employer or perspective employer’s right
to give hiring preference to citizens over non-citizens, all else being equal; being eligible
to obtain the full range of federal grants and scholarships available for education and
other purposes; and being able to gain access to all public/governmental benefits.
Additional benefits afforded to naturalized citizens include being given priority over non-
72

citizens when requesting that family members be brought to the U.S. and having a
freedom from the legal permanent resident’s prospect of deportation in extremely rare
circumstances (e.g., conviction for a felony). All of these benefits are deemed to be
inalienable once the naturalization process is complete, barring the discovery of
misrepresented facts on the naturalization application (Bloemraad et al., 2008; Bratsberg
et al., 2002; DeSipio, 2011; Mazzolari, 2009; U.S. Citizenship and Immigration Services,
2007). These benefits do, however, come at a premium. Prospective citizens must
complete a lengthy form, demonstrate knowledge about U.S. civics and history,
understand and accept the principles of the U.S. Constitution, and pay a $675 application
fee (DeSipio, 2011; U.S. Citizenship and Immigration Services, 2007).
Another benefit that has been linked to naturalization is increased levels of
economic security. Several studies have shown that immigrants who are naturalized
citizens have higher levels of income than non-citizen immigrants (e.g., Gassoumis et al.,
2010; Sumption & Flamm, 2012). Growth in income has also been linked to
naturalization, using cross-sectional (Chiswick, 1978), repeated cross-sectional
(Mazzolari, 2009), and longitudinal samples (Bratsberg et al., 2002). Using longitudinal
data to detect a direct relationship between naturalization and income growth, Bratsberg
and colleagues (2002) identified three possible mechanisms for a link between income
and naturalization: the increased access to employment, discussed above; a commitment
to remain in the U.S. that drives the acquisition of human capital, including citizenship;
and unmeasured differences in productivity between those who naturalize and those who
do not. Based on their analysis model, though, the continued acquisition of human
73

capital explanation could not have driven the relationship between naturalization and
income growth.

Naturalization and Segmented Assimilation
In discussing the relationship between naturalization and income growth,
Bratsberg and colleagues (2002) specified that the act of naturalization can accelerate
immigrants’ labor market assimilation, a perspective that is in-line with the classical
straight-line model of immigrant assimilation. However, the link between naturalization
and income can also be framed in terms of segmented assimilation, a prominent approach
to the study of immigrant assimilation. According to the theory of segmented
assimilation, immigrants can utilize three strategies for assimilation into the host country:
1) assimilating toward the majority norm, similar to the traditional model of straight-line
assimilation; 2) selective assimilation, in which upward mobility is achieved through
embracing one’s cultural values and solidarity; and 3) “downward assimilation” into a
niche within the majority society that reinforces poverty, such as gang membership
(Bloemraad et al., 2008; Portes & Rumbaut, 2001; Portes & Zhou, 1993). While the
strategies and trajectories under segmented assimilation were originally developed with
the immigrant second generation in mind (Portes & Zhou, 1993), their typology can also
be applied to the assimilation of immigrants themselves.
Within the segmented assimilation framework, naturalization can be seen in two
ways. In line with traditional straight-line assimilation, it can be viewed as a step down
the path of an immigrant assimilating into majority society. Alternatively, it can
74

represent a move away from a pattern of downward assimilation and toward one of the
other two assimilational approaches in search of upward economic mobility.

Age Differences in Lifespan Income Trajectories
Research on income and aging has long acknowledged that income growth is not
consistent across the lifespan. Early in the worklife, incomes exhibit the steepest growth
rates. Workers tend to accept lower paid positions to gain entry into the workforce and
begin to build their occupational human capital. As workers gain experience and see
increases in their human capital, their incomes rise as well (Ben-Porath, 1967; Mincer,
1997). Younger workers are more likely to exhibit career mobility and less likely to get
stuck in their jobs through the forces of occupational embeddedness (Feldman & Ng,
2007). By contrast, workers later in the worklife will have fewer gains to their human
capital and may begin to reduce their working hours in the face of family commitments
and/or the shift to retirement, resulting in slower growth and eventual downturn in
earnings around midlife (Mincer, 1997; Willson, 2003). These general patterns of growth
and decline can be observed across racial/ethnic lines (Willson, 2003).
Of the three existing studies on income growth and naturalization, one focused
exclusively on younger age groups (Bratsberg et al., 2002), and as such avoided the effect
of heterogeneity in income growth across the worklife. The other two analyzed all age
groups together and generated results for the naturalizing immigrant population as a
whole. Based on the ties between age and income, a fixed effect of naturalization on
income growth across the worklife seems improbable. The upward shift in income that
comes from naturalizing is unlikely to be comparable for young adults, who are still in
75

the stage of acquiring human capital and exhibit increased career mobility, and older
working-age adults, who are largely stable in their careers and exhibit less growth, and
even downturns, in income.

Hypotheses
This study takes an initial step toward disaggregating by age the effect of
naturalization on income growth. It first seeks to use growth curve modeling to replicate
past empirical findings across the entire lifespan, then subsets the analysis for different
age groups. Two specific hypotheses are tested:

Hypothesis 1: Those who naturalize during the study period will see higher levels of
income growth than the growth seen for the population in general.
Hypothesis 2: There are differential effects of naturalization on income growth by age,
with younger workers exhibiting more growth than older workers.

Methods
The present study assessed the growth of individual income over time using data
from the Survey of Income and Program Participation (SIPP). SIPP is a large-scale
survey administered by the U.S. Census Bureau, which collects nationally representative
panel data on individuals and households in the U.S. New panels are initiated every four
to six years and in recent years have included roughly 50,000 households in the initial
wave. Individuals from the first wave are tracked for several years, even if they move
households. Data are collected on household composition, sociodemographic
76

characteristics, employment information, sources and amounts of income and wealth,
participation in public benefit programs, and health insurance status.
The 2004 SIPP panel was the first wave to collect citizenship status longitudinally
during the study period. The 2004 panel began in February 2004 with 51,379 households
(110,269 individuals) and tracked them until the final data collection in January 2008. As
is standard practice with the SIPP, data collection is conducted year-round in four
rotation groups, with each group being surveyed three times a year. At each survey,
respondents are asked to report their income for the preceding four months; as such,
income data for the 2004 panel are for October 2003 through December 2007.

Measures
The outcome variable of interest for this study was individual income. Individual
income included all sources and was reported over the course of 48 months for each of
the rotation groups. For this analysis, consistency in the duration of reporting was given
priority over consistency of time periods reported, so the first 12 months of income data
were aggregated to create year 1 income, regardless of the data collection group (year 1
of income represented either Oct 2003-Oct 2004, Nov 2003-Nov 2004, Dec 2003-Dec
2004, or Jan 2004-Jan 2005). Since these timing of the data collection periods was so
close, inflationary weights were deemed unnecessary. Additionally, inflationary
adjustments were not made based on study year; therefore, income metrics reflect
nominal and not real dollars.
Income is a construct known to contain considerable skew, but since zero and
negative values are possible, logarithmic transformation is unwieldy. Therefore, an
77

alternative transformation was used—the inverse hyperbolic sine (IHS)—that transforms
data in a manner similar to the logarithmic transformation for values above 1 and
provides comparable transformations of all values less than 1 (Burbidge, Magee, & Robb,
1988; Pence, 2006). Following the work of Burbidge and colleagues, the value of 0.0001
was used for θ (Gale & Pence, 2006).
The primary predictor variable of interest to this study is citizenship status. For
the analyses described below, four citizenship groups were used: 1) those individuals
who reported being born as U.S. citizens; 2) those individuals who reported having
attained U.S. citizenship by the beginning of the study period, either through
naturalization, serving in the U.S. military or having a spouse serve in the military, or
being adopted by a U.S. citizen parent; 3) those individuals who were non-citizens at the
beginning of the study period but reported becoming citizens during the study period,
either through naturalization or military service; and 4) those individuals who were non-
citizens throughout the study. Those respondents with conflicting citizenship information
across the survey’s waves (n = 2,062; 2.6%) were excluded. The most common errors
were being reported as a U.S.-born citizen followed by being reported as a non-citizen
(n = 600) or a naturalized citizen (n = 443), and being reported as a naturalized citizen
followed by being reported as a U.S.-born citizen (n = 398). For convenience, all groups
that have acquired citizenship are referred to hereafter as naturalized citizens, regardless
of their route to citizen attainment.
Other predictor variables included age, age squared (to allow for a curvilinear
effect of age), gender (coded as female), race/ethnicity (non-Latino white, non-Latino
black, Latino, non-Latino Asian/Pacific Islander, or other non-Latino), marital status
78

(married, widowed, divorced/separated, or never married), and educational attainment
(less than high school, high school degree, some college, or college degree or higher).
Stratification by age was accomplished by dividing the sample into those aged under 40
vs. those aged 40 and above. This cutpoint was chosen to be in line with the federal
definition of an older worker for the purposes of age discrimination, as established by the
Age Discrimination in Employment Act of 1967 (29 USC §631(a)). As the workforce
was of primary interest in this analysis, all individuals younger than age 25 or older than
64 when the study began were excluded from the sample.

Analyses
The income data from SIPP’s 2004 wave were analyzed using linear growth curve
modeling. This technique can be thought of as modeling income data for every
individual in a sample to achieve each individual’s unadjusted intercept and slope for
income, then analyzing those slopes and intercepts at the sample level. This is done with
a system of equations on two levels. The first level of analysis contains a single equation
to calculate the intercepts and slopes for income trajectory (IHS-transformed) at the
individual level—the random effects.
(1) y
i,t
= g
0,i
+ g
1,i
* t + ε
i,t

In Equation1, y
i,t
represents income for each individual i at each time t, g
0,i

represents the intercept for income, and g
1,i
represents the slope for income. As discussed
in the introduction, income generally follows a curvilinear pattern over the worklife;
however, a linear curve was deemed sufficient given the study’s four-year timeframe.
79

The second level of analysis contains two equations to predict the intercepts and slopes at
the sample level.
(2) g
0,i
= α
0
+ X
0,i
* β
0
+ ε
0,i

(3) g
1,i
= α
1
+ X
1,i
* β
1
+ ε
1,i

Where α represents the constant (or intercept) for intercept and slope, X represents
the vector of predictor variables, and β represents the vector of parameter estimates for
these predictor variables—the fixed effects. The same predictor variables made up X
0
in
the intercept model and X
1
in the slope model: age, age squared, gender, race/ethnicity,
marital status, educational attainment, and naturalization status. The marginal effects at
the conditional means for each citizenship group were calculated for both intercept and
slope, using a calculation method similar to that which would be employed for a log-
linear model.
Inclusion criteria were being surveyed in the first wave of data collection and
being between 25 and 64 years old as of October 2003; exclusion criteria were
naturalizing within the final year of data collection and containing errors in citizenship
status across waves. After applying these criteria, the final analytic sample included
55,830 individuals. Weights were applied to make the sample nationally representative,
with separate weights generated for each sub-sample analyzed. The analysis dataset was
compiled and descriptive statistics were computed in SAS version 9.2 (SAS Institute Inc.,
Cary, NC), and growth curve analyses were run using Mplus version 6.12 (Muthén &
Muthén, Los Angeles, CA).

80

Results
Characteristics of the full sample (N = 55,830) and comparisons of the younger
and older sub-samples (n = 21,821 and 34,009, respectively) are shown in Table 4.1. Due
to weighting, differences reflect population-based demographic differences between the
age groups and were seen on all variables except for gender. The younger sample was far
more Latino and slightly more black and Asian than the older sample. There were more
immigrants in the younger sample (17.1% vs. 11.6%), but far fewer had naturalized at the
beginning of the study. More of the younger sample naturalized by the study’s end than
the older sample, but at the end of the study 68.2% of the younger immigrants still lacked
citizenship (vs. 40.8% of the older immigrants). Three times as many of the younger
sample had never been married, with almost none of the younger sample having been
widowed. The younger sample had slightly higher levels of education, though the
differences were minimal. Mean annual income in year 1 was 17.8% higher among the
older sample, but median income was only 8.6% higher, suggesting greater skew among
the older sample, a realization of cumulative inequalities across the lifespan (Ferraro &
Shippee, 2009).
Table 4.2 shows characteristics of immigrant groups within the younger sub-
sample along with two comparisons: those who naturalized prior to the study (A) vs.
those who naturalized during the study (B), and those who naturalized during the study
(B) vs. those who remained non-citizens at the study’s close (C). Small differences were
seen in age, with the naturalized citizens older than the non-citizens. There were no
significant differences by gender or marital status, though slightly more non-citizens were
male and never married. All three groups were different on race/ethnicity, with those
81

Table 4.1. Demographic and economic characteristics.

Total
Sample
Age Group Samples
25-39 40-64
n = 55,830 n = 21,821 n = 34,009 χ²/t statistic
Age at Baseline
43.36 32.19 50.53 386.60 ***
(10.76) (4.33) (6.85)
25-39 39.08 100.00 0.00
40-64 60.92 0.00 100.00
Female 51.07 50.56 51.40 3.77
Race/Ethnicity 989.97 ***
Latino 11.51 16.23 8.49
Non-Latino
White 70.98 64.45 75.16
Black 11.49 12.31 10.97
Asian/Pacific Islander 3.53 4.30 3.04
Other 2.48 2.71 2.34
Citizenship Status 1,035.47 ***
Born U.S. citizens 86.25 82.94 88.38
Naturalized before Oct 2003 5.35 4.12 6.14

Naturalized during study
period 0.96 1.30 0.75
Non-citizens 7.43 11.63 4.74
Marital Status 4,059.41 ***
Married 64.88 59.42 68.39
Widowed 2.10 0.35 3.22
Divorced/separated 15.69 11.06 18.66
Never married 17.33 29.17 9.73
Education 38.98 ***
< High school 11.11 11.12 11.10
High school degree 24.72 23.55 25.47
Some college 35.78 35.71 35.83
College degree & beyond 28.40 29.63 27.60
Individual Income, annual
$35,207 $31,758 $37,420 16.42 ***
($42,137) ($35,114) ($45,872)
Median income $26,061 $24,914 $27,046

Transformed income¹
16,077 15,470 16,466 13.15 ***
(8,805) (8,550) (8,941)

Note: Units are percentages, by column, or mean (standard deviation) for continuous variables, unless
otherwise specified. Income measures are taken from baseline. ¹Transformed using the inverse
hyperbolic sine. *p < .05; **p < .01; ***p < .001

82

Table 4.2. Demographic and economic characteristics of younger (age 25-39) immigrants, by citizenship status.

A: Naturalized
Prior to Study
B: Naturalized
During Study
C: Non-citizens A vs. B B vs. C
n = 884 n = 279 n = 2,493
χ²/ t
statistic
χ²/ t
statistic
Age at Baseline 33.33 (4.36) 32.46 (4.30) 31.52 (4.44) 2.91 ** 3.41 ***
25-39 100.00 100.00 100.00
40-64 0.00 0.00 0.00
Female 53.31 51.68 45.74 0.23 3.61
Race/Ethnicity 21.51 *** 20.61 ***
Latino 30.15 41.11 55.00
Non-Latino
White 28.70 27.53 21.36
Black 6.84 9.73 6.39
Asian/Pacific Islander 32.52 20.25 16.26
Other 1.79 1.38 0.99
Citizenship Status N/A N/A
Born U.S. citizens 0.00 0.00 0.00
Naturalized before Oct 2003 100.00 0.00 0.00
Naturalized during study period 0.00 100.00 0.00
Non-citizens 0.00 0.00 100.00

Note: Units are percentages, by column, or mean (standard deviation) for continuous variables, unless otherwise specified.
¹Transformed using the inverse hyperbolic sine. * p < .05; **p < .01; ***p < .001

83

Table 4.2 (cont’d). Demographic and economic characteristics of younger (age 25-39) immigrants, by citizenship status.

A: Naturalized
Prior to Study
B: Naturalized
During Study
C: Non-citizens A vs. B B vs. C
n = 2,625 n = 469 n = 3,730 χ²/ t statistic χ²/ t statistic
Marital Status 3.62 5.34
Married 68.86 71.32 65.71
Widowed 0.26 0.58 0.28
Divorced/separated 7.78 4.77 7.31
Never married 23.11 23.33 26.69
Education 50.05 *** 37.94 ***
< High school 9.58 25.10 38.76
High school degree 18.61 19.61 20.56
Some college 32.57 29.21 16.14
College degree & beyond 39.23 26.09 24.55
Individual Income, annual $36,547 ($36,660) $25,506 ($26,836) $21,375 ($27,107) 5.45 *** 2.42 *
Median income $29,100 $19,266 $15,896
Transformed income¹ 16,747 (9,272) 13,426 (9,194) 11,914 (8,569) 5.23 *** 2.77 **

Note: Units are percentages, by column, or mean (standard deviation) for continuous variables, unless otherwise specified. ¹Transformed using
the inverse hyperbolic sine. *p < .05; **p < .01; ***p < .001

84

who naturalized more likely to be Asian and those who didn’t naturalize more likely to be
Latino. Large differences were apparent for education, with 38.8% of the non-citizen
group having less than a high school education, compared with 25.1% of those who
became naturalized during the study and just 9.6% of those who had naturalized before
the study. Mean and median income at baseline were highest for those who had
naturalized prior to the study, with non-citizens earning the least.
Comparable comparisons for the older sub-sample are shown in Table 4.3. No
significant age difference was seen between the two naturalization groups, but both were
slightly older than non-citizens. No significant differences were seen for gender or
marital status. As with the younger group, those who naturalized were more likely to be
Asian and those who didn’t naturalize more likely to be Latino, and naturalized groups
had higher education than non-citizens. Income at baseline was higher for the older
sample who had naturalized prior to the study than for the other two groups; however, no
difference was seen between the non-citizen group and the group that naturalized during
the study period.

Growth Curve Results
The growth curve model for the full sample revealed multiple predictors for the
model intercept, but few for the model slope (see Table 4.4). Of the citizenship status
variables, there was no difference in intercept or slope between those who were born as
U.S. citizens and those who had naturalized prior to the study period. Being a non-citizen
at baseline and remaining one throughout the study period was found to have a
considerably negative effect on intercept (β
0
= -2377) and a positive effect on slope

85

Table 4.3. Demographic and economic characteristics of older (age 40-64) immigrants, by citizenship status.

A: Naturalized
Prior to Study
B: Naturalized
During Study
C: Non-citizens A vs. B B vs. C
n = 2,097 n = 255 n = 1,618
χ²/ t
statistic
χ²/ t
statistic
Age at Baseline 50.98 (7.53) 50.13 (7.62) 48.60 (6.98) 1.69 3.41 ***
25-39 0.00 0.00 0.00
40-64 100.00 100.00 100.00
Female 54.46 50.27 50.29 1.60 3.56
Race/Ethnicity 30.04 *** 20.33 ***
Latino 27.59 41.81 49.70
Non-Latino
White 31.51 25.86 27.35
Black 9.46 12.26 7.43
Asian/Pacific Islander 29.44 18.28 14.42
Other 2.01 1.78 1.10
Citizenship Status N/A N/A
Born U.S. citizens 0.00 0.00 0.00
Naturalized before Oct 2003 100.00 0.00 0.00
Naturalized during study period 0.00 100.00 0.00
Non-citizens 0.00 0.00 100.00

Note: Units are percentages, by column, or mean (standard deviation) for continuous variables, unless otherwise specified.
¹Transformed using the inverse hyperbolic sine. * p < .05; **p < .01; ***p < .001

86

Table 4.3 (cont’d). Demographic and economic characteristics of older (age 40-64) immigrants, by citizenship status.

A: Naturalized
Prior to Study
B: Naturalized
During Study
C: Non-citizens A vs. B B vs. C
n = 2,097 n = 255 n = 1,618 χ²/ t statistic χ²/ t statistic
Marital Status 1.96 5.26
Married 76.61 76.58 72.96
Widowed 3.20 2.68 2.99
Divorced/separated 14.05 12.61 14.44
Never married 6.14 8.13 9.61
Education 64.07 *** 37.42 ***
< High school 18.44 38.17 46.46
High school degree 20.20 18.92 15.12
Some college 28.39 26.77 17.92
College degree & beyond 32.97 16.13 20.49
Individual Income, annual $36,979 ($54,434) $23,497 ($48,418) $22,678 ($36,830) 4.14 *** 0.26
Median income $24,744 $16,416 $15,060
Transformed income¹ 15,878 (10,122) 12,244 (8,454) 11,943 (8,900) 6.34 *** 0.51

Note: Units are percentages, by column, or mean (standard deviation) for continuous variables, unless otherwise specified. ¹Transformed using
the inverse hyperbolic sine. *p < .05; **p < .01; ***p < .001

87

(β
1
= 158), which is consistent with both an immigrant entry effect and “catch-up” or
assimilation effect (Banerjee, 2009; Li, 2003). Those who naturalized during the study
period also had a negative effect on intercept (β
0
= -1756); although it was slightly lower
in magnitude than that of non-citizens, a Wald test revealed no significant difference
between these two parameter estimates (Wald’s χ² = 2.57, df = 1; p = .109), in line with
previous findings (Bratsberg et al., 2002). Despite the expectations of hypothesis 1,
however, there was no effect of naturalizing during the study period on slope. In contrast
to previous studies, this analysis failed to find an effect of naturalization on income
growth for the entire age sample. The model predicted over a quarter of the variance in
the intercept (R
2
0
= 0.287) but less than one percent of the variance in the slope
(R
2
1
= 0.007). Nonetheless, available fit statistics indicated acceptable goodness-of-fit
(RMSEA = 0.021; SRMR = 0.008).
Results for the younger sub-sample’s growth curve were generally similar to
those for the entire sample (Table 4.4). Among citizenship status groups, the primary
difference when compared to the full sample is that the younger sample revealed an
intercept effect for those who had naturalized before baseline (β
0
= 965). Negative
effects were detected for those who naturalized during the study and non-citizens that
were comparable to those for the full sample, and the parameter estimates again showed
no significant difference from each other (Wald’s χ² = 2.25, df = 1; p = .134). Non-
citizen was again the only citizenship group that significantly affected slope, with an
effect larger in magnitude than what it was for the entire sample (β
1
= 224). Of relevance
to other parts of this dissertation, the growth parameter for Asians was significantly
positive (β
1
= 312), whereas no other race/ethnicity variable had a significant effect on

88

Table 4.4. Linear growth curve models, full sample and by age group.
Full Sample Younger (aged 25-39) Older (aged 40-64)
(n = 55,830) (n = 21,550) (n = 34,280)
Β St. β p β St. β p β St. β p
Intercept
Age at baseline 236.504 0.309 *** 319.72 0.173 *** 212.58 0.174 ***
Age squared -5.911 -0.301 *** -12.164 -0.097 *** -5.626 -0.244 ***
Female -5443.69 -0.330 *** -5668.87 -0.356 *** -5254.83 -0.312 ***
Race/Ethnicity (vs. white, non-Latino)
Latino -721.28 -0.028 *** -575.00 -0.027 ** -911.27 -0.030 ***
Black, non-Latino -1088.80 -0.042 *** -970.71 -0.040 *** -1126.86 -0.042 ***
Asian/Pacific Islander, non-Latino -889.40 -0.020 *** -1261.64 -0.032 *** -498.32 -0.010
Other, non-Latino -1447.38 -0.027 *** -862.68 -0.018 ** -1868.84 -0.034 ***
Citizenship Status (vs. born U.S. citizen)
Naturalized before Oct 2003 99.90 0.003 965.11 0.024 ** -333.45 -0.010
Naturalized during study period -1755.98 -0.021 *** -1415.62 -0.020 ** -2044.52 -0.021 ***
Non-citizens -2377.31 -0.076 *** -2231.43 -0.090 *** -2592.62 -0.065 ***
Marital Status (vs. married)
Widowed 1464.95 0.025 *** 2405.24 0.018 ** 1308.47 0.027 ***
Divorced/Separated 348.01 0.015 *** 594.95 0.023 *** 205.25 0.009
Never married -878.55 -0.040 *** -267.72 -0.015 * -1759.91 -0.062 ***
Education (vs. < high school)
High School degree 2717.73 0.142 *** 2536.83 0.135 *** 2737.13 0.142 ***
Some college 4779.97 0.277 *** 4524.02 0.273 *** 4858.02 0.277 ***
College degree & beyond 9508.42 0.519 *** 8904.02 0.511 *** 9852.69 0.523 ***

89
Table 4.4 (cont’d). Linear Growth Curve Models, Full Sample and by Age Group.
Full Sample Younger (aged 25-39) Older (aged 40-64)
β St. β p β St. β p β St. β p
Slope
Age at baseline -23.918 -0.149 *** -45.444 -0.108 * -52.855 -0.218 **
Age squared 0.378 0.092 ** 2.03 0.071 0.917 0.201 **
Female 27.48 0.008 -66.82 -0.018 77.61 0.023 *
Race/Ethnicity (vs. white, non-Latino)
Latino -28.92 -0.005 -112.45 -0.023 80.18 0.013
Black, non-Latino -65.86 -0.012 -16.29 -0.003 -98.20 -0.018
Asian/Pacific Islander, non-Latino 64.27 0.007 311.86 0.035 * -165.25 -0.017
Other, non-Latino -5.65 -0.001 -109.63 -0.010 82.64 0.007
Citizenship Status (vs. born U.S. citizen)
Naturalized before Oct 2003 -36.49 -0.005 -79.36 -0.009 -2.25 0.000
Naturalized during study period 93.97 0.005 147.32 0.009 13.27 0.001
Non-citizens 157.58 0.024 * 224.38 0.040 * 74.47 0.009
Marital Status (vs. married)
Widowed -13.65 -0.001 -112.28 -0.004 -22.81 -0.002
Divorced/Separated 21.67 0.005 -16.06 -0.003 40.06 0.009
Never married 121.67 0.027 ** 59.64 0.015 192.01 0.034 ***
Education (vs. < high school)
High School degree -37.68 -0.009 -57.31 -0.013 -23.32 -0.006
Some college -16.74 -0.005 63.40 0.017 -55.51 -0.016
College degree & beyond -70.85 -0.018 42.77 0.011 -133.76 -0.036 *
Model Fit
χ² (df; p) 1061.59 (df=40; p<.001) 541.19 (df=40; p<.001) 577.84 (df=40; p<.001)
RMSEA 0.021 0.024 0.020
SRMR 0.008 0.010 0.007

Note: RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual.
90

growth in this or the full sample model. The younger sub-sample model predicted similar
amounts of variance as the full sample model (R
2
0
= 0.289; R
2
1
= 0.007) and also showed
acceptable goodness-of-fit based on available statistics (RMSEA = 0.024; SRMR =
0.010).
The older sample’s growth curve was also similar to those from the other two
analyses. Contrary to the younger sample, there was no significant intercept effect of
having naturalized prior to the study. However, similar to the younger sample,
naturalizing during the study period and being a non-citizen throughout the study’s
duration both had negative effects on the intercept that were not significantly different
from each other (Wald’s χ² = 1.00, df = 1; p = .323). None of the citizenship status
variables had an effect on the slope of the older sub-sample. Paired with the similar null
finding for the naturalization parameter on slope for the younger sub-sample, this finding
fails to support hypothesis 2. An interesting race/ethnicity finding of tangential relevance
to this study is that the Asian/Pacific Islander effect on the intercept was not significant,
whereas all other race/ethnicity variables across all models had significant intercept
effects. The older sample model also accounted for similar amounts of variance as the
full sample model (R
2
0
= 0.285; R
2
1
= 0.005) and showed acceptable goodness-of-fit
based on available statistics (RMSEA = 0.020; SRMR = 0.007).

Marginal Effects
Translating the IHS-transformed results above into marginal effects at the
conditional mean (see Table 4.5) revealed a $5,645 disparity in the income intercept
among non-citizens compared with those born as U.S. citizens, but a $259 larger annual
91

Table 4.5. Marginal effects of citizenship status on income (intercept and slope), full sample and
by age group.

Full
Sample
Younger
(aged 25-39)
Older
(aged 40-64)
Intercept
Citizenship Status
Born U.S. citizens -- -- --
Naturalized before Oct 2003 $265 $2,532 -$897
Naturalized during study period -$4,284 -$3,331 -$5,091
Non-citizens -$5,645 -$5,068 -$6,303
Slope
Citizenship Status
Born U.S. citizens -- -- --
Naturalized before Oct 2003 -$94 -$154 -$12
Naturalized during study period $151 $245 -$7
Non-citizens $259 $341 $117

change in income for non-citizens (an elasticity at the conditional mean of 1.38%). The
direction of these effects was consistent across the two age groups, though the intercept
effect size was larger for the older cohort and the slope effect size was larger for the
younger cohort. The slopes represent a 1.90% elasticity for the younger cohort and a
0.61% elasticity for the older cohort at their conditional means.
Those who became naturalized citizens during the study period exhibited a $4,284
disparity in income intercept compared to those born as U.S. citizens, but a $151 larger
annual change in income (an elasticity of 0.75% at the conditional mean). Both of these
effects were smaller than for the non-citizens, but in the same direction. Again, the older
sample exhibited a larger marginal effect size than the younger sample for the intercept;
however, the annual change in income for the older sample was slightly negative and
negligible (-$7; an elasticity of 0.03%), compared to a fairly sizeable effect for the
92

members of the younger cohort who naturalized during the study period ($245; an
elasticity of 1.25%).
Finally, those who had naturalized prior to the study period showed a small
intercept effect for the full sample when compared to those born as U.S. citizens ($265)
but slightly more sizeable slope effects (-$94; an elasticity of -0.38% at the conditional
mean). The small intercept effect masked sizeable advantages seen in the younger
sample’s intercept ($2,532) compared to modest but real disadvantages among the older
cohort (-$897). The disparity in one-year change for members of the older cohort who
had naturalized prior to the study period was negligible (-$12; an elasticity of 0.04%) but
slightly more pronounced for their counterparts in the younger cohort (-$154; an
elasticity of -0.60%).

Discussion
To detect the effect of naturalization on immigrants’ income growth at different
points in the worklife, this study employed linear growth curve modeling on data from
SIPP’s 2004 panel. Contrary to the expectations set by previous research (Bratsberg et
al., 2002; Chiswick, 1978; Mazzolari, 2009) and laid out in the study’s two hypotheses,
no relationship was found between income growth and naturalization during the study
period. The effect was not detected on a sample level (hypothesis 1), nor was there
variation found based on age group (hypothesis 2), a puzzling discrepancy when
compared with the existing literature. In fact, the predictor variables used in this study
appeared to have minimal relationships with slope in the model, with less than 1% of
variance explained in each model. Although an effect of naturalization may have
93

appeared were more years of data available, it was not detected over the 4-year study
period.
There was, however, an unexpected difference between the age groups in the
effect of naturalization on the intercept model. Individuals in the younger sample who
had naturalized prior to the study’s start had higher intercepts than U.S.-born citizens,
whereas no difference was observed for the older sample. The presence of this effect for
the younger sample appears to be analogous to a “catch-up” or assimilation effect on
immigrant income (Banerjee, 2009; Li, 2003). These assimilation effects have been
shown in a Canadian sample to result in income levels that surpassed the mean income of
the non-immigrant population (Meng, 1987); perhaps similar forces were at work for the
younger sample in this study. The older sample, by contrast, appears to display
traditional assimilation effects, in which the immigrant population catches up to but does
not surpass the income of the non-immigrant population (Li, 2003), which is similar to
the findings for naturalized populations reported in a U.S. sample of adults late in the
worklife from 2000 (Gassoumis et al., 2010).
A third interesting finding to emerge from the study came out of the cross-
tabulations by age group. In the younger sample, those who naturalized during the study
represented a socioeconomic midpoint of sorts between the non-citizens and those who
had naturalized prior to the study. They were roughly halfway between the other two
groups in terms of racial/ethnic composition, education, and income. By comparison,
those in the older sample who naturalized during the study period were qualitatively
more similar socioeconomically to those in the sample who were still non-citizens at the
end of the study than they were to those who had naturalized prior to the study. This
94

appeared to be the case for the race/ethnicity and education variables, but it was most
clear for income, on which there was no statistically significant difference between those
who naturalized during the study period and non-citizens.
Within the framework of segmented assimilation and naturalization as proposed
in the discussion, this differentiation by age group between those who naturalize could
only be explained by younger non-citizens incorporating naturalization into their straight-
line or selective assimilation, while older individuals use it as a means to escaping
patterns of downward assimilation. This explanation, while possible, is perhaps not the
most straightforward of explanations.
Two alternatives emerge to the segmented assimilationist interpretation of these
findings. Both involve similar reasons for naturalizing at younger ages as was provided
in the explanation above: the motivation of the economic benefits associated with U.S.
citizenship. The first alternative to the explanation of naturalization at older ages also
involves the tangible benefits provided by citizenship. Those who naturalize at older
ages may do so with the primary motivation being to take advantage of the non-economic
benefits of citizenship. Namely, the driver may be the ability to gain priority in
sponsoring family members to enter the U.S., a long-recognized issue of great importance
in naturalization (Jasso & Rosenzweig, 1986; Yang, 1994). The second alternative
explanation to naturalization at older ages involves the costs of naturalizing. In weighing
the decision to naturalize, it may deemed too costly in terms of time or finances to
warrant pursuing at a young age, whereas the cost is less of a deterrent to older workers.
The driving force behind these age differences may, of course, be a combination of these
three explanations. It may also be due to other forces, including differences based on the
95

immigrant’s source country or emblematic of trends toward older immigrant populations
(Carr & Tienda, 2013). Further research is necessary to determine which of these
explanations is behind the age-based differences in naturalizing immigrants revealed in
this paper.

Limitations
When evaluating these findings, or lack thereof, it is important to consider several
possible limitations. First, the income measures driving this study were collected during
a time of great economic turmoil in the U.S. Roughly halfway through the study period,
the U.S. economy began its decline into the throes of the “Great Recession”, a period that
had severe economic consequences across both racial/ethnic lines and age groups
(Gassoumis, 2012; Kochhar, Fry, & Taylor, 2011). Crucially, the recession had
differential impacts based on nativity and citizenship status, with immigrant households
exhibiting declines in income—particularly large among non-citizens—while U.S.-born
households saw mild increases in income (Kochhar, 2008). These dynamics may have
impacted the presence of naturalization-based income effects during the study’s time
period. Coupled with the study’s relatively short 4-year duration, and the relatively few
individuals who naturalized during this period, the ability to detect an effect may have
been compromised. More nuanced, time-varying data about naturalization may also have
impacted the ability to detect an effect of naturalization; however, the citizenship data
available in SIPP’s 2004 panel did not allow for this level of differentiation.
For the sake of parsimony, this study considers the sociodemographic variables—
age, gender, marital status, education, and citizenship status—to have a homogenous
96

effect across different racial/ethnic groups. This assumption has been shown to be
invalid in other samples for education (Sandefur & Pahari, 1989); future studies on larger
samples should allow for heterogenous effects across racial/ethnic groups wherever
possible. This is underscored by the likelihood that educational attainment, a key
component of income and income disparity, can have different ramifications for
immigrants who may have received their education abroad versus U.S.-born populations
who likely received a domestic education; the marginal effect of one year of education
may vary based on where and when that education was attained.
In investigating naturalization’s effect on income, the methodological allowance
for a shift in level or slope at the time of naturalization—as employed by Bratsberg and
colleagues (2002)—would be the ideal approach; however, the 2008 panel of SIPP did
not provide enough information about naturalization or a long enough window of time in
which to implement this technique. Future research using the SIPP and other datasets
would be well suited to adopt these approaches wherever possible. Additionally, date of
naturalization for those who gained citizenship before the initiation of short-term panels
such as the SIPP, as well as time since immigration for immigrant groups in general,
would aid investigation into immigrant populations. Information on documentation
status would also be useful in these analyses, as undocumented individuals are inherently
unable to naturalize. However, research indicates that even immigrants with
documentation may have low uptake in naturalization (Gonzalez-Barrera, Lopez, Passel,
& Taylor, 2013), minimizing these concerns.

97

Implications
Ensuring economic sufficiency in the U.S. immigrant population is crucial not
only to support individual financial health in the worklife and retirement but also to
alleviate the macroeconomic pressures associated with supporting an economically
disparate underclass. Reducing disparities based on immigrant status is an important step
to achieving levels of economic sufficiency across immigrant and citizenship groups.
Solidifying the immigrants’ position in the U.S. through naturalization has been
suggested as a possible route to reducing these economic disparities. While this study
was unable to replicate the findings from prior studies, it did point to important
differences related to naturalization based on age. Future research into naturalization and
the economic integration of an immigrant workforce should be sure to consider age,
especially when discussing the interplay between naturalization and race/ethnicity.
It is crucial to note that all immigration and naturalization dynamics occur in a
context of federal and state policies, which can have very real effects on patterns and
dynamics within immigrant groups (e.g., Vink, Prokic-Breuer, & Dronkers, 2013). Past
changes to immigration policy have not always survived the political pressures post-
implementation; however, repeal and modification has resulted in a degree of confusion
among the immigrant and non-immigrant populations alike (Zimmerman & Tumlin,
1999). Future changes to immigration policy, especially when they involve repeals and
modifications, should be communicated with great care to minimize the extent to which
this economically disadvantaged population acts or fails to act based on misinformation.
98

CHAPTER 5: CONCLUSION

This dissertation adds to the literature on economic security and its links to
race/ethnicity, immigration, and naturalization through three discrete empirical papers.
An introductory chapter presents the context of the Baby Boom generational cohort, the
Latino population in the U.S., and the historical landscape of immigration between Latin
America and the U.S., as well as introducing the theories that frame the data and
discussions in subsequent chapters. Chapter 2 follows, describing the sociodemographic
and economic characteristics of the Latino Baby Boom Generation cohort and identifying
stark differences based on citizenship status. Non-citizen Latino boomers were found to
be much more socioeconomically disadvantaged than other Latino boomer groups, and
native-born Latino boomers were found to be comparable socioeconomially to the non-
Latino white population. Chapter 3 provides a cross-sectional look at the predictors of
income and wealth for pre-retirement age baby boomers and shows that their racial/ethnic
disparities are considerably reduced when taking into account sociodemographic
characteristics. The remaining race/ethnicity-based structural disadvantage in wealth is
considerably less than that for the Silent Generation when they were the same age, but the
structural disadvantage for income was no lower for boomers. Chapter 4 lays out
naturalization and income growth in a longitudinal context, generating null findings for
the hypothesized positive relationship between the two variables. Despite the null
findings, the sample analyses reveal important patterns and relationships when
integrating age and the study of naturalization; namely, that those who naturalize when
99

aged 40-64 look considerably more like the non-citizen population of the same age than
do those who naturalize between ages 25 and 39.
These three chapters have provided a focus on the Baby Boom cohort—explicitly
in 2 and 3 and implicitly in 4—with an aim of shedding light on the cohort’s
characteristics and dynamics in the period leading up to their retirement age. This serves
two purposes: 1) providing insights into the characteristics, demographic history, and
socioeconomic patterns of the next cohort of retires and 2) providing insights to inform
and modify practice and policy for the next pre-retirement cohort: Generation X. The use
of generational cohorts when discussing this dissertation’s findings—instead of the
smaller age- or birth year-based cohorts that are more typical in the social sciences—is a
deliberate choice made with policymakers in mind. As the names of the common
generational cohorts (baby boomers, Generation X, etc.) exist across much of the public
consciousness, they are already engrained as concepts for policymakers. Framing social
science results that involve cohorts in terms of generations instead, even if partial
generations, has anecdotally been successful in driving home the findings and making
them memorable for a non-scientific audience. And while a generational cohort frame is
sometimes applied to research within subfields of ethnic studies and migration studies,
the aging of these generations and issues surrounding population aging receive relatively
little attention in these arenas.

Implications for Theory
Chapter 2 lays out select sociodemographic, economic, and health characteristics
of Latino boomers in the context of the cumulative disadvantage/cumulative inequality
100

theory (Dannefer, 1987; Ferraro & Shippee, 2009). For this descriptive analysis, the
concept of cumulative disadvantage is used as a tool to frame the discussion. As such, no
implications for theory were developed in this chapter.
Chapter 3 also draws on cumulative inequality theory; in this chapter, it informs
the selection of variables for inclusion in the model. Based on cumulative inequality’s
provision for modifiable trajectories through human agency, the decision was made to
include variables for citizenship status and to focus on education in the presentation of
results. The theory that guided this chapter’s analysis approach, however, was the theory
of structural inequalities by race/ethnicity (Alba & Nee, 2003; Sandefur & Pahari, 1989;
Waters & Eschbach, 1995). The work of Sandefur and Pahari (1989) in particular
suggested that racial/ethnic structural inequalities would diminish in future years. By
comparing two generational cohorts at comparable ages, no evidence was found for a
reduction in structural inequality for income, but a decrease in the structural inequality
for wealth did indeed emerge. Further research should probe the structural inequalities of
race/ethnicity on income, to confirm whether it is indeed diminishing over time or if the
results in this analysis are due to either the underlying economic turbulence or the close
chronological proximity of the two cohorts. A portion of the analysis for the third
empirical chapter seems to corroborate the original proposition that structural inequalities
will decrease over time, but this result is possibly confounded by underlying age effects
in the data, which the study design of Chapter 4 was not able to address.
Chapter 4, the final empirical chapter, uses and informs the theory of segmented
assimilation as it applies to naturalization. The two hypothesis tests both revealed a null
effect, which calls into question the universality of naturalization’s effect on income
101

growth. The main theoretical developments from this chapter, however, developed from
the shortcomings of segmented assimilation to explain the differences between those who
naturalized during the study period from the younger sample versus those from the older
sample. The older sample’s naturalizing group and non-citizen group were very similar
socioeconomically, whereas the younger sample’s naturalizing group fell
socioeconomically right between the already naturalized and non-citizen groups. One
explanation for this phenomenon was proposed based on segmented assimilation, as well
as two alternative explanations: that those who naturalized in the older sample either
were driven by a desire to sponsor family members in other countries for immigration or
had foregone naturalization earlier in life due to the relatively high financial and time cost
of attaining citizenship. The most appropriate explanation of the three will need to be
determined by future research.

Implications for Research and Policy
The three chapters lay out several practical implications in addition to theoretical
developments. After presenting select sociodemographic, economic, and health
characteristics, Chapter 2 makes an admonishment to policymakers to consider the full
Baby Boom generational cohort in future discussions of entitlement reform. There is
considerable variation within the cohort, including a great deal among Latino boomers,
and only by considering the profile of the entire generational cohort can policy
approaches—particularly for entitlement programs—hope to be universally equitable.
The cross-sectional analysis in Chapter 3, using ACS and HRS, revealed
interesting results in spite of the data having been collected during the economic
102

recession. The finding of reduced racial/ethnic structural inequalities in wealth among
the baby boomers compared to the Silent Generation matches expectations (Sandefur &
Pahari, 1989) and should be monitored across generational cohorts to ensure replicability
and sustainability over time. The lack of a reduction in the racial ethnic structural
inequalities in income may have been an aspect of the economic turmoil at the time of
data collection or may have been indicative of increasing structural inequalities in
younger generational cohorts. Researchers should assess which of these is the case; if it
is found that structural inequalities are indeed widening across multiple cohorts, the issue
must be turned over to policymakers to implement changes that promote the reduction of
inequalities among racial/ethnic groups for future generations.
Finally, the longitudinal analysis of naturalization on income in Chapter 4 has
many implications for future research. As detailed above, research efforts should
characterize the motivations for naturalizing and/or delaying naturalization, focusing on
the interplay between age and time since immigration and/or documentation.
Furthermore, it will be important to replicate this study’s longitudinal approach to the
relationship between naturalization and income growth, but with a dataset that contains a
longer study period. This will allow for a confirmation of this study’s findings or,
alternatively, determine if the economic turmoil during the time of data collection and/or
the relatively short duration of data collection for the SIPP panels was driving the null
findings.

103

Concluding Remarks
Due to mostly rising levels of immigration in the United States over the past 30
years, the number of older immigrants in the U.S. will grow at a rapid pace in the near
future. Economic security in retirement is an especially pressing factor for immigrants,
many of whom are not eligible to receive Social Security. These immigrants, whose
incomes tend to be lower than those of the U.S.-born population, will face the challenge
of bolstering their economic security enough to support themselves in retirement. In the
absence of Social Security, continuing employment income, or high levels of savings,
they are left to rely on family support, whatever savings they have accumulated, and any
public assistance that is available to them. To promote retirement security among this
population, it is crucial to build economic security during the worklife by reducing or
eliminating disparities based on race/ethnicity and immigrant/citizenship status.
Barring a dramatic liberalization of Social Security benefits, future retirement
security for the current and subsequent generations of retirees will be largely dependent
on the ability to amass savings and contributing to defined contribution accounts such as
401(k) plans or Individual Retirement Accounts (IRAs). But in order for this
accumulation of wealth to be feasible, income sufficiency is of tantamount importance.
In light of this shifting retirement income system, the sharp disparities in worklife
income across racial/ethnic lines will, if left unaddressed, result in generations of
racial/ethnic minority retirees who have inadequate private assets to support their
retirement security and will continue to rely disproportionately on Social Security
income. By highlighting the state of worklife economic security for Latino baby
boomers, this dissertation has pointed to some of the key sociodemographic relationships
104

with income and wealth in the pre-retirement years for the largest minority group in the
U.S. The results can be used to inform future research and policy as part of an effort to
reduce the disparities in both worklife economic security and retirement security for
racial/ethnic minorities.

105

REFERENCES

Alba, R., & Nee, V. (2003). Remaking the American mainstream: Assimilation and
contemporary immigration. Cambridge, MA: Harvard University Press.
Allison, P. D. (2002) Missing data. Thousand Oaks, CA: Sage.
Altman, N. (2005). The battle for Social Security: From FDR’s vision to Bush’s gamble.
Hoboken, NJ: John Wiley & Sons, Inc.
Banerjee, R. (2009). Income growth of new immigrants in Canada: Evidence from the
Survey of Labour and Income Dynamics. Industrial Relations, 64(3), 466-488,
doi:10.7202/038552ar.
Baughman, R. A. (2005). Evaluating the impact of the Earned Income Tax Credit on
health insurance coverage. National Tax Journal, 58(4), 665-684.
Beedon, L., & Wu, K. B. (2004). Hispanics 65 and older: Sources of retirement income
(Fact Sheet No. FS 101). Washington, DC: AARP Public Policy Institute.
Ben-Porath, Y. (1967). The production of human capital and the life cycle of earnings.
Journal of Political Economy, 75(4), 352-365.
Bloemraad, I., Korteweg, A., & Yurdakul, G. (2008). Citizenship and immigration:
Multiculturalism, assimilation, and challenges to the nation-state. Annual Review
of Sociology, 34, 153-179, doi:10.1146/annurev.soc.34.040507.134608.
Board of Trustees. (2013). The 2013 annual report of the Board of Trustees of the federal
Old-Age and Survivors Insurance and federal Disability Insurance trust funds.
Washington, DC: U.S. Social Security Administration.
Bonacich, E. (1972). A theory of elite antagonism: The split labor market. American
Sociological Review, 37(5), 547-559.
Bratsberg, B., & Ragan, J. F., Jr. (2002). The impact of host-country schooling on
earnings: A study of male immigrants in the United States. The Journal of Human
Resources, 37(1), 63-105.
Bratsberg, B., Ragan, J. F., Jr., & Nasir, Z. M. (2002). The effect of naturalization on
wage growth: A panel study of young male immigrants. Journal of Labor
Economics, 20(3), 568-597, doi:10.1086/339616.
Burbidge, J. B., Magee, L., & Robb, A. L. (1988). Alternative transformations to handle
extreme values of the dependent variable. Journal of the American Statistical
Association, 83(401), 123-127.
106

Burdman, P. (2005). The student debt dilemma: Debt aversion as a barrier to college
access (Report No. CSHE.13.05). Berkeley, CA: Center for Studies in Higher
Education.
Butrica, B. A., Iams, H. M., Smith, K. E., & Toder, E. J. (2009). The disappearing
defined benefit pension and its potential impact on the retirement incomes of baby
boomers. Social Security Bulletin, 69(3), 1-27.
Carr, S., & Tienda, M. (2013). Family sponsorship and late-age immigration in aging
America: Revised and expanded estimates of chained migration. Population
Research and Policy Review, 32(6), 825-849, doi:10.1007/s11113-013-9300-y.
Chen, M. K. (2013). The effect of language on economic behavior: Evidence from
savings rates, health behaviors, and retirement assets. American Economic
Review, 103(2), 690-731.
Chiswick, B. R. (1978). The effect of Americanization on the earnings of foreign-born
men. The Journal of Political Economy, 86(5), 897-921.
Crimmins, E. M., Kim, J. K., Alley, D. E., Karlamangla, A., & Seeman, T. (2007).
Hispanic paradox in biological risk profiles. American Journal of Public Health,
97(7), 1305-1310.
Cruz, E. H., & Carpenter, S. (2011). We want you when we need you, otherwise get out:
The historical struggle of Mexican immigrants to obtain lawful permanent
residency in the United States. Law Journal for Social Justice, 1(1), 50-74.
Crystal, S., & Shea, D. (1990). Cumulative advantage, cumulative disadvantage, and
inequality among elderly people. The Gerontologist, 30(4), 437-443,
doi:10.1093/geront/30.4.437.
Daguerre, A. (2011). US social policy in the 21st century: The difficulties of
comprehensive social reform. Social Policy & Administration, 45(4), 389-407,
doi:10.1111/j.1467-9515.2011.00781.x.
Dannefer, D. (1987). Aging as intracohort differentiation: Accentuation, the Matthew
Effect, and the life course. Sociological Forum, 2(2), 211-236.
DeSipio, L. (2011). Immigrant incorporation in an era of weak civic institutions:
Immigrant civic & political participation in the United States. American
Behavioral Scientist, 55(9), 1189-1213, doi:10.1177/0002764211407842.
Dunlop, D. D., Song, J., Manheim, L. M., Daviglus, M. L., & Chang, R. W. (2007).
Racial/ethnic differences in the development of disability among older adults.
American Journal of Public Health, 97(12), 2209-2215.
107

Dupre, M. E. (2007). Educational differences in age-related patterns of disease:
Reconsidering the Cumulative Disadvantage and Age-As-Leveler hypotheses.
Journal of Health and Social Behavior, 48(1), 1-15.
Feldman, D. C., & Ng, T. W. H. (2007). Careers: Mobility, embeddedness, and success.
Journal of Management, 33(3): 350-377, doi:10.1177/0149206307300815.
Ferraro, K. F., & Shippee, T. P. (2009). Aging and cumulative inequality: How does
inequality get under the skin? The Gerontologist: 49(3), 333-343,
doi:10.1093/geront/gnp034.
Ferraro, K. F., Shippee, T. P., & Schafer, M. H. (2009). Cumulative inequality theory for
research on aging and the lifecourse. In V. Bengtson, M. Silverstein, N. Putney, &
D. Gans (Eds.), Handbook of theories of aging (2nd ed.) (pp. 413-433). New
York: Springer.
Fisher, P. J., & Hsu, C. (2012). Differences in household saving between non-Hispanic
white and Hispanic households. Hispanic Journal of Behavioral Sciences, 34(1),
137-159, doi:10.1177/0739986311428891.
Forte, K. S. (2013). Educating for financial literacy: A case study with a sociocultural
lens. Adult Education Quarterly, 63(3), 215-235,
doi:10.1177/0741713612460267.
Franzini, L., Ribble, J. C., & Keddie, A. M. (2001). Understanding the Hispanic paradox.
Ethnicity & Disease, 11(3), 496-518.
Gabriel, S. A., & Painter, G. (2008). Mobility, residential location and the American
dream: The intrametropolitan geography of minority homeownership. Real Estate
Economics, 36(3), 499-531.
Gale, W. G., & Pence, K. M. (2006). Are successive generations getting wealthier, and if
so, why? Evidence from the 1990s. Brookings papers on economic activity, 1,
155-234.
Gassoumis, Z. D. (2012). The recession’s impact on racial and ethnic minority elders:
Wealth loss differences by age, race and ethnicity. Los Angeles, CA: USC
Edward R. Roybal Institute on Aging.
Gassoumis, Z. D., Lincoln, K. D., & Vega, W. A. (2011). How low-income minorities get
by in retirement: Poverty levels and income sources. Los Angeles, CA: USC
Edward R. Roybal Institute on Aging.
Gassoumis, Z. D., Wilber, K. H., Baker, L. A. and Torres-Gil, F. M. (2010). Who are the
Latino baby boomers? Demographic and economic characteristics of a hidden
population. Journal of Aging & Social Policy, 22(1), 53-68.
108

Gentry, W. M., & Hubbard, R. G. (2004). “Success taxes,” entrepreneurial entry, and
innovation (Working Paper No. 10551). Cambridge, MA: National Bureau of
Economic Research.
Gonzales, R. G. (2011). Learning to be illegal: Undocumented youth and shifting legal
contexts in the transition to adulthood. American Sociological Review, 76(4), 602-
619.
Gonzalez-Barrera, A., Lopez, M. H., Passel, J. S., & Taylor, P. (2013). The path not
taken: Two-thirds of legal Mexican immigrants are not U.S. citizens. Washington,
DC: Pew Hispanic Center.
Green, T. K. (2003). Discrimination in workplace dynamics: Toward a structural account
of disparate treatment theory. Harvard Civil Rights-Civil Liberties Law Review,
38(1), 91-157.
Gregoire, T. K., Kilty, K., & Richardson, V. (2002). Gender and racial inequities in
retirement resources. Journal of Women & Aging, 14(3/4), 25-39.
Gross, J. P., Torres, V., & Zerquera, D. (2013). Financial aid and attainment among
students in a state with changing demographics. Research in Higher Education,
54(4), 383-406, doi:10.1007/s11162-012-9276-1.
Guzmán, B. (2001). The Hispanic population (Brief No. C2KBR/01-3). Washington, DC:
U.S. Census Bureau.
Halliwell, P. A., Gassoumis, Z. D., & Wilber, K. H. (2007). Social Security reform: How
various reform options will affect Latino retirees. (Research Report No. 2). Los
Angeles, CA: UCLA Center for Policy Research on Aging.
Hero, R. E., & Preuhs, R. R. (2007). Immigration and the evolving American welfare
state: Examining policies in the U.S. states. American Journal of Political
Science, 51(3), 498-517.
Ibarra, B., & Rodriguez, E. (2006). Closing the wealth gap: Eliminating structural
barriers to building assets in the Latino community. Harvard Journal of Hispanic
Policy, 18, 25-38.
Jasso, G., & Rosenzweig, M. R. (1986). Family reunification and the immigration
multiplier: U.S. immigration law, origin-country conditions, and the reproduction
of immigrants. Demography, 23(3), 291-311.
Kim, J., DesJardins, S. L., & McCall, B. P. (2009). Exploring the effects of student
expectations about financial aid on postsecondary choice: A focus on income and
racial/ethnic differences. Research in Higher Education, 50(8), 741-774,
doi:10.1007/s11162-009-9143-x.
109

Kochhar, R. (2004). The wealth of Hispanic households: 1996 to 2002. Washington, DC:
Pew Hispanic Center.
Kochhar, R. (2008). Sharp decline in income for non-citizen immigrant households,
2006-2007. Washington, DC: Pew Hispanic Center.
Kochhar, R., Fry, R., & Taylor, P. (2011). Wealth gaps rise to record highs between
whites, blacks and Hispanics. Washington, DC: Pew Social & Demographic
Trends.
Krivo, L. J., & Kaufman, R. L. (2004). Housing and wealth inequality: Racial-ethnic
differences in home equity in the United States. Demography, 41(3), 585-605.
Li, P. S. (2003). Initial earnings and catch-up capacity of immigrants. Canadian Public
Policy, 29(3), 319-337.
Markides, K. S., & Eschbach, K. (2005). Aging, migration, & mortality: Current status of
research on the Hispanic paradox. The Journals of Gerontology: Psychological
Sciences and Social Sciences, 60B(Special Issue II), 68-75.
Massey, D. S., Durand, J., & Malone, N. J. (2002). Beyond smoke and mirrors: Mexican
immigration in an era of economic integration. New York: Russell Sage
Foundation.
Massey, D. S., & Pren, K. A. (2012). Unintended consequences of US immigration
policy: Explaining the post-1965 surge from Latin America. Population and
Development Review, 38(1), 1-29, doi:10.1111/j.1728-4457.2012.00470.x.
Mazzolari, F. (2009). Dual citizenship rights: Do they make more and richer citizens?
Demography, 46(1), 169-191.
McDonough, P. M., & Calderone, S. (2006). The meaning of money: Perceptual
differences between college counselors and low-income families about college
costs and financial aid. American Behavioral Scientist, 49(12), 1703-1718, doi:
10.1177/0002764206289140.
Meng, R. (1987). The earnings of Canadian immigrant and native-born males. Applied
Economics, 19(8), 1107-1119, doi:10.1080/00036848700000051.
Mincer, J. (1997). The production of human capital and the life cycle of earnings:
Variations on a theme. Journal of Labor Economics, 15(1), S26-S47.
Myers, D., Calnan, R., Jacobsen, A., & Wheeler, J. (2012). California roller coaster:
Income and housing in the boom and bust, 1990-2010. Los Angeles, CA:
Population Dynamics Research Group, USC Sol Price School of Public Policy.
110

Office of Management and Budget. (1997). Revisions to the standards for the
classification of federal data on race and ethnicity. Washington, DC: Author.
Retrieved January 23, 2014, from
http://www.whitehouse.gov/omb/fedreg/1997standards.html
Office of Management and Budget (2013). Fiscal year 2014 historical tables: Budget of
the U.S. Government. Washington, DC: U.S. Government Printing Office.
Orszag, P. R., & Rodriguez, E. (2005). Retirement security for Latinos: Bolstering
coverage, savings and adequacy (Report No. 2005-7). Washington, DC: The
Retirement Security Project.
Ouellette, T., Burstein, N., Long, D., & Beecroft, E. (2004). Measures of material
hardship: Final report (Contract No. 282-98-0006). Washington, DC: Office of
the Assistant Secretary for Planning and Evaluation.
Palloni, A., & Arias, E. (2004). Paradox lost: Explaining the Hispanic adult mortality
advantage. Demography, 41(3), 385-415.
Passel, J. S. (2007). Growing share of immigrants choosing naturalization. Washington,
DC: Pew Hispanic Center.
Pence, K. M. (2006). The role of wealth transformations: An application to estimating the
effect of tax incentives on saving. Contributions to Economic Analysis & Policy,
5(1), 1-24.
Pew Hispanic Center. (2008). Statistical portrait of Hispanics in the United States, 2006.
Washington, DC: Author.
Plath, D. A., & Stevenson, T. H. (2005). Financial services consumption behavior across
Hispanic American consumers. Journal of Business Research, 58(8), 1089-1099,
doi:10.1016/j.jbusres.2004.03.003.
Ponce, N. A., Cochran, S. D., Mays, V. M., Chia, J., & Brown, E. R. (2008). Health
coverage of low-income citizen and noncitizen wage earners: Sources and
disparities. Journal of Immigrant and Minority Health, 10(2), 167-176.
Portes, A., & Rumbaut, R. G. (2001). Legacies: The story of the immigrant second
generation. Berkeley, CA: University of California Press.
Portes, A., & Zhou, M. (1993). The new second generation: Segmented assimilation and
its variants. Annals of the American Academy of Political and Social Science
530(1): 74-96.
RAND HRS Data, Version L (2011, December). Produced by the RAND Center for the
Study of Aging, with funding from the National Institute on Aging and the Social
Security Administration. Santa Monica, CA.
111

Reschovsky, J. D., Hadley, J., & Nichols, L. (2007). Why do Hispanics have so little
employer-sponsored health insurance? Inquiry, 44(3), 257-279.
Restifo, S. J., Roscigno, V. J., & Qian, Z. (2013). Segmented assimilation, split labor
markets, and racial/ethnic inequality: The case of early-twentieth-century New
York. American Sociological Review, 78(5), 897-924,
doi:10.1177/0003122413501071.
Roberts, B. K. (1996). Socially expected durations and the economic adjustment of
immigrants. In A. Portes (Ed.), The economic sociology of immigration (pp. 42-
86). New York: Russell Sage.
Rudkin, L., Markides, K. S., & Espino, D. V. (1997). Functional disability in older
Mexican Americans. Topics in Geriatric Rehabilitation, 12(3), 38-46.
Ruggles, S., Alexander, J. T., Genadek, K., Goeken, R., Schroeder, M. B., & Sobek, M.
(2010). Integrated Public Use Microdata Series: Version 5.0 [Machine-readable
database]. Minneapolis: University of Minnesota.
Rumbaut, R. G. (2004). Ages, life stages, and generational cohorts: Decomposing the
immigrant first and second generations in the United States. International
Migration Review, 38(3), 1160-1205.
Sandefur, G. D., & Pahari, A. (1989). Racial and ethnic inequality in earnings and
educational attainment. Social Service Review, 63(2), 199-221.
Sherraden, M., Schreiner, M., & Beverly, S. (2003). Income, institutions, and saving
performance in Individual Development Accounts. Economic Development
Quarterly, 17(1), 95-112, doi:10.1177/0891242402239200.
Shuey, K. M., & Willson, A. E. (2008). Cumulative disadvantage and black-white
disparities in life-course health trajectories. Research on Aging, 30(2), 200-225.
Smith, J. P. (1997). Wealth inequality among older Americans. Journals of Gerontology
Series B: Psychological Sciences and Social Sciences, 52B(Special Issue), 74-81.
Snyder, T. D., Dillow, S. A., & Hoffman, C. M. (2007). Digest of education statistics
2006 (Report No. NCES 2007-017). Washington, DC: National Center for
Education Statistics, Institute of Education Sciences, U.S. Department of
Education. Retrieved January 23, 2014, from
http://nces.ed.gov/programs/digest/d06/index.asp
Social Security/Medicare. (2009). Tax Management Financial Planning Journal, 25(1),
19-20.
112

Spader, J., Ratcliffe, J., Montoya, J., & Skillern, P. (2009). The bold and the bankable:
How the Nuestro Barrio telenovela reaches Latino immigrants with financial
education. The Journal of Consumer Affairs, 43(1), 56-79, doi:10.1111/j.1745-
6606.2008.01127.x.
Sumption, M., & Flamm, S. (2012). The economic value of citizenship for immigrants in
the United States. Washington, DC: Migration Policy Institutes.
Torres-Gil, F., Gassoumis, Z., & Wilber, K. (2012). Latino baby boomers: New elders for
a diverse America. In R. R. Verdugo (Ed.), The demography of the Hispanic
population: Selected essays (pp. 125-144). Charlotte, NC: Information Age
Publishing.
Torres-Gil, F., & Treas, J. (2008). Immigration and aging: The nexus of complexity and
promise. Generations, 32(4), 6-10.
U.S. Census Bureau. (2005). Supplemental measures of material well-being: Basic needs,
consumer durables, energy, and poverty, 1981 to 2002 (Report No. P23-202).
Washington, DC: Author.
U.S. Census Bureau. (2011). Profile of general population and housing characteristics:
2010 demographic profile data (Table No. DP-1). Washington, DC: Author.
U.S. Census Bureau. (2012a). Educational attainment of the population 18 years and
over, by age, sex, race, and Hispanic origin: 2012 (Tables No. 1-03 through 1-06,
Current Population Survey, Annual Social and Economic Supplement, 2012).
Washington, DC: Author.
U.S. Census Bureau. (2012b). 2012 national population projections: Downloadable files
(Table No. NP2012-D4). Washington, DC: Author.
U.S. Census Bureau. (2013). Population by sex, age, Hispanic origin, and race: 2012
(Current Population Survey, Annual Social and Economic Supplement, 2012;
Table 1). Washington, DC: Author.
U.S. Citizenship and Immigration Services. (2007). Welcome to the United States: A
guide for new immigrants (Publication No. M-618). Washington, DC: U.S.
Government Printing Office.
U.S. Immigration and Customs Enforcement (2013). FY 2013 ICE immigration removals.
Washington, DC: Author.
U.S. Public Health Service. (1948-1966). Vital statistics of the United States (1946-
1964). Washington, DC: U.S. Government Printing Office.
U.S. Social Security Administration. (2013). Social Security is important to Hispanics.
Baltimore: Author. Retrieved January 23, 2014, from
http://www.ssa.gov/pressoffice/factsheets/hispanics.htm
113

Vink, M. P., Prokic-Breuer, T., & Dronkers, J. (2013). Immigrant naturalization in the
context of institutional diversity: Policy matters, but to whom? International
Migration, 51(5), 1-20, doi:10.1111/imig.12106.
Wadhia, S. S. (2013). Response: In defense of DACA, deferred action, and the DREAM
Act. Texas Law Review, 91, 59-71.
Wang, Q. (2005). Disability and American Families: 2000 (Report No. CENSR-23).
Washington, DC: U.S. Census Bureau.
Waters, M. C., & Eschbach, K. (1995). Immigration and ethnic and racial inequality in
the United States. Annual Review of Sociology, 21, 419-446, doi:
10.1146/annurev.so.21.080195.002223.
Wickrama, K. A. S., Mancini, J. A., Kwag, K., & Kwon, J. (2013). Heterogeneity in
multidimensional health trajectories of late old years and socioeconomic
stratification: A latent trajectory class analysis. Journals of Gerontology Series B:
Psychological Sciences and Social Sciences, 68(2), 290–297,
doi:10.1093/geronb/gbs111.
Willson, A. E. (2003). Race and women's income trajectories: Employment, marriage, &
income security over the life course. Social Problems, 50(1), 87-110,
doi:10.1525/sp.2003.50.1.87.
Wu, K. B. (2013). Income and poverty of older Americans, 2011 (Fact Sheet No. 287).
Washington, DC: AARP Public Policy Institute.
Yang, P. Q. (1994). Explaining immigrant naturalization. International Migration
Review, 28(3), 449-477, doi:10.2307/2546816.
Zhan, M., & Schreiner, M. (2005). Saving for post-secondary education in Individual
Development Accounts. Journal of Sociology and Social Welfare, 32(3), 139-163.
Zimmerman, W., & Tumlin, K. C. (1999). Patchwork policies: State assistance for
immigrants under welfare reform (Occasional Paper No. 24). Washington, DC:
The Urban Institute.

Abstract (if available)

Abstract The United States is facing dramatic demographic changes due to the aging of the Baby Boom Generation and increasing diversity, including rapid growth of the Latino population. Questions have been raised regarding the economic security of the aging baby boomers’ generational cohort once they retire, which are of particular relevance to minority and Latino members of the cohort. Latinos tend to have lower levels of financial security than their white, non‐Latino counterparts, but there is little research that examines individuals who fall into the intersection of these two groups: the Latino baby boomers. Because Latino boomers are a largely hidden population, their economic status and prospects are difficult to estimate. After laying out the historical and theoretical contexts in an overarching introduction, this dissertation integrates three empirical chapters to advance knowledge in this area, by: 1) laying out selected sociodemographic, economic, and health characteristics of the cohort and drawing implications for national social insurance policies

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The dollars and sense of caregiving: exploring the financial financial realities of providing informal care to an older adult

PDF

Investigating the effectiveness of a social work intervention on reducing hospital readmissions among older adults

PDF

Fall related injuries among older adults in the Los Angeles region

PDF

Residential care in Los Angeles: policy and planning for an aging population

PDF

Estimating survival in the face of pain: evidence from the health and retirement study

PDF

A better method for measuring housing affordability and the role that affordability played in the mobility outcomes of Latino-immigrants following the Great Recession

PDF

Intergenerational transmission of values and behaviors over the family life course

PDF

The role of genetic ancestry in estimation of the risk of age-related degeneration (AMD) in the Los Angeles Latino population

PDF

Longitudinal assessment of neural stem-cell aging

PDF

The role of critical literacy and the school-to-prison pipeline: what was learned from the life histories and literacy experiences of formerly incarcerated young Latino males

PDF

Self-perceptions of Aging in the Context of Neighborhood and Their Interplay in Late-life Cognitive Health

PDF

To adapt or not to adapt: the age-specific and sex-dependent differences in the adaptive stress response

PDF

Internet communication use, psychological functioning and social connectedness at older ages

PDF

Burden of illness in hemophilia A: taking the patient’s perspective

PDF

Forensic markers of physical elder abuse in medical and community contexts: implications for criminal justice interventions

PDF

The everyday Olympics: an introduction to the many faces of Special Olympics

PDF

Urban universities' campus expansion projects in the 21st century: a case study of the University of Southern Calfornia's "Village at USC" project and its potential economic and social impacts on...

PDF

Healthcare provider recommendations: a panacea to improving influenza vaccination rates?

PDF

Assessment of challenges and strategies employed by presidents and chancellors at California community colleges responding to the Completion Agenda

PDF

Homegrown violent extremism: designing a community-based model to reduce the risk of recruitment and radicalization

Asset Metadata

Creator Gassoumis, Zachary Demetrius (author)

Core Title The economic security of an aging minority population: a profile of Latino baby boomers to inform future retirees

School Leonard Davis School of Gerontology

Degree Doctor of Philosophy

Degree Program Gerontology

Publication Date 04/24/2014

Defense Date 02/13/2014

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag baby boomers,disparities,Hispanic,Income,Latino,OAI-PMH Harvest,Wealth

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Wilber, Kathleen H. (committee chair), Nichol, Michael B. (committee member), Torres-Gil, Fernando M. (committee member)

Creator Email gassoumi@usc.edu,zachary.gassoumis@dunelm.org.uk

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-384529

Unique identifier UC11297496

Identifier etd-GassoumisZ-2410.pdf (filename),usctheses-c3-384529 (legacy record id)

Legacy Identifier etd-GassoumisZ-2410.pdf

Dmrecord 384529

Document Type Dissertation

Format application/pdf (imt)

Rights Gassoumis, Zachary Demetrius

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA