‘ - I. n
' II II I I I I I I I m I

300+

This is to certify that the
dissertation entitled

A Latent Class Analysis of Latino Political Assimilation

presented by

Gia Elise Barboza

has been accepted towards fulﬁllment
of the requirements for the

Ph.D. degree in Political Science

 

 

11W C 0L7 Mam

Major Professor’s Signature

ﬁne/M

 

Date

MSU is an Aﬁinnative Action/Equal Opportunity Institution

._A _ V‘—v--.-n-o-Q-O-l-ﬁ-o—o-n-O-O-Q-I-O-O-I-l-C-I-.-l_-_

-C--.-.--.-.-.--.-.-.-.-.--—‘—‘- ~

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p‘IClRC/DateDue.indd-p.1

A LATENT CLASS ANALYSIS OF LATINO
POLITICAL ASSIMILATION

By

Gia Elise Barboza

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Political Science

2007

ABSTRACT

A LATENT CLASS ANALYSIS OF LATINO
POLITICAL ASSIMILATION

By

Gia Elise Barboza

Latino/ Hispanic Americans as a broad category and Mexican, Puerto Rican and
Cubans in particular, are one of the newest and fastest growing ethnic populations
that are undergoing the political incorporation process. This dissertation provides an
empirical analysis of Latino political incorporation into the United States polity. It is
a response to scholars and policymakers who have asserted that immigrants threaten
the national unity of this country and contribute little, if anything, to American
political culture. To critically examine the multiple levels of political incorporation
among Latino immigrants, this research uses data from the 2002 Kaiser/ Pew Survey
of Latinos and the 2004 Kaiser/Pew Latino Survey: Political and Civic Incorporation.
One goal of this dissertation is to comprehensively evaluate the predictions of three
competing theories derived from the sociological literature on immigrant assimilation

as they are applied to patterns of Latino political participation. More specifically,

I contribute to the current literature on Latino political participation by examining
three generations of Cuban, Mexican and Puerto Rican immigrants and by evalu-
ating whether the predictions of these competing models hold to empirical scrutiny.
Overall, the degree to which context of reception, acculturation and group-based re-
source variables explain the variance in political behavior differs by national origin
and generational status and is contingent on the type of resource under considera-
tion. A latent class analysis revealed two typologies of Latino political participation,
namely individuals who are highly responsive to the electoral system and those who
are highly unresponsive. A second goal of this dissertation is to explore the contours
of American national identity among Latino subgroups in the United States. A latent
class analysis revealed an interpretable solution of Latino national identity into three
classes that can be loosely characterized as ethnocultural Americanism, multicultur—
alism and modiﬁed multiculturalism. The simultaneous latent structure of national
identity among Mexicans, Puerto Ricans and Cubans further revealed homogeneity
across each national origin subgroup, indicating a substantially similar structure be-
tween them. A multinomial latent class regression was also performed on several
covariates related to immigration status and group identity. The analysis revealed
that context of reception and group consciousness signiﬁcantly predicts membership
in each class. A major ﬁnding of this dissertation is that Latinos show overwhelm-
ing support for the contours of ethnocultural Americanism as evidenced by the large
prevalence of this class. Overall, I found a high level of support for American values
among Latinos and high levels of electoral participation an indication that Latinos

are assimilating both politically and culturally, despite popular belief to the contrary.

© 2007

Gia Elise Barboza

All Rights Reserved

This dissertation is dedicated to three individuals who struggled to
make this possible but who were taken from me too soon, my beloved
mother, Othelia Roberta Salerno, my beloved grandmother, Susan

Salerno, and my beloved friend Kimberly Rosario Perez.

ACKNOWLEDGMENTS

This dissertation would not be possible without the loving support of many people.
First and foremost is my wonderful, loving and supportive husband, I’khala, who has
always believed in me and encouraged me to succeed, even in the face of tremendous
adversity. All of whatever I do in my life, my accomplishments and achievements, my
struggles and failures, are done so that my ﬁve children, A’ssunta, Maia, Romello,
Sunﬂower-Mei and Oji, can have a better life than I had. They are instrumental in
my ability to ﬁnish this dissertation, despite the fact that so many people blame them
as being an obstacle to success rather than a facilitator of it.

During my tenure at Michigan State University, I have met many wonderful peo—
ple, all of whom must be mentioned here because without them it would have been
impossible to ﬁnish this dissertation. A special thank you for the emotional guidance,
ﬁnancial support and general encouragement of Dr. Yevonne Smith of the Graduate
School. Dr. Smith, you have encouraged me to move beyond and move past so many
times when I was on the verge of giving up. You have never once let me down. You
have always been there for me in any way I needed you to be, and you are a very
instrumental part, and in many ways the sole reason, that I am able to complete

this degree. Another incredibly supportive person has been Dr. Karen Klomparens.

vi

Thank you for being so student oriented, fair and balanced and open to alternatives.
The Graduate School has done everything possible to make this happen, and I will
never forget that.

Words cannot even begin to express the gratitude I have for my dissertation ad-
visor, Dr. Kenneth Williams. As I learned the hard way, very few individuals have
the ability to guide students in a professional, decent and respectful manner while
simultaneously imparting to them key ingredients of their future success. Ken, thank
you so much for not giving up on me and the topic I was interested in pursuing.
My very future has depended so heavily on your guidance and support, and your
genuineness is something that I will never forget despite all the difficult times I have
encountered. I would also like to express special gratitude to Dr. Ada Finifter and
Dr. Reginald Sheehan for their helpful comments on the dissertation. Undoubtedly,
I am grateful to have been privileged with such a wonderful, supportive dissertation
committee with some of the best scholars in this field.

The “heart of gold” of my good friend Dr. David Imig is something I am truly
blessed to have had the privilege to be blessed with for the past eight years. The many
discussions we have had has helped me begin to develop as a scholar and a person.
There are very few individuals in life that have compassion, gratitude and respect
for all persons as do you, and knowing that such a scholar is close by and always
accessible has been a huge motivational factor for me never giving up. Likewise, Dr.
Fransisco Villarruel has been such a good friend to me and always a positive inﬂuence.
Thank you both very much.

When I think about my best friend Rohan Williams, tears well up in my eyes. You

vii

have been and will always be so giving and you never ask for anything. Lord knows, I
have taken so much from you. You have always been there for myself and my family
and have become part of my family in the process. I truly love you like a brother and
could not have made it through this without you. All of the incidents, however much
they seemed comical and absurd, were never really funny but rather sabotaging to
both of us. This, my friend, is the price we pay for being intuitive individuals who
care and who want to make a difference and did not. merely pursue higher education
as a means to an end. To my other close friend, Steven Korzeneiwski, my deepest
gratitude. We are so much alike, and I am so grateful that you were there to reinforce
my values and beliefs during the last stages of this dissertation. We will always be
close friends, even if we are apart.

Kim, I never got a chance to thank you. You came to the political science depart-
ment for a reason and it was not to complete a degree, it was to bless my life with
yours. You left a huge imprint on my heart, and as I write this I still cannot help
Crying and wondering why you were taken from this planet so drastically too soon. I

know you were with me as I struggled to get through this. Thank you.

A huge part of my life for the past couple of years has involved statistics and the
individuals who I have idolized and have given me something to strive for. First is
Dr. Jeffrey Wooldridge who has been hugely inﬂuential in motivating me to pursue
Statistics because of his wonderful and what turns out to be a rare ability to teach it
and to make it extremely interesting. I am so blessed with the exceptional opportunity
to have had the absolute pleasure to converse with him on several occasions and to

have him be a part of this dissertation. Thank you for never saying no. As for my

viii

good friend Dr. James Stapleton I have truly cherished your friendship and your
encouragement to pursue statistics. You have made the past few years so enjoyable
regardless of whatever else was happening in my life.

I would like to express appreciation for the especially insightful comments of Dr.
Karen Kauffman of the University of Maryland whose comments at last year’s South-
ern Political Science Association truly guided me and helped me to reorganize this
thesis.

There are many more individuals that deserve my complete gratitude for always
being supportive and positive and believing in me. These people are Dr. Larry
Shiamberg, Dr. James Oehmke, Dr. Robert Griffore, Dr. Teresa Mastin, Rajesh La]
and Robin Weatherill.

My “sisters” Arwa Mohamed and Violet Rahme have been a huge part of my life
for the past two years. Arwa, thank you for always reading my papers and providing
insightful critique and comments. You both are truly the best friends a person can

ever have.

Last but not least are the two people who were most influential in my life, my
mother “farta” and my grandmother, “grannie.” Nobody can understand, as can we,
how devastated I have been since I lost you both and how much it has taken to get
me to this point from the deepest, darkest moments in my life that occurred after
you left me. Your struggles, as a single parent, as an immigrant, as individuals who
IIlade tremendous sacriﬁces for your children, have not. left this planet totally in vein.
And, while in the general scheme of things this dissertation does not matter a huge

deal, it does give at least five people a chance in life that we never had. I love you

and thank you.

By the way, I would like to thank everybody who tried their best to not make
this possible. I am truly honored to have received so much attention from peOple in
the department who wanted to destroy me and my family . thank you. What you do
not understand is that I am going to succeed with or without you. I pity you. You

know who you are.

TABLE OF CONTENTS

List of Tables xiv

List of Figures xv

1 Introduction

1.1 The Demographics of Immigration .................... 1
1.1.1 The Importance of Latino Political Incorporation to American
Politics ............................... 6
1.2 Main Argument and Research Questions ................ 9
1.2.1 Main Argument .......................... 13
1.2.2 Summary of Research Questions and Major Findings ..... 16
1.2.3 Organization of the Dissertation ................. 19
2 Newcomers in the American Polity: Theoretical Models of Inter-
group Relations Between Immigrants and Americans 22
2.1 Introduction ................................ 22
2.2 Early Models of Incorporation: Classical or “Straightline” Assimilation
Theory ................................... 23
2.3 The Ethnic Competition Perspective on Immigrant Assimilation . . . 28
2.3.1 Extensions to the Ethnic Disadvantage Perspective of Immi-
grant Assimilation ........................ 31
Theoretical Perspectives on Identity Development 35
3.1 Group Interest and Identity Politics ................... 35
3.2 Group Identity and Consciousness as a Political Group Resource . . . 42
3.3 The Process of Identiﬁcation ....................... 51
3.4 Ethnic Identity and Acculturation Theory ............... 53
3.5 Social Identity Theory .......................... 56
Data, Research Methodology and Hypotheses 59
4.1 Data .................................... 59
4.1.1 The 2002 National Survey of Latinos in America ........ 59
4.1.2 The 2004 National Survey of Latinos: Politics and Civic Life . 61
4.2 Components of the Straightline Model of Assimilation ......... 61

xi

4.3
4.4
4.5

4.6
4.7

4.2.1 Cultural Indicators of Assimilation ............... 62
4.2.2 Operational Definitions, Exogenous Predictors and Research

Hypotheses ............................ 63
The Ethnic Disadvantage Model of Assimilation ............ 81
The Multiple Pathways of Assimilation ................. 88

The Components of Political Assimilation: The Dependent Variables 88
4.5.1 Extending the Assimilation Framework to Political Outcomes 88

4.5.2 Electoral Activity ......................... 92
4.5.3 Nonelectoral Participation .................... 93
4.5.4 Citizenship ............................ 94
4.5.5 Political Culture and American National Identity ....... 94
4.5.6 Control Variables ......................... 98
An Empirical Model of Political Incorporation ............. 99
Statistical Methodology ......................... 103
4.7.1 Binary, Ordinal and Multinomial Logistic Regression ..... 103
4.7.2 Ordinal Logit Models: Proportional Odds Models ....... 108
4.7.3 Path Analysis and Structural Equation Modeling ....... 114
4.7.4 Latent Class and Logistic Latent Class Analysis ........ 115
4.7.5 Poisson and Negative Binomial Regression ........... 120

The Meaning of Americanism and National Attachment from an

Immigrants Perspective 122
5.1 Overview .................................. 122
5.2 Review of Research ............................ 123
5.2.1 American National and Political Identity ............ 129
5.3 Social Identity Theory, Symbolic Politics and Group Attachment . . . 134
5.3.1 Hypotheses ............................ 135
5.4 Results ................................... 136
5.4.1 Ethnocultural Americanism (Class I) .............. 150
5.4.2 Multiculturalism (Class II) .................... 151
5.4.3 American Liberalism (Class I) .................. 156
5.4.4 Governmental Paternalism (Class II) .............. 156
5.5 How Does Acceptance of American Liberalism change with Accultur-
ation, Exposure and Context of Reception in US Society ....... 158

5.6
5.7

5.8

5.5.1 A Multinomial Logistic Latent Class Analysis of Americanism 159
The Nature of Ethnic Identity Choices Among Immigrant Populations 165

The Distribution of Identity Choices Among Latino Immigrants . . . 165

5.7.1 A Multinomial Logistic Model of Identity Among Latino Immi-
grants ............................... 169

Conclusion: The Integration of Identity ................. 183

xii

similation Theories
6.1 Introduction ................................
6.2 Review of Research ............................
6.3 Assimilation as a multidimensional concept ...............
6.4 Results ...................................
6.5 Group—Based Resources and Latino Political Participation ......
6.5.1 Bivariate Relationships Among Group-Based Resources and
Political Participation ......................
6.6 A typology of Latino voting behavior ..................
6.7 A negative binomial model of Latino citizenship acquisition ......
6.7.1 Results ...............................
6.8 Summary and Conclusions ........................

6 A Test of the Straightline, Ethnic Disadvantage and Segmented As-

189
189
191
200
201
220

242
253
256
268

Group Solidarity, Pan-Ethnic Identity and Acculturation: A Re-

source Model of Latino Nonelectoral Political Participation 273
7.1 Introduction ................................ 273
7.2 Data and Method ............................. 275
7.3 Results ................................... 278
7.3.1 Sustaining Participation Beyond A Single Act ......... 283

7.3.2 A Continuation Ratio Logistic Model of Nonelectoral Partici-
pation Among Latinos ...................... 284
7.3.3 A Latent Class Analysis of Latino Empowerment ....... 295
7.4 Summary and Conclusions ........................ 302
Summary and Conclusion 303
8.1 Policy Implications ............................ 311
Appendices 313
Bibliography 317

xiii

List of Tables

1.1 US Residents Born Abroad by Origin and Citizenship Status, US Es-
timates, 1990 - 2000 ........................... 3

4.1 Hypothesized Effects of Explanatory Variables on Political Participation 87

xiv

5.1
5.2
5.3
5.4

6.1

6.2

6.3

6.4

6.5

6.6

6.7

List of Figures

Components of American Liberalism ..................
Distribution of American Liberalism with Covariates .........
Mode of Identiﬁcation By Age ......................
Predicted Probability of Identification given Perceptions of Group-

based discrimination Among First Generation Latinos (Foreign Born)

Predicted Probability of Voting in US Elections and Ties to Mexico by
Identiﬁcation (Mexicans Only) ......................
Predicted Probability of Voting in Mexican Elections and US Accul—
turation ..................................
Latent Class Model of Latino Voting ..................
Gamma Parameter Estimates for Latent Class Membership (Voting) .
Rho Parameter Estimates of Electoral Participation ..........
Negative Binomial and Poisson Predicted Probabilities vs. Observered
Proportion for Model Fit .........................

Negative Binomial Predicted Probabilities Based on Estimated Means

XV

174

176

214

219
243
248
249

257
258

Chapter 1

Introduction

In this dissertation, I address two broadly construed questions that have not been
adequately addressed in the literature. The ﬁrst set of questions concerns the extent
to which existing theories of assimilation explain the political incorporation of Latino
immigrants (de la Garza, 2004). A separate but related set of questions pertains
to the role that group solidarity, group consciousness and discrimination plays in

facilitating or prohibiting political incorporation.

1.1 The Demographics of Immigration

Immigrants leave their country of origin for many reasons, some because of adverse
economic conditions, and others in order to escape political or religious persecution
or simply to fulﬁll the American dream. Following the liberalization of restrictive
and discriminatory immigration laws in the 19608, immigration totals in the US have

increased during the last decades. Since then, immigrants have continued to arrive

in the United States at an increasing rate, truly making this country a “nation of
immigrants” (Wong, 2000). During the period spanning 1991 - 1998 the average an-
nual number of immigrants admitted to the United States was approximately 700,000
(Bean and Stevens, 2003). As can be seen from Table 1.1, the overwhelming major-
ity of immigrants arriving in the US. today come from either Mexico or Asia (61%)
rather than Europe (12%), in contrast to the mass migration of the early 19005 where

most migrants were of European descent (Wong, 2000).

ooom - om? .mcésﬁmm m2 £33m aEmsmEﬁO was EmEO c3 veoﬁaa Eom 352QO m2 AA wish

 

 

£02: 832...: $2: «maﬁa: use: 3mg? Lace
a Sign a 83% H mama 888

a 93.80. a 2.32 a 80.2: 8:82 seem

a Seam a 82% m. Sada 853. 3:80

w 89mg a 23% S £3: 503:8
an 8.933 em 8326 M: £38 0252
8. Rama 8. Exam 3. Sea assoc
e .833 a 232. a =ng 8:2
a 2335 mm Seams oe moose. sag
as $38; as 83%; 0a: Beam 325
Eomémwoaoh we Eomémmoaom H508 cmNEOéoZ 05 cmﬂﬂoéoz “832.8352 mks 3856 6323232 Ewco

A particularly salient aspect of this population change besides the overall level of
immigration is the changing face of immigration (Schildkraut, 2003). In 1970, only
5% of the population was Hispanic; by 1990, Latinos comprised 9% of the United
States population 1. The Census Bureau estimates that as of March 2004, approx—
imately 40 million Latinos reside in the United States, 12% of whom are foreign
born2. Population projections suggest that Latinos will comprise roughly 25% of the
total American population by 2050 and that Latinos will be a majority in a number
of states (US Census Bureau, Population Projection Division, March, 2004). The
increase in sheer numerical strength is coupled with signiﬁcant regional disparity, as
Hispanics are concentrated in the Western and Southern regions of the United States,
particularly in heavily populated states such as Texas and California (Highton and
Burris, 2002).

Driven largely by waves of these “new” immigrants, the Latino population has
soared since 1990, reaching parity with African Americans as the largest “minority”
group. Not only are contemporary immigrants growing in number, they and their
children are becoming a larger segment of the American political system. Rising
levels of immigration coupled with an image of these newcomers as “non-white” has
had major consequences for American politics. As a result, scholars and policymakers
alike have increasingly turned their attention to Latino political incorporation. For
example, mobilizing the Hispanic population to ensure their political participation has

recently played a central role in election campaigns. In the 2000 election, Republicans

 

1http: //www.jsri.msu.edu/latinospec/
2US Census, November 2004 Voters Supplement

employed several devices that were meant to entice Hispanic voters to abandon their
Democratic partisan affiliation and also to mobilize the Hispanic vote (New York
Times, 2000). Using data from the 2000 and 2004 National Annenberg Election
Surveys (NAES), Kenski and Tisinger (2006) examined the extent to which Bush
and the Republicans made gains among Hispanic Americans. They found that while
Bush improved his support among Hispanics, Hispanic party identiﬁcation in 2004 was
comparable to identiﬁcation in 2000. The four states with large Latino populations,
California, New York, Florida, and Texas, are crucial in deciding future presidential
races, and a couple of dozen congressional districts could determine which party
controls the House of Representatives. According to Henry Cisneros, president of

Univision and a former secretary of housing and urban development

In every election from now into the future. Latinos will have a decisive

say, because of the concentration of Latinos in important states.

Clearly, then, attention to immigrant incorporation into the US political system
is a critical ingredient to understanding future developments in American politics.

The research presented here focuses on the political participation and assimila—
tion of Latino immigrants in the United States who are of Mexican, Puerto Rican
and Cuban origin. Since immigrants from Latin America comprise over 50% of the
foreign born population (Wong, 2000), particular emphasis is placed on the foreign-
born/native-born dichotomy. As can be seen by Table 1.2, approximately 33% of
Latino citizens are foreign born. Comparisons between native and foreign born pop-

ulations are conducive to understanding how differences in assimilation trajectories,

including the degree to which Latinos have common goals, expectations and percep-
tions of identity, affect degrees of political incorporation.

Table 1.2: Birthplace By Ethnicity

 

Foreign born Native born Total
Hispanic 33% 67 100%
Non-Hispanic 4% 96 100%

 

Source: US. Census Bureau, Current Population
Survey (CPS), Voting Supplement (2004).

1.1.1 The Importance of Latino Political Incorporation to
American Politics

Some scholars and policymakers alike believe that immigration is subversive of the
American “way” and on that basis are opposed to it. In general, the reason given
to justify this opposition is that immigrants are a threat to the political cohesion
of this nation (de la Garza, 2004). The most prominent scholar to use this logic of
justiﬁcation is Huntington (1996, 1997, 2000), who claims that Latinos are relatively
noncommittal with respect to the “national interest” and that immigration threatens
American “cultural integrity, national identity and perhaps its very future as a coun-
try” (Huntington, 2004). According to Huntington, “this [new] wave of immigration
is profoundly different from any before it and therefore more dangerous to Ameri—
can identity” (2004, p. 34). He continues that “[m]any Mexican immigrants and their
offspring simply do not appear to identify primarily with the United States” (p. 40).
The response among the Latino community to Huntington and others with similar
viewpoints has been overwhelming. In fact, a June 2006 symposium entitled “Im-

migration and National Identity,” that appeared in the “Perspectives on Politics” is

comprised of four articles which consider different aspects of the social and political
incorporation of Latino immigrants in the US. In the introductory article, Segura

(2006, p. 278) claims that

[d]ebates about immigration and national identity cut to the core of our
national self-image as a nation of immigrants, and invariably includes
allusions to the past - real and idealized — as a way of understanding and

coping with social and demographic changes today.

The ﬁrst article, entitled “Culture Clash? Contesting Notions of American Iden-
tity and the Effects of Latin American Immigration,” examined immigration and na-
tional identity in the context of American political development (Segura and Fraga,
2006). The second article. “Mexican Americans and the American Dream” addressed
claims that “Mexicans are on their way to forming a separate nation within the US

(Alba, 2006, p. 289).” Alba argues that

In the Mexican-American case, the cross—sectional comparison of genera-
tions is misleading, as we will see below, in part. because different genera-
tions originate in different periods of Mexican immigration and settlement
and in part because the institutional discrimination of the pre—civil rights

era thwarted mobility and interfered with past intergenerational advance.

The third article, “Cuban Emigres and the American Dream,” discusses the Cuban

immigrant experience and the claim that “Latin Americans are eroding our country’s

13

core Anglo—Protestant values. Finally, the fourth article, “Mexican Immigrant Po-

litical and Economic Incorporation,” investigated the empirical basis of the claim

that American identity and culture is being undermined by a “trend toward cultural
bifurcation” driven largely by Latino immigration.

While these series of articles were written specifically to address Huntington’s
theories of Latino political and social incorporation, with limited exception propo—
nents from neither camp have provided empirical evidence to support their positions.
Only recently have scholars opined that these theoretical claims must be rigorously
tested (see e.g. Staton, Jackson and Canache, 2005). Scholars must respond to these
claims systematically and intelligibly before this type of armchair speculation begins
to reinforce an ideology of resentment and prejudice towards individuals who in many
circumstances have every right to be here. Furthermore, what characterizes a national
identity and to what degree do immigrants adopt American values? Assuming this
identity is comprised of acceptance of democratic processes (Cisneros, 2004) then it
is possible to engage in a scholarly analysis that systematically tests Huntington’s
theories.

Apart from speculation regarding the impact that immigrants have or will have
on this country, developing greater knowledge of the nature and processes that char-
acterizes Latino political incorporation is important for other reasons as well. First,
despite the fact that Latinos represent the largest minority group in the United States
and that immigration is the primary factor driving population growth, little is known
about how they relate to the political system (Bedolla, 2003). Second, given the
growing numbers of Latinos, the future of American politics depends crucially on
their incorporation into the US. political system (Wong, 2000). Most importantly,

political empowerment and adequate representation will enhance Latinos quality of

life in the United States. If the rhetoric of equality is to have any bite, Latinos must

be adequately represented and given an active voice in United States politics.

1.2 Main Argument and Research Questions

Ethnic identity refers to the component, of the self that includes both a knowledge
and evaluation of one’s membership in an ethnocultural group (Tajfel, 1981). Eth-
nocultural identity considers a person’s ethnicity as well as the extent to which he
or she actually identifies with and practices the lifestyle of that group. In groups
undergoing acculturation, there can be considerable variation in the extent of ethno-
cultural identity with a particular cultural tradition. Assimilationist thought argues
that the process of etlmocultural identification plays a central role in determining
one’s level of acculturation. Understanding the extent to which Latinos identify with
their ethnocultural heritage, then, is a crucial component of understanding the de—
gree to which they are assimilated, which in turn is important in understanding their
political attitudes and behaviors.

The experience of newly arrived immigrants provides a unique lens with which to
view the interplay between the emergence of ethnocultural identity and the political,
economic and social institutions that redeﬁne and transform those identities. Schol-
arly focus on the role of ethnic identity on the political incorporation of Latinos is
desperately needed yet “political scientists have. generally not analyzed how issues of
language and identity formation affect political attitudes (Bedolla, 2003, p. 265).” In

a recent article, Citrin et a1. (2003) stated that, in addition to the variables normally

used by political scientists, “the role of ethnic identity [in facilitating political assimi-
lation] - particularly as it surfaced among Latinos - needs further exploration” (p.
22). I argue that models applicable to Latino immigrants must include psychological
variables designed to measure ethnic and national identity because to some, they sig-
nify national integrity and commitment to American political values (Sears, et. a1,
2003).

Since immigrants must decide the degree to which they are willing to assimilate
into American society, it seems reasonable to assume that issues surrounding eth-
nocultural identiﬁcation have considerable inﬂuence for Latino political orientations.
Some researchers have suggested that the political incorporation of Latinos depends
on the particular mode of assimilation into the American polity, which in turn is
related to membership in speciﬁc interpersonal networks and the ability to use group
identity to gain access to political resources (Torres, 2001). But what role does ethnic
identity play in facilitating Latino political incorporation? Does political incorpora-
tion depend more heavily on certain kinds of identities as opposed to others? Does
heightened ethnic consciousness motivate Latinos to be more or less likely to adopt
American political ideals and/ or be politically active? In addition, how does being
conscious of ethnic identity compare across racial and ethnic categories?

The traditional model of assimilation based on the experiences of EurOpean immi-
grants posits that the values and beliefs of newly arrived immigrants will eventually
converge to that of the dominant culture (Marquez, 2004), eventually leading to the
wholesale adoption of an American identity. However this “linear trend of incorpo-

ration” (Marquez, 2004, p. 4) may not fully capture the dynamics of Latino incor-

10

poration because for them the ﬂuidity that characterizes ethnic identity is inﬂuenced
by, among other things, levels of discrimination, govermnental policies, educational
achievement and family composition (Portes and Rumbaut, 2000). More recent the-
ories of assimilation argue that as immigrants assimilate they should be more likely
to describe themselves in panethnic terms and be politically motivated by Latino-
specific causes and issues. By insisting that Latino subgroups are distinct, however,
Latino scholars have failed to acknowledge the possibility of a shared Latino political
identity, and therefore the role of panethnicity is not well developed in the literature
on Latino political incorporation (Claassen, 2003). Is the recent rise in panethnic
self-identiﬁcation reﬂective of a new Latino political identity (de la Garza, 2004)?
Similarly, does the increase in usage parallel the assimilation process such that as
immigrants are socialized into American culture they are more likely to identify in
panethnic terms? Political scientists studying African American political behavior
have shown that black consciousness contributes to the combination of political effi-
cacy and political mistrust. Does the same hold true among Latinos? A priori, there
is reason to believe that that Latino immigrants are more conscious of their ethnic
identity than are native born populations, if for no other reason than because the
inability to speak English forces some Latinos to confront their ethnic identity on a
daily basis.

Clearly, issues of identity are at the crux of both these perspectives and the ques-
tion of whether there is a shared Latino identity versus a unique identity based on
national origin has not been explored in the political science literature and hence re—

mains unanswered. In this dissertation, I address these questions by assessing the role

11

that identity plays in promoting political affiliations, including the importance placed
on political participation, and political attitudes and values. An important goal of
the dissertation is to sort out these competing hypotheses to gain deeper insight into
Latino political behavior.

To further contextualize the importance of personal resources and psychological
orientations, I focus on how the broader social and institutional environment shapes
the political behavior and attitudes that promote political incorporation (Marschall,
2004). While social context clearly plays an important role in determining an in-
dividual’s involvement in politics, existing research has not been clear about how
social networks and other contextual factors affect the level of political incorporation
among Latinos. In particular, the context of reception should shed light on why some
immigrants have successfully assimilated into the US polity while others have not.
Latino immigrants who are deﬁned as outsiders to mainstream society by natives are
viewed as inferior and this can create or reinforce discrimination (Castles and Miller,
1998). It may be that, for some Latino subgroups, the context of reception, which is
revealed in the opinions and attitudes of natives, particularly whites, promotes group
consciousness. Importantly, according to some assimilation theories (Portes, 1984),
it is the perception of discrimination and about one’s relative position in the social
hierarchy that matters the most. Do positive perceptions about life in the United
States relative to one’s country of origin lead to an outright acceptance of Ameri-
can political values and attitudes? Alternatively, does the awareness of institutional
barriers in the host country awaken one’s sense of group consciousness and motivate

political behavior?

12

1 .2.1 Main Argument

Differences in how immigrants become involved in the political system can be at—
tributed to the context of reception and transformations in the way individuals come
to View themselves as contact with American culture increases. Although Mexicans,
Puerto Ricans and Cubans have distinct cultural and historical experiences they have
had similar experiences upon coming to the United States. For Mexican-origin pop-
ulations, the long history of economic, legal, and educational discrimination has left
them disproportionately overrepresented in lower socioeconomic positions in Ameri-
can society (Agguire, 2001). Agguire (2001), for example, argues that negative beliefs
about Mexican Americans that they are prone to criminality, are lazy, un—American,
and gang oriented, for example have legitimized discrimination against them. The
distinct cultural characteristics of Puerto Ricans have similarly tended to reinforce
an ideology of prejudice and resentment. Because the island of Puerto Rico has been
a US colony since the late 1800s, the cultural differences between Puerto Ricans and
whites are less critical than perceived biological differences (Agguire, 2001). Amer-
icans hold many negative stereotypes about Puerto Ricans. They are viewed as a
drain on the welfare system and social services, and are considered lazy, submis-
sive and immoral (Wagenheim, 1973; Lopez, 1973; Diaz-Cotto, 1996; Agguire, 2001).
American perceptions of Cubans when they ﬁrst began arriving to the United States,
however, were quite different due to their anti—Castro and anticommunist sentiment.
Cubans were viewed more as allies in a common cause than as territorial minority

groups (Agguire, 2001). As Cubans began to assimilate into the US, they were in—

13

creasingly seen as industrious, intelligent and law—abiding. The Mariel boatlift3 in
1980, however, dramatically changed the perception of Cubans and hostility towards
them has escalated, creating latent white and manifest African American resentment
towards them. Exemplary of this treatment is the approval in 1980 in Dade County
of an English-only ordinance that reversed the policy of official bilingualism estab—
lished in 1973“. Clearly, initial positive perceptions of Cubans has given way to
more negative beliefs based on their language and cultural roots and their success in

7

transforming south Florida into a Latin American “haven.’ Given these and other
historical circumstances, my argument is that the context of discrimination is in large
part responsible for an awakened group consciousness and the politicization of ethnic
identity.

An important component of ethnic identity is one’s relationship to their home
country and the extent to which immigrants become re—socialized into the American

polity. In their discussion of Australian migrants, (Finifter and Finifter, 1989, p. 629)

eloquently summarize this learning process:

In sum, substantial new political learning, based on migrants’ current

social experiences, does take place. However, it is quite clear that this

 

3T he Mariel boatlift refers to the mass movement of Cubans who departed from
Cuba’s Mariel Harbor to the United States between April 15 and October 31, 1980.
The “wave” of Cuban immigrants who came during this period are widely known
to be very different than previous waves of Cuban immigrants. In contrast to their
predecessors, immigrants who arrived in 1980 were considered to be “misﬁts” by
the Cuban government, because, for example, they were either sexual deviants or
criminals.

4This ordinance prohibited any county expenditure for the purpose of utilizing any
language other than English, or promoting any culture other than that of the United
States.

14

new learning takes place most readily among those predisposed by prior
attitudes and behavior to continue their participant patterns. Prior so-
cialization provides the seed for new learning, but the current environment

provides the culture in which it is nurtured.

Clearly, the permanency of intercultural contact matters as well for political in—
corporation because if an individual is not “psychologically committed” to residing
in the United States, he or she will not be motivated to become a citizen, acquire
partisan affiliation or otherwise participate in politics (F inifter and Finifter, 1989).
An additional question is whether Latino political attitudes more reﬂective of Amer-
ican attitudes or are they more representative of the values and interests of their
country of origin? Relatedly, are immigrants who are politically active in their home
country more or less active in the United States, and if so what factors explain these
differences? Is it possible, as Jones-Correa (1998) suggests, for immigrants, particu-
larly non-citizens, to participate in American politics, even as they remain “officially”
outside the polity as non-citizens? F inifter and Finifter (1989) argue that partisan—
ship and overall ideology are important factors in helping migrants adapt to new
environments, and that new political learning is generally dependent on previously
established political attitudes. According to classical assimilation theory, the fre—
quency of the contact between the host country’s institutions and Latino immigrants
should provide sufﬁcient exposure to result in their acceptance of American political
norms (Citrin, 2003). Succinctly put, differences in assimilation trajectories create

distinct political identities that are 1) derivative of the socio—political environment;

and 2) depend on the perception of shared political goals and expectations.

1.2.2 Summary of Research Questions and Major Findings

The agents of socialization, their content and how they structure Latino political
identity are important to understanding whether the traditional “indicators” of in-
corporation will be forthcoming and to understand why the path to Latino political
incorporation has looked so different from that of other groups. Two predominant
theories of immigrant acculturation are considered here. The ﬁrst is classical or
straightline assimilation theory. The second is the ethnic disadvantage model of as-
similation and its close cousin segmented assimilation theory.

The main difference between these theories lies with the importance of group
consciousness and ethnicity in motivating individual political behavior. Group con-
sciousness is a well-developed literature in political science as it pertains to blacks
but not as it pertains to Latinos and certainly not Latino immigrants. Crucial to the
notion of a Latino group consciousness, is the idea that the political system should
be responsive to Latino commonality.

Panethnicity derives from membership in a Latino/Hispanic group. Because of
substantial differences among subgroups of Latinos, this notion has been hotly con-
tested by Latino scholars, who argue that they are too distinct to be thought of as
a single ethnic group. Recently, this is beginning to change, and some are acknowl-
edging that panethnic identity is situation speciﬁc but could indeed arise in some

instances. Certainly, the presence of discrimination is one of these circumstances

16

(Portes, 1984; Portes and Zhou, 1993). The recent demonstrations in Washington,
Chicago, Los Angeles and other cities reﬂect a fear among Latinos that restrictive
immigration laws will be implemented, and this has motivated political activity. The

,

demonstrations also suggest that Latinos feel they are “in this together,’ so to speak.
Identity plays a role in this because Latinos have no identiﬁcation with the constructs
Hispanic or Latino before coming to the US; they are much more prone to identify as
members of their country of origin. One contribution to existing theories of political
assimilation is made by showing that the processes that compel Latinos to identify
with the American polity are best viewed within the lens of social identity theory.
Using quantitative research methods involving analysis of survey data, I ﬁnd that
at the individual level, acculturation is a very powerful predictor of Latino political
behavior, especially among immigrant populations. That is, one of the most impor-
tant determinants of political participation among immigrants today relates to their
socioeconomic status and level of acculturation into American society. In addition,
politicization of identity appears to result from exposure to American culture and
identiﬁcation as an American seems to be a natural progression among contempo-
rary immigrants from Latin America. I also ﬁnd that group based resources explain
the variance in political behavior over and above traditional resources based on so-
cioeconomic status alone. More speciﬁcally, to some extent, perceptions of group
consciousness, identity and discrimination all play an important role in explaining
variation in political participation among Mexican, Cuban and Puerto Rican immi-
grants. For example, I ﬁnd that Cubans who identify as “American” are signiﬁcantly

more likely to participate in electoral politics, but Puerto Ricans are motivated more

17

by their level of income, education and economic attachment. Additionally, identify-
ing in panethnic terms tends to suppress voting behavior among Cubans much more
so than any other group. Perhaps the most consistent and widespread ﬁnding in this
dissertation is the role of discrimination in motivating political activity. Perceived
discrimination was found to be a signiﬁcant predictor across all types of political
activities presented here, and with limited exceptions tended to hold true across sub-
groups of Latinos as well. For example, perceptions of group based discrimination
diminish responses favorable to ethnocultural Americanism and lessen the probability
of identifying as an American versus as a member of one’s country of origin or as a
member of a panethnic group. Somewhat paradoxically, discrimination is positively
related to some types of electoral activity, as I discuss further in Chapter 6.

A comparison of Mexican, Cuban and Puerto Rican immigrants allows for the
assessment of the interrelationship between identity and political attitudes and be-
havior. Therefore, I pay particular attention to how well these models compare across
Latino racial and ethnic categories. Very few studies have explored the heterogene-
ity of Latinos by studying the variation in Latino political participation by subgroup
(Martinez, 2005). This is surprising given vast differences in demographic proﬁles and
settlement patterns among Cubans, Mexicans and Puerto Ricans. Even when ethnic

differences have been accounted for, full comparisons have generally been omitted.5

 

5The general method of comparison that has been used by others is to allow for
different intercepts for Mexicans, Cubans and Puerto Ricans, with one omitted cat-
egory (usually non-Hispanic white or “other Latino”). Each group is then compared
with the baseline for a total of two comparisons; however this approach excludes all
possible (3) = 3 comparisons. Alternatively, models oftentimes include three ethnic
“dummies” with the baseline being all other Latinos. This, however, also negates full

18

By comparing subgroups of Latinos instead of analyzing them as a common group, I
am able to identify differences in assimilation trajectories and to assess the role that

group based resources play in motivating political behavior.

1.2.3 Organization of the Dissertation

The next two chapters present several theoretical models of immigrant acculturation,
focusing speciﬁcally on classical assimilation theories by Gordon (1964) and Fuchs
(1990), and later revisions of these theories such as the ethnic disadvantage and seg-
mented assimilation perspectives. The theories represent the most recent advance-
ment of assimilation theory proposed by sociologists. The common thread among
them is that assimilation has generally not occurred among Latino immigrants, even
in the third generation. In contrast to classical assimilation theories, proponents of
these latter models argue that patterns of assimilation do not necessarily follow a
straight line, but rather for some immigrants assimilation depends on the mode of
reception into the host country. In addition, while color-based racism and religious
persecution may exist, discrimination based on race may not occur in an immigrant’s
home country ,_ it is largely a US phenomenon ~— and therefore immigrant awareness
of discrimination based on their racial distinctiveness alone takes time. This type
of discrimination may cause an additional assimilatory path among immigrants that
essentially parallels that of the minority underclass in the US. The expectations de-

rived from these theories are presented in Chapter 4, along with the methods and

 

comparison between Mexicans, Cubans and Puerto Ricans and muddles any substan-
tive interpretation. Whenever possible, I have avoided these types of analyses.

19

 

 

data used in this analysis. In addition to traditional methods employed by political
scientists to analyze attitudes and behaviors, I use latent class analysis, structural
equation modeling and other innovative methodological techniques in order to ex—
amine Latinos’ latent propensity to be politically engaged in the American polity.
As described further in Chapter 4, latent variable and simultaneous latent structure
models incorporate measurement error making their use particularly warranted and
preferable over logistic regression models. An additional beneﬁt is that these mod-
els allow me to statistically test for homogeneity across subgroups. Chapter 5 sets
the stage for the dissertation by arguing that ethnicity is multidimensional and that
in particular contextual determinants are responsible for the wholesale transforma-
tion of ethnic identity. This chapter also explores the meaning of Americanism and
national identity from an immigrants own perspective. The results of this chapter
suggest that Latinos accept ethnocultural Americanism and have strong attachments
to the American polity as evidenced by their preference towards American identi-
ﬁcation over time. I turn to electoral participation including voting, registering to
vote and naturalization in Chapter 6. The commonalities and distinctions between
Mexican, Puerto Rican and Cuban immigrants are examined. In addition to showing
how socioeconomic status and time related processes, such as length of residence and
language skills affect immigrants’ participation in electoral politics, I Show how group
based resources such as panethnicity, the perception of linked fate, shared political
goals and discrimination (i.e., the components of ethnicity discussed in Chapter 3)
impact it as well. Because a large portion of immigrants are non-citizens, nonelectoral

participation is an important indicator of political incorporation. Therefore, chapter

20

7 analyzes nonelectoral forms of participation including contacting an elected ofﬁcial,
contributing money and protesting behavior. I address an important question that
to this date has remained unexamined by students of Latino politics, namely “What
factors sustain political participation among immigrant groups after its inception?”
Another question addressed in this chapter is the role that being conscious of one’s
ethnicity plays in motivating nonelectoral participation. As is further discussed in this
chapter, a latent class analysis shows that participation clusters along three dimen-
sions, one group of individuals that will participate irrespective of feelings of ethnic
empowerment, a second group that feels highly empowered but does not participate,
and a ﬁnal group that is highly empowered and highly participatory. This chapter
discusses the implications of these ﬁndings. Finally, Chapter 8 discusses the policy

implications of this study and provides directions for future research.

21

Chapter 2

Newcomers in the American
Polity: Theoretical Models of
Intergroup Relations Between

Immigrants and Americans

2. 1 Introduction

Due to increasing levels of immigration from Latin America, Latino communities in
the United States are undergoing a constant state of ﬂux (Kaufman, 2003; DeSipio,
1996). The acculturation process that foreign born Latinos go through is central to
understanding the nature and extent of their interaction with the American political

system. Accordingly, I discuss the three predominating theories’ predictions of the

22

likely outcome of Latino political acculturation1 below.

2.2 Early Models of Incorporation: Classical or

“Straightline” Assimilation Theory

The conceptual framework of assimilation theory as a model of immigrant experiences
with American culture was ﬁrst proposed by Park and Burgess (1924) and expanded
upon by Warner and Srole (1945) and Gordon (1964). Park and Burgess (1924, p.

735) viewed assimilation as

a process of interpretation and fusion in which persons and groups acquire
the memories, sentiments, and attitudes of other groups, and, by sharing
their experience and history, are incorporated with them into a common

cultural life.

Park proposed a multistage model to account for the processes that lead to im-
migrant assimilation into American society. Park’s (1916) assimilation theory was
couched in a larger ecological framework for analyzing urban areas which emphasized
that living patterns in urban areas are produced by competition for scarce resources
such as land, housing and jobs. As populations migrate to urban areas, they accel-
erate the level of competition for resources with current occupants. The ﬁrst stage,

which Park termed the competitive phase, was viewed as an outgrowth of this con-

 

1Acculturation is thought to occur when groups of individuals having different
cultures come into proximate contact with one another resulting in a transformation
of either group (Umaﬁa—Taylor, 2003)

23

tact among diverse ethnic groups. In this phase, ethnic populations competed over
resources, such as jobs, living space and political representation. Park believed that
intergroup competition sets in motion accommodation and eventually assimilation.
The accommodation stage was characterized by change and adaptation to a new
environment. During this phase, relations between immigrant groups and natives
stabilize, even if immigrants are forced to occupy low socioeconomic positions in so-
ciety. Despite the fact that ethnic stratiﬁcation impedes the pace of assimilation, he
believed that all immigrants, even those of lower socioeconomic status, could eventu-
ally enter the last stage and become fully assimilated. The assimilatory phase, then,
was characterized by the wholesale transformation of ethnic culture.

Building upon this notion of linear progression, Warner and Srole (1945) devel-
oped the idea of “straightline assimilation,” which proposed that the foreign-born and
their offspring must acculturate and seek acceptance among native—born Americans
as a prerequisite for social, economic and political advancement. As the name implies,
straightline assimilation theory views the process of assimilation as “linear” and irre—
versible. Exemplary of the sequential nature of the assimilation process, straightline
theory posits that political incorporation follows structural assimilation, which in turn
is achievable only after an immigrant has been successful in culturally assimilating.
By irreversible, proponents of the theory meant that once assimilation occurred, it
was permanent. These researchers believed that assimilation was inevitable because
the forces of social mobility and egalitarianism act together to dissolve distinctions be-
tween ethnic groups. Part of the assimilation process, according to Warner and Srole,

was that immigrants would unlearn their cultural traits and become re—socialized into

24

the American way of life.

Gordon (1964) extended the work of Parks and Burgess (1924) and Warner and
Srole (1945), to more explicitly delineate the role of the host society and culture
to which the migrant group must adapt. Gordon was one of the ﬁrst scholars to
suggest that in order to become fully assimilated into American society, minority
groups must navigate through a series of stages that he believed occurred in pro—
gression. He identiﬁed seven assimilation dimensions, including cultural, structural,
marital, identiﬁcational, attitudinal receptional, behavior receptional, and civic as-
similation. According to Gordon, cultural assimilation occurred when the values,
beliefs, ideology, language and other symbols of the dominant culture were adopted
by migrants. Structural assimilation occurred when migrant ethnic groups inhabit
primary groups existing in the predominant culture via educational and economic
mobility (Martinez, 2005). Structural assimilation was a key phase in the assimila-
tion process because once it occurred other types of assimilation were guaranteed to
follow. Gordon viewed structural assimilation as being difﬁcult to achieve because it
required increasing levels of contact with dominant groups. In addition, once immi—
grants were able to achieve economic and educational mobility, they were expected
to be virtually indistinguishable from the dominant group(s) in society. The next as-
similatory phases, identiﬁcational and marital assimilation, were viewed as dependent
on the level of cultural and structural assimilation that immigrant groups were able

to achieve. Marital assimilation was deﬁned as high rates of intermarriage between

25

immigrants and the dominant group”. On the other hand, identiﬁcational assimila—
tion was believed to have occurred when individuals no longer viewed themselves as
members of their own ethnic group, their personal identities having become imbued
with participation and success in mainstream institutions. The lack of prejudicial
attitudes and stereotyping on the part of the dominant group demarcated the begin-
ning of the next phase, namely attitudinal receptional assimilation. Closely related
was what Gordon described as behavioral receptional assimilation, or the absence
of intentional discrimination by dominant groups against subordinate groups. Only
after traversing these stages of assimilation could immigrants be expected to par-
ticipate fully in the United States political system. This last stage, which Gordon
termed civic assimilation, was characterized by the reduction of conﬂict between mi-
grants and dominant groups with respect to their political attitudes and behaviors.
A deﬁning characteristic of this model is that over time, immigrants were presumed
to possess enough resources to overcome structural barriers that tend to prohibit full
participation in American life.

The process of ethnic identiﬁcation is an important component of classical as—
similation theories. The “ethnic enclosure” model, derived from the work of Gordon
(1964), describes how immigrants come to shed their ethnic identity in favor of Ameri-
canization. According to this model, an immigrant’s ethnic identity is derivative of the
shared cultural heritage brought to the host society and is maintained as long as they

remain physically and socially isolated from outsiders (Hwang and Murdock, 1991).

 

2In this dissertation, I was unable to assess the degree to which marital assimilation
affects Latino political incorporation because none of the datasets included a variable
measuring intermarriage. It is certainly a fruitful area for future research.

26

As immigrants begin to assimilate into the larger society, meaning they learn the
host’s language, achieve higher socioeconomic standing and interact more frequently
with natives, ethnic identity begins to erode. The ethnic enclosure thesis maintains
that lack of opportunities for intergroup contacts and socioeconomic advancement
sustains ethnic identity whereas structural integration results in the relinquishment
of ethnic identity in favor of assimilation. Clearly, the transformation of ethnocul-
tural identity is the catalyst that motivates immigrants to assimilate. Overall, this
model assumes that assimilation culminates in a fully “Americanized” immigrant who
identiﬁes with American political attitudes and adopts American political behaviors.

The theoretical insights of classical assimilation were based on pre—1965 European
immigrants who are much less diverse and not. at all similar to America’s new im-
migrants. For example, pre-1965 immigrants earned slightly more than natives but
post-1965 immigrants earn, on average, 16% less than natives (Brimelow, 1995). In
addition, post-1965 immigrants have darker skin color which makes assimilation more
difﬁcult. On this basis, sociologist Portes (1984) suggested that scholars re-evaluate
the applicability of straightline assimilation theory as it applies to post-1965 immi-
grants. Noting that the assimilation trajectories of these immigrants have generally
not been linear, scholars of late have proposed modiﬁcations to this model (Alba
and Nee, 1997; Gans, 1992; Portes and Bach, 1985; Portes, 1984; Portes, Parker and
Cobas, 1980). Consequently, two particularly salient models have emerged to explain
modes of immigrant assimilation, the ethnic competition model and the segmented

assimilation model.

27

2.3 The Ethnic Competition Perspective on Immi-

grant Assimilation

In reaction to the “melting pot” assumption of classical assimilation theories, scholars
began to emphasize the process of maintaining “patterns of ethnicity” (Portes, 1984).
Opponents of classical assimilation theories maintain that ethnic groups preserve their
unique cultural characteristics as a way of providing support and guidance in coping
with the consequences of discrimination (Aguirre and Turner, 2001). Researchers
began to realize that ethnic groups not only selectively retain their cultural charac-
teristics, but they often construct and create new ways of adjusting to discrimination
(Aguirre, 2003). Consequently, proponents of the ethnic disadvantage model have
recently placed increasing emphasis on the role of ethnicity in understanding the be-
havior of ethnic groups. Nevertheless, they acknowledge that some assimilation into
dominant society is inevitable. Scholars Nathan Glazer and Daniel Moynihan (1970),
for example, were among the ﬁrst to emphasize that partially assimilated immigrants
continue to reveal residential, behavioral, organizational, and cultural patterns that
mark their distinct ethnic identity.

The ethnic disadvantage perspective that derives from Glazer and Moynihan
(1970), Greeley (1971) and more recently Portes and his collaborators (Portes 1984;
Portes and Bach, 1985; Portes, Parker and Cobas, 1980) was an attempt to address
the limitations of classical assimilation theory. As noted by Portes in 1984, immi-
grants and their descendants have generally not “melted” into the mainstream which

suggests that immigrants may prefer to remain distinct. In particular, proponents of

28

this model note that structural assimilation has not followed cultural assimilation as
a matter of course, even in the absence of discriminatory and institutional barriers
(Portes, 1984). He argued that socioeconomic opportunities for the ﬁrst generation
are evaluated relative to those in their country of origin, and as a result second and
third generation immigrants are more likely to believe that assimilation is more dif-
ﬁcult and will take longer than they previously believed. This realization, in part, is
responsible for the reconstitution of ethnic culture (Portes and Bach, 1985).

A key question is not only why immigrants choose to retain their own cultures but
by what process does this occur? The answer is provided from newly revised theories
of ethnic competition. In contrast to Park’s conceptualization of ethnic competition,
recent theories stress that competition for resources often escalates conﬂict between
ethnic subpopulations, forcing subordinate ethnic groups into segregated housing and
limiting economic opportunities. This process reinforces the distinctiveness of one’s
ethnic group and makes the group a likely target of discrimination. The ethnic compe-
tition perspective suggests that competition with other groups heightens awareness of
racial and ethnic differences which form the basis for participation and mobilization.
These theories emphasize the relative size of the ethnic subpopulations, their patterns
of migration, their movement into various social niches, and their competition with
other ethnic groups for economic and social status.

The ethnic disadvantage perspective views ethnic identity as a dormant political
consciousness aroused among minorities as they confront prejudice and discrimination
from the dominant society. These scholars assert that because racial discrimination

in their home country is virtually nonexistent, immigrants" sense of ethnic awareness

29

is derivative of structural barriers in the host society rather than being residual of
adverse conditions in their country of origin. This perspective entails the expectation
that awareness of discriminatory conditions develops over time as minorities’ knowl-
edge of the customs and practices of the host country increase and as interaction with
natives becomes more frequent. Portes, Parker and Cobas prediction was that “the
greater the level of cultural preparedness and the higher the relative socioeconomic
standing of immigrants in the US, the greater the perception and experiences of dis-
crimination and the more critical their overall assessment of the host society (Portes,
Parker and Cobas, 1980, p. 205).” In sum, the ethnic competition thesis asserts
that greater intergroup contact intensiﬁes minorities’ sense of ethnic identity that is
requisite for political action. The main theoretical insight is that greater familiarity
and economic success allows immigrants to gain a realistic understanding of inequal-
ity and discrimination against them as they compete against natives for resources
(Aguirre et al., 1989).

Proponents of ethnic disadvantage place ethnic resilience, deﬁned by Portes (1984)
as the racial and cultural differences and the solidarity on which they are based, as
a central feature in the experience of most immigrant groups. The persistence and
salience of ethnic perceptions is an important component in his deﬁnition of ethnic
resilience. As a result, ethnic identiﬁcation, ethnic consciousness and solidarity have
become major research topics in the last two decades (Olzak, 1983; Alba and Chamlin,
1983; Portes et al., 1980; Portes, 1984). The notion of “group interests” is central to
the conceptualization of the ethnic disadvantage perspective of intergroup relations.

The re—emergence of ethnic identity stemming from discriminatory conditions implies

30

that Latinos have unique politically relevant characteristics and /or special interests

to which a representative could or should respond (Sapiro, 1981).

2.3.1 Extensions to the Ethnic Disadvantage Perspective of
Immigrant Assimilation

Current research has been fairly contentious with respect to the pathways of assimila-
tion among post-1965 immigrants (Xie and Greenman, 2001). While many researchers
have argued that the changing face of immigration has set these “new” immigrant
groups apart from their European counterparts, others have been more optimistic
about the prospect of their gradual assimilation into the American mainstream. Re-
cently, a compromise between these two competing viewpoints has emerged as the
predominant theoretical paradigm, coined by Portes and Zhou (1993) as the theory
of segmented assimilation.

Their segmented assimilation model of immigrant assimilation is an outgrowth
of work on second generation immigrants that found that racial discrimination and
a restructuring economy created distinct paths of assimilation for newer, nonwhite
immigrant populations and tended to decouple acculturation and economic mobil—
ity. Segmented assimilation theory extended the work of Gans (1992), who outlined
several distinct trajectories that second generation immigrants can follow. These
trajectories are a consequence of stratiﬁcation and inequality in American social in-
stitutions. Consequently, it is difficult for many immigrant. groups to assimilate into

L

certain ‘segments” of society. In one of the most well cited exemplars of segmented

31

assimilation, Portes and Zhou (1993) wrote

Instead of a relatively uniform mainstream whose mores and prejudices
dictate a common path of integration, we observe today several distinct
forms of adaptation. One of them replicates the time honored portrayal of
growing acculturation and parallel integration into the white middle-class;
a second leads straight in the opposite direction to permanent poverty and
assimilation into the underclass; still a third associates rapid economic
advancement with deliberate preservation of the immigrant community’s

values and tight solidarity (1993, p. 82).

In discussing post-1965 immigrants, Portes and Zhou (1993) noted that the process
of assimilation oscillates between smooth acceptance and confrontation depending on
the context of reception into American society. Their theory was a response to the
growing recognition among scholars that assimilation into American society was not
occurring even among third generation immigrants. Portes and Zhou (1993) sug-
gested that earlier assimilation theories are not invalidated by the apparently uneven
patterns of assimilation among immigrants; rather, for some, the process of assimi-
lation is better characterized as “bumpy” as opposed to “straightline”. Examining
the experiences of more recent and diverse (post-1965) immigrant groups, Portes and
Zhou (1993) argued that members of different immigrant groups may follow different
paths and participate in different segments of society, rather than moving toward the
same kind of participation in American culture. They suspected that there are multi-

ple pathways to incorporation that depend on national origin, socio—economic status,

32

the context of reception in the United States, and family resources, both social and
ﬁnancial. In particular, they argued that poor outcomes are a function of contextual
factors such as racial discrimination, urban subcultures, and labor market opportuni-
ties, and that these factors are obstacles to complete social, academic, and economic
assimilation. These various factors inﬂuence the social context that is primarily re—
sponsible for determining the mode of assimilation. The question for them was, “into
what sector of American society a particular immigrant group assimilates?” (Portes
and Zhou, 1983, p. 82). In their attempt to answer this question, they proposed
three trajectories of immigrant assimilation. The ﬁrst was that of the straightline
model proposed many years before. The second closely paralleled the earlier ethnic
disadvantage model and but was renamed selective assimilation. The third trajectory
was called segmented assimilation3.

The main theoretical contribution of segmented assimilation was the recognition
that some immigrants acculturate into a minority underclass as opposed to the main-
stream. Accordingly, the context of reception that an immigrant ﬁnds upon arrival in
this country is crucial to understanding his or her mode of incorporation. The modes
of incorporation depend on the perceptions of immigrants regarding the context in
which they are received as welcoming, indifferent or hostile as well as community
characteristics such as size and diversiﬁcation of the occupational structure. The
communities that receive present day immigrants may discourage school participa-

tion, which harms an immigrant’s chance at upward mobility. This places immigrants

 

3Confusingly, segmented assimilation is the name of the overall theory and also a
name of a particular mode of incorporation.

33

at risk of acculturating into an oppositional culture found among those in their sur-
roundings. If immigrants assimilate too fully into the surrounding environment, they
may experience dissonant acculturation and lose access to the cultural resources of
the ethnic community. Under these circumstances, the segmented assimilation frame-
work asserts that maintaining the culture of origin has a protective effect (Xie and
Greenman, 2001).

As noted by Citrin et al. (2003), immigrants who cannot assimilate to traditional
mainstream culture will contribute to the development of a minority underclass and
will begin to view themselves in racialized terms. According to these researchers,
stagnant or “downward mobility” will result even as fellow immigrants are able to
follow the straight line model towards assimilation into mainstream society. Overall,
at least three factors appear necessary for the segmented model of socioeconomic
success but continued ethnic group attachment: 1) high volume of migration from a
given area; 2) sustained flows of large numbers over time; and 3) ethnic residential
concentration“. By emphasizing downward mobility into an American underclass, the
segmented assimilation model assumes low levels of community resources as well as
a highly racialized population (Miles, 1989) with structural barriers curbing the life

chances of groups differentiated from the majority on the basis of cultural differences”

 

4“In this dissertation, I am only able to test the aspect of this model that pertains
to ethnic residential concentration.

5It is interesting to note that all three assimilation theories contain assumptions
that are questionable. For example, the straightline model assumes that convergence
towards Anglos is desirable for immigrant populations. On the other hand, the seg-
mented model assumes that interacting with African Americans leads to “stagnation”
and “downward assimilation.”

34

Chapter 3

Theoretical Perspectives on

Identity Development

3.1 Group Interest and Identity Politics

Two questions arise when discussing Latino group interests. The ﬁrst is whether the
category of “Latino” even exists (Dominguez, 1994) and the second is whether, if
it does, Latinos share common attitudes and values? The heterogeneity of Latino
subgroups has resulted in the failure of political scientists to incorporate traditional
pluralist models in theorizing about Latino political behavior (de la Garza, 2004).
These scholars have argued that distinct identity groups such Mexicans, Puerto Ri-
cans and Cubans occupy objectively different positions in racialized socio—political
hierarchies, have appreciably different life experiences and life chances, and therefore
develop unique political attitudes and interests. Apart from diverse historical circum-

stances, they point to the fact that Mexicans and Cubans more than Puerto Ricans

35

have made the United States a permanent borne, Mexicans have the highest propor-
tion born in the United States and the highest level of bilingualism and that Cubans
have more readily adapted to American life (Calderon, 1992). In addition, the term
Latino does not apply to a group with common biological descent and not all Latinos
even speak the same language (Fox, 1996). On this basis, scholars have insisted that
Latino subgroups are too heterogeneous to be analyzed in terms of membership in
a panethnic group. They argue that group heterogeneity precludes the possibility of
identifying public policies that promote the interests of Latinos as a panethnic group.

Nevertheless, repeated contact with American institutions has resulted in com-
monalities that compel Latinos to unite for collective action purposes (Calderon,
1992) despite signiﬁcant differences. Of late, students of Latino politics have begun
to take note and are beginning to pay attention to the possibility of shared politi-
cal identities. Claassen (2003), for example, argued that the most effective way of
studying Latinos is to undifferentiate them in order to understand the issues that are
important to them as a group. Only then will we see their inclusion in democratic
processes - in effect, the category of “Latino” does exist and has meaning.

The increasing tendency to believe that minority or ethnic groups possess sub-
stantive interests that ought to be represented (Thernstrom, 1989) coupled with the
fact that Mexicans, Cubans and Puerto Ricans have been the target of substantial
discriminatory practices, will fundamentally change the way scholars theorize about
Latino politics. The utility of a pluralistic framework (de la Garza, 2004) to char-
acterize Latino political life is contingent on the highly contested assumption that

Mexicans, Puerto Ricans and Cubans share particular social, economic, or political

36

problems that do not closely match those of other groups and also that they share a
particular viewpoint on solutions to political problems (Sapiro, 1981, p. 703). This
panethnic identity, which arose from external forces, such as the media, US. Census
Bureau, and politicians, rather than from an internal sense of intergroup cohesion
(Moore and Pachon, 1985; Munoz, 1999), may be responsible for the tendency of
some groups to unite around what is perceived to be a common cause. For exam-
ple, community leaders in Chicago have popularized the term Latino to represent the
collective concerns of the Spanish-speaking population in response “to common struc-
tural conditions in the areas of education, politics and economics” (Padilla, 1985, p.
40). According to Padilla (1985), Mexicans and Puerto Ricans in Chicago have a

history of collective mobilization based on a common Latino panethnic identity.

37

Amoomv 6::0002 can 52:90
4 23 beEEDm ooom was 23 mash. >..:~EE:m ommﬁ ,semesm .3980 .m.D “meadow

 

 

 

2: 2: we @333: masons...“ 333$ 58.

mm mm a 83m? 3.38.2 mmeswos oeaaam 5&0

a m 2 SEQ 8932 games; sense

2 2 mm «mama $39.5 5.53 52m 055

as 8 mm ESE $53.8 $33.2 case:

88 82 $55 ea casszz 88 as: assesses
GOSDQEQEQ acoouonm Doomnommﬁ nomadso Hon—552

 

Sea Ea 82 .mssm Bee: 2: s eosaasaoa smegma 2: a eoasaoeessaam S was.

38

The question of whether a common political and cultural collectivity exists among
Latinos is perhaps better answered by Latinos themselves. As can be seen from
Tables 3.1 and 3.2, a large proportion of the US population reported ‘other Hispanic’
origins in the 2000 decennial census (Guzman and McConnell, 2002). The table
demonstrates that more than 10 million individuals who self-identify as Hispanic
did not choose the Mexican, Puerto Rican or Cuban categories. Furthermore, of
all Latinos who chose the ‘other Spanish/Latino/Hispanic’ category 40% wrote in
the term Hispanic and an additional 19% wrote in the term ‘Spanish’ or ‘Latino’
(Guzman and McConnell, 2002). Attempting to explain this phenomenon, Guzman
and McConnell (2002) argued that Latino self-identiﬁcation may have changed over
the decade spanning 1990-2000. , As further evidence, they point to differences in 2000
versus 1990, when the ﬁrst large scale Latino political survey was conducted by de la
Garza and his associates. The Latino National Political Survey (“LN PS”) found that
Latinos preferred to identify with their country of origin over opting for a panethnic
identity. By 2000, however, this has changed, and more Latinos preferred to categorize
themselves in panethnic terms. Guzman and McConnell suggest that this preference
is further supported by the fact that 91% of the respondents who chose the ‘other’
category did not identify with a particular country, preferring instead to write in the
terms ‘Hispanic’, ‘Spanish’, ‘Latino’, or checking the box for ‘other Hispanic’ without
entering a more speciﬁc response. Clearly, as evidenced by those who ﬁlled in that
they would prefer a panethnic identity on a survey question, panethnic identiﬁcation
is a particularly salient form of identiﬁcation among some Latino groups. It reﬂects a

growing trend among Latinos to believe that they have a common interest and that

39

they are in fact a political community, albeit an imagined one (Anderson, 1997).

Table 3.2. Breakdown of “All other Hispanic” response by type, 2000

 

 

 

2000
Number Percent
All other Hispanic or Latino 6,111,665 61
Checkbox only, other Hispanic 1,733,274 17
Write-in Spanish 686,004 7
Write—in Hispanic 1,454,529 25
Write-in Latino 450,769 5
Not elsewhere classiﬁed 787,089 8
Total 10,017,244 100

 

Source: US. Census Bureau, 1990 Summary Tape File
and 2000 Summary File 1, Guzman and McConnel (2002)

Latino panethnicity presumes the existence of a uniﬁed and distinctive Latino
community that has resulted from the context of and contact with American institu-
tions. Evidence that the rise in panethnicity is derivative of contact with American
institutions was provided by Jones-Correa and Leal (1996). Their analysis of data
taken from the LNPS found that characteristics usually associated with assimilation,
such as English proﬁciency, lack of support for bilingual education, higher level of edu-
cation and longer generational history in the United States, are highly correlated with
the adoption of a panethnic identity. Of particular interest is their ﬁnding that the rise
of panethnic identity has accompanied a decline in ethnic attachments. The question
is whether the tendency to identify in panethnic terms is a result of assimilation and
if so is it inevitable or is it a result of ethnic resilience and disadvantage? In total, the
evidence seems to be that the emergence of panethnic identity quite possibly arises
from a common language, an awareness of similarities between Latino subgroups, a

low standard of living, and a common desire to eliminate the inequalities imposed on

40

the Spanish-speaking by the larger society only later developing into a newly formed
Latino political consciousness that is situation-speciﬁc and strengthened as a result
of contact with American institutions. In any event, Latino panethnicity provides us
with an alternative orientation to understanding differences in political attitudes and
behaviors among and between Latinos that is better conceptualized by emphasizing
Latinos’ common interests and homogeneity. In order to understand high levels of
Latino political incorporation or conversely low levels of political alienation, scholars
should focus less on intergroup distinctions in favor of similarities across groups that
arise from their socialization into the American polity (de la Garza, 2004).

A more careful examination of these arguments, however, reveals that they may be
based on a false dichotomy. It is possible to analyze Latino subgroups separately while
simultaneously analyzing the extent to which their political interests revolve around
multiethnic lines. In addition, by separating Cubans, Mexicans and Puerto Ricans
it is possible to focus on the range of differences and similarities with respect to the
degree to which they have internalized a common culture among Latino groups, and
to understand and provide explanations for both. The existence of common group
interests need not presume that these identity groups are economically or socially

homogenous or politically monolithic.

41

3.2 Group Identity and Consciousness as a Politi-

cal Group Resource

Participation in United States politics has long been established as a function of re-
sources, engagement and recruitment or mobilization (Verba et al., 1995). Among
historically disenfranchised minority groups, such as Latinos, group—based resources
— resources that are collective in nature to the minority group (e.g., self-identiﬁcation
as panethnic and the belief that Latinos share a common culture or common “linked
fate”) _ may also play a critical role in mobilizing political participation (Alex-
Assensoh and Assensoh, 2001; Bobo and Gilliam, 1990; Chong and Rogers, 2003;
Harris, 1994; Miller et al., 1981; Shingles, 1981; Tate, 1993; \Nong et al., 2005). P0-
litical scientists have offered three justiﬁcations for the existence of identiﬁable and
distinctive group interests: 1) the notion of “linked fates”; 2) the prevalence of group
consciousness sustained through indigenous socio—political institutions; and 3) the
process of political socialization. Because the ethnic disadvantage model presumes
the existence of a distinct group interest, these justiﬁcations are further elaborated

on below.

Group Based Resources and Political Participation

Dawson’s book Behind the Mule was one of the ﬁrst manuscripts to develop a theory
of African American group interest that was grounded in the concept of linked fates:
“the relationship between a Black person’s sense of his or her own interests and the

same person’s sense of the interests of the racial group is the key to the apparent

42

political homogeneity of African Americans” (Dawson, 1994, p. 46). Dawson (1994)
and others (Gurin et al., 1989; Tate, 1994; Herring et al., 1999) have noted that,
among African Americans, linked fate stems from a long history of discrimination
and segregation in the United States. The development of “objective” interests is le-
gitimized via differences between African Americans and other groups such as whites.
“A new wave of empirical research conclusively demonstrates that where one can live,
for whom one is likely to vote, whether one is likely to encounter discrimination based
on stereotypes when entering the labor market, and whether one’s culture and intel-
ligence are considered inferior were still structured by race in the 19908” (Dawson,
2001, p. 42).

Scholars have consistently found that life chances are tied to ethnic group mem-
bership. For example, in their study of a national sample of blacks in the United
States, Hughes and Hertel (1990) found that 1) light skin color is associated with
higher socioeconomic status; 2) dark skin color is related to higher levels of black
consciousness and a stronger black identity; and 3) that the effect of skin color on
socioeconomic status changed relatively little in the period spanning 1950 - 1980. The
authors concluded that “the association between skin color and life chances appears
to be an aspect of black life in America that persists in spite of many social, political,
and cultural changes that have affected black Americans in the present century” (p.
1110). Therefore, because one’s life chances are inexorably linked to one’s racial or
ethnic group membership, members of a particular racial or ethnic group come to
understand that group members have a common linked fate.

Murguia’s study of phenotypes among Mexicans provides further support that

43

life’s chances depend on race or ethnicity. In his article Phenotype and Schooling,
Murguia shows how individuals who are Mexican, dark and Indian- looking are treated
differently than lighter skin Mexicans and that this differential treatment is repro—
duced in the educational system. They concluded that after controlling for variables
known to affect education, darker and Indian looking phenotypes had a signiﬁcant
negative effect on educational attainment for those of Mexican origin. While these
ﬁndings were not articulated in terms of a political context, it is not much of a stretch
to image that the discrepancy between darker and lighter skin individuals is also re-
inforced by the political system. For example, the policies that were meant to beneﬁt
members of minority groups are likely to have gone largely to the lightest skinned
members of that group. In addition, a substantial literature has documented that
education is an important resource for political action and engagement (Almond and
Verba, 1963; Barnes and Kaase, 1979; Conway, 1991; Jennings, van Deth, et a1. 1989;
Milbrath and Goel 1977; Verba, Nie, and Kim 1971).

The majority of research on the role of group—based resources was based on stud-
ies of African Americans (Calhoun—Brown, 1996; Harris, 1994; Jackson, 1987; Miller
et al., 1981; Olsen, 1970; Shingles, 1981; Tate, 1991; Verba and Nie, 1972; W'ong,
2000). These studies have revealed that African Americans tend to believe that their
personal welfare is heavily inﬂuenced by what happens to the entire racial group. Con-
sequently, they tend to evaluate policies, political parties, and candidates in terms of
their impact on group interests (Dawson 1994, 57; Gillmor 1993). Some recent studies
have questioned the relationship between group-based resources and political partic-

ipation and empowerment (Bobo and Gilliam, 1990) while others have found that

44

different forms of group-related resources, especially linked fate and group conscious—
ness, matter for distinct types of participation. For example, Wong (2003) found that
the concept of “linked fates” did not exhibit a strong relationship to registration or
voting but did matter for participation beyond voting. The notion of linked fates is
also applicable to Latinos as demonstrated by the fact that Mexican Americans are
democrats because of the tendency on the part of republicans to demonize Mexican
immigrants in the 1990s (Baldassare, 2000). According to Dawson, “the fact that two
African Americans can believe that their fate is linked to that of the race does not
mean that they agree on how best to advance their own and racial interests” (Dawson,
2001, p. 11). In similar fashion, Latino interests are not necessarily monolithic or

politically uniﬁed.

Group Consciousness and Group Interests

Ethnic awareness is deﬁned as the perception by members of a minority group of
the social distance separating them from the dominant group and the existence of
discrimination based on racial or cultural differences (Miller et al., 1981; Portes,
1984). The opposite of ethnic awareness is a sense of equality and the belief that
one can freely mingle with anyone in society (Portes, 1984). This belief is said to
characterize immigrants who are fully assimilated into mainstream society. Awareness
of social distance and of negative stereotypes has been said to be a decisive factor
affecting the behavior of minority-group members (Nagel and Olzak, 1982). Portes
and Zhou (1984), in fact, argued that the Latinos who are aware of discrimination

and who sense the relative deprivation that accompanies it, are more likely to develop

45

a strong sense of ethnic awareness. Clearly Portes’ (1984) conceptualization of ethnic
awareness is a synonym for “group consciousness,” a term political scientists use to
describe similar perceptions among minority groups. As I discuss further below, the
potential for group consciousness to develop among Latinos is more likely than ever,
as evidenced by the increase in panethnic identiﬁcation along with the corresponding
rise in Latinos’ perception of discrimination against them. Latinos are distinct from
other minority groups, for example, African Americans, because Latinos’ collective
identity, assuming for the moment it exists, is not self-imposed and hence remains
precarious.

In political science, group consciousness has been deﬁned as identiﬁcation with a
political group. Verba and Nie (1972) and Olsen (1970) have shown that blacks tend
to participate at rates higher than whites once socioeconomic status is controlled.
Olsen (1970), in particular, focused on blacks who identiﬁed themselves as members
of an ethnic minority versus those who did not. The sense of commonality and
collectivity among blacks led Olsen to propose the “ethnic community” thesis to
account for high rates of political participation among blacks in his sample. The
ethnic community thesis posited that the black community serves as the referent
group for “ethnic identiﬁers” and membership in such an activist community was
responsible for increasing levels of black participation. Since then, many studies of
African American political behavior have established the role of group consciousness
on African American participation (Bobo and Gilliam, 1990; Dawson, 1994; Gurin,
et al., 1989; Guterbock and London, 1983; Jackson, 1987; Shingles, 1981). Despite

economic and social cleavages in the African American community, African American

46

political behavior has been largely homogenous because individual self-interest stems
from group interests (Bobo et al., 1991; Dawson, 1994; Kaufman, 2000).

Individuals who identify with one group over another manifest a conscious aware—
ness of objective membership in that group as well as a psychological attachment to
that group (Conover, 1984). Social or economic circumstances alone are not enough
to foster an overall sense of belonging, the individual must also perceive that the
group to which he belongs lacks access to resources when compared to other groups
(Miller, 1981; Stokes, 2003). Miller et al., (1981) ﬁrst drew the distinction between
identiﬁcation with a group and politicized group consciousness. According to the
authors, group identiﬁcation connotes a perceived self-location within a particular so-
cial stratum, along with a psychological feeling of belonging to that stratum. Group
consciousness, on the other hand, involves identiﬁcation with a group as well as a
political awareness or ideology regarding the group’s relative position in society and
a commitment to collective action aimed at realizing that group’s interest. Group
consciousness implies an awareness of “shared status as an unjustly deprived and
oppressed group” (Bobo and Gilliam, 1990, p. 377) and a strong sense of panethnic
community (Dawson, 1994; Singles, 1981; Verba and Nie, 1972; Wong et al., 2005).

Miller et al., (1981) criticized Olsen and Verba and Nie for attempting to directly
associate group identiﬁcation and group consciousness. According to the authors,
there is no theoretical reason “to expect a simple direct relationship between group
identiﬁcation and political participation (Miller et al., 1981, p. 495).” While this
is arguably true among groups such as African Americans, it may not be true with

respect to immigrants. The processes that immigrants undergo may very well be

47

reﬂective of a politicized identity even in the absence of perceived or actual injustice.
When an immigrant self-identiﬁes as an American, for example, this is evidence that
his or her identity has undergone a radical transformation. Immigrants are unlikely to
have claimed an American identity before arrival in the United States. Quite possibly,
feeling proud to be an American goes hand in hand with this newly formed identity.
Political participation would be one manifestation of that identity that may or may
not have anything to do with a feeling of relative deprivation, in fact, the opposite
may be true. Nevertheless, because it is possible for an individual with no ethnic
group identity to develop one after a consciousness raising experience, scholars have
continued to deﬁne group consciousness as an individual’s awareness of the group’s
status relative to other groups and an overall commitment to further those interests
(Antunes and Gaitz, 1975; Gurin et al., 1980; Jackman and J ackman, 1973; Miller et
al., 1981; Wilcox and Gomez, 1990; Stokes, 2003; Wong et al., 2005).

Contrary to the tenets of the standard socioeconomic model of voting, which
explains racial differences in political behavior on the basis of blacks’ educational, oc-
cupational and economic achievement, Bobo and Gilliam (1990) concluded, as have
others before them, that when socioeconomic status is controlled, blacks actually
participated at higher rates than whites. This led them to explain participation in
terms of resources based on group membership. By doing so, they emphasized the im-
portance of context to representation and empowerment in the development of black
political consciousness. For example, in Race, Sociopolitical Participation, and Black
Empowerment, Bobo and Gilliam (1990), address the factors that motivate political

activity among African Americans. They found that in high empowerment areas,

48

operationalized as those areas possessing a black mayor, a greater degree of trust and
efficacy existed among blacks, which they deemed to be causally related to higher lev-
els of political participation. They reasoned that descriptive representation increased
the participation rate of blacks because the degree of sociodemographic similarity cre—
ated a ”macro- level cue” that fundamentally changed how blacks perceive the costs
and beneﬁts associated with voting. In other words, there was a net gain to voting in
high empowerment areas. An unintended consequence of black representation was a
transformation in the nature of black- white differences because black empowerment
tended to suppress white political participation in those areas as well.

The potential benefit of group-based resources on political participation has been
extended to Latinos as well on the assumption that Latinos perceive that access to
both material and group resources and to opportunities are strongly linked to paneth-
nic group identity. Cain, Kiewiet and Uhlaner (1991) provide evidence that Mexicans,
Puerto Ricans and Cubans can be considered a community that shares identiﬁable
substantive interests. In an examination of Latino partisanship, they concluded that
groups that experience discrimination and perceive unequal opportunities are more
likely to identify as Democrats, given the Democratic party’s image of being more
supportive of policies that favor disadvantaged groups. De la Garza et a1. (1992) pro-
vide evidence to the contrary. For example, a 1989 survey analyzed by de la Garza
and his associates reported that Latinos had very few experiences with discrimination
(LNPS, 1989). In fact, evidence from the 19908 paints a relatively conﬂicted view of
discrimination among Latinos. In 1990, only 39% of Mexican Americans claimed to

have personally experienced discrimination (de la Garza et al., 1992; de la Garza,

49

2004). In contrast, nearly two-thirds of Puerto Ricans reported having been the vic-
tim of discrimination. On the other hand, the 2000 Knight Ridder / San Jose Mercury
News reported that only 18% of Latinos had experienced discrimination in the 5 years
prior to the survey. In Chicago, only 6% of Latinos mentioned discrimination as a lo-
cal concern. These low percentages led de la Garza to speculate that “discrimination
does not appear to be so pervasive as to motivate Hispanic citizens to ban together
and vote in pursuit of a common electoral agenda, as Latino advocates expect” (de
la Garza, 2004, p. 98).

Several important clariﬁcations/ distinctions are warranted. First, in Portes and
Zhou’s original theoretical formulation of ethnic disadvantage, they emphasized the
perception of discrimination in the reconstitution of ethnic awareness and solidarity.
Compare the results of the 2004 National Survey of Latinos: Politics and Civic Life.
In this survey, only 26.6% of Mexicans, 1% of Puerto Ricans. and 36% of Cubans
reported that race relations were as good in the United States as they are in their
country of origin. In addition, when asked whether the respondent believed that
discrimination against Latinos in general is a “major” or “minor” problem or not a
problem at all, the overwhelming majority of Latinos responded afﬁrmatively (i.e.,
that it was a major or minor problem; 82.2% of Mexicans, 87.4% of Puerto Ricans, and
73.7% of Cubans). This is true despite the fact that Latinos on the whole reported
fairly low levels of direct or personal discrimination (41.5% of Mexicans, 42.8% of
Puerto Ricans and 18.7% of Cubans). This lends support to the possibility that the
conﬂicted ﬁndings of the 19908 were due to an increasing awareness of discriminatory

conditions among immigrants in the decade spanning 1990-2000. An examination

50

of the relative context in which immigrant assimilation is occurring, as well as the
perception of discrimination in the US, may indeed reveal that Latinos are motivated
to unite in the pursuit of common interests via the American political system. The
evidence since the late 19908 points to the fact that when Latinos are distinguished
from other groups on the basis of observable characteristics or along lines of power or
class, they respond panethnically (Calderon, 1994). Indeed, a thorough examination

of these issues is one of the theoretical contributions of this dissertation.

3.3 The Process of Identiﬁcation

In 2001, Leonie Huddy (p. 128) made the following claim

Despite the recent emergence of identity politics around the world, re—
searchers of political behavior have been slow to incorporate the concept
of identity into their empirical studies. This seems odd, given that de—
mands for group respect and recognition are at the heart of new social

movements that argue for the rights of diverse ethnic and racial groups.

She continues that “it is extremely important to understand why some indi-
viduals in a given social and political context adopt a group identity, whereas others
in identical circumstances do not.”

The social—psychological literature on identity formation provides a useful theo-
retical framework in which to understand how group consciousness emerges via the

identiﬁcation process. Tajfel (1981) deﬁned ethnic identity as an aspect of a person’s

51

self concept that derives from his or her knowledge of n‘iembership in social groups to—
gether with value or emotional signiﬁcance attached to that membership. Rotheram
and Phinney (1987) deﬁned ethnic identity as one’s sense of belonging to a group,
and a part of one’s thinking and perceptions as well as feelings that are part of being
a member of that group.

Ethnic identity is an important component of overall identity formation. Ethnic
identity is a psychological construct reﬂecting identiﬁcation with, and membership
in, an ethnic group (Cuellar, et al., 1997). Ethnic identity includes self-identiﬁcation
(Rumbaut, 1994), feelings of belongingness and commitment to a group, a sense of
shared values, and attitudes towards one’s own ethnic group (Phinney, 2001). Exist-
ing research on identity development suggests that reformulations of one’s identity can
occur at any point in the lifecycle (Umana—Taylor, 2002). Ethnic identity is dynamic
and changes in response to contextual factors including socialization experiences in
the family, the ethnic community, and the larger setting (Phinney, 2001). There—
fore, identity formation involves the integration of a number of relational contexts
(Josselson, 1994). Among these and of central importance is embeddedness where
individuals construct their identity in relation to others including their ethnic group.
In recognition, Markstrom—Adams (1992) suggested that ethnic group membership is
one of the social contextual environments that inﬂuences identity formation.

Among newcomers to American society, in particular, ethnicity is especially salient
because immigrants are faced with the difﬁcult task of ethnic re-identiﬁcation, forced
to identify with a less powerful group, and confront several negative stereotypes asso—

ciated with that group (Rotheram and Phinney, 1987; Spencer and Bornbusch, 1990).

52

The theoretical perspectives that have been proposed by social psychologists to ex-
plain the development of ethnic identity and self-esteem among Latinos are discussed

belowl.

3.4 Ethnic Identity and Acculturation Theory

One theoretical approach that has proven useful in structuring the study of ethnic
identity is based on levels of acculturation. The acculturation theory of ethnic iden-
tity argues that ethnic identity changes as a function of acculturation processes and
inﬂuences. The focus has been on interactions between members of different cultures
and on the coexistence of minority groups in two cultures (Bautista del Demanicoet
al., 1994). Ethnic identity becomes salient as part of the acculturation process that
takes place when immigrants come to a new society (Phinney et al., 2001). The
distinction between the constructs of ethnic identity and acculturation is often un-
clear (Liebkind, 2001; Phinney, 1990; Phinney, 1998). Nevertheless, Cuellar (1997)
found that ethnic identity formation and acculturation, while related, are separate
processes. Acculturation is a broader construct that encompasses a wide range of
behaviors, attitudes and values that change when two or more different cultures come
into contact with one another (Phinney et al., 2001). Ethnic identity formation is
that aspect of acculturation that focuses on the subjective sense of belonging to a

group or culture (Phinney, 1990).

 

1These theories have been developed and applied speciﬁcally with reference to
Latino adolescents, but the process of negotiating identity and identity formation is
similar among adolescents and newly immigrant populations

53

Classical assimilation theories view acculturation as a linear process of change
requiring the relinquishment of one’s culture of origin and assimilation into a new
culture. On the other hand, the ethnic disadvantage and segmented assimilation
models are best understood as two-dimensional. Somewhat ironically, psychologists
studying acculturation have also tended to view the processes related to ethnic identi-
ﬁcation as a two—dimensional process (Berry, 1990; LaFromboise, et al., 1993; Nguyen
et al., 1999). These models are based largely on the work of Berry (1990, 1997)
who argued that two dominant aspects of acculturation, preservation of one’s her-
itage culture and adaptation to the host society, are conceptually distinct and vary
independently. Berry suggested the following two questions as a means of identify—
ing strategies used by immigrants in dealing with acculturation: Is it considered to
be of value to maintain one’s cultural heritage? Is it considered to be of value to
develop relationships with the larger society? The combination of yes/ no responses
deﬁne the four types of acculturation strategies. Integration is deﬁned by positive
answers to both questions while marginalization is deﬁned by negative answers to
both questions. A positive response to the former and negative response to the latter
deﬁnes separation and the reverse deﬁnes assimilation. This model is useful in that it
highlights the fact that acculturation can proceed in diverse ways and that it is not
necessary for immigrants to give up their culture of origin in order to adapt to a new
society (Phinney et al., 2001). It also allows for the possibility and reality of multi-
culturalism. Additionally, it acknowledges that there may be alternative pathways to
incorporation in contrast to earlier theories that viewed immigrant acculturation as

resulting in either assimilation or marginalization.

Of particular interest to the current research endeavor is the extent to which immi-
grants adopt the values of the mainstream culture while at the same time maintaining
the values of their own ethnic group. Several researchers have found that identifying
with one’s ethnic group is the healthiest form of ethnic identity. Researchers work-
ing within this framework have grouped individuals into categories based on their
inclination to adopt mainstream values or maintain the values of their country of
origin. For example, Bautista de Domanico et al., (1994) found that biculturals had
signiﬁcantly higher levels of self-esteem than monoculturals and concluded that a
bicultural identity is related to greater flexibility which facilitates adaptation. In ad-
dition, they speculated that individuals with bicultural identities can relate to both
cultures and feel less alienated from either. Other studies, however, report different
results. Rotheram-Bows et al., (1996), for example, found that strong ethnic identi-
ﬁers scored signiﬁcantly higher than individuals in the other two groups. Similarly,
Cuellar et a1. (1997) studied ethnic identity in speciﬁc acculturated subgroups of
a Mexican origin population. They found that ethnic identity scores were highest
among ﬁrst generation Mexicans, less acculturated subjects, and traditional accul-
turative types. They further found that higher levels of acculturation were associated
with diminished feelings of afﬁrmation and belonging and lower feelings of ethnic iden-
tity achievement. More speciﬁcally, ‘High Biculturals’ were found to obtain higher
scores in ethnic identity than ‘Low Biculturals’ and ‘High Biculturals’ were found
to be oriented more toward others than those who were classiﬁed as ‘Traditional’ or
‘Assimilated.’ Phinney and Devich-Navarro’s (1997) research with Mexican Ameri-

can adolescents in the United States showed that ethnic identity, but not American

55

identity, was a predictor of self-esteem. These relationships are influenced by the par—
ticular setting and immigrants’ perceptions of their place in that setting (Phinney,
2001). For example, Birman et al., (2002) found that immigrants who adopted a na-
tional identity manifested higher levels of school integration than those who adopted
an ethnic identity. As explanation for these ﬁndings, the authors offered that “be-
ing American-oriented is advantageous in contexts that demand acculturation to the
American culture, that is school and American peers” (Birman et al., 2002, p. 599).
A few studies in political science have found that having strong ties to one’s ethnic
or national origin group is not an impediment to feeling proud to be an American
(Citrin et al., 2002; de la Garza et al., 1996). When pressures for assimilation are
strong, opting for an American identity may make it easier for immigrants to adjust

to the surrounding environment (Phinney, 2001).

3.5 Social Identity Theory

In Berry’s theoretical framework, acculturation strategies that immigrants adopt
moderate the adaptational outcomes of acculturation. Additional factors also in-
fluence the acculturation process including age, gender, cultural distance from host
society, experiences of discrimination, social support (Phinney, 2001) and contextual
factors such as size of the ethnic population and attitudes of the host society (Berry,
1990).

Social identity theory (Ta jfel, 1981; Tajfel and Turner, 1986) provides a framework

for understanding how ethnic identity and psychological well—being changes as a result

of contextual relations in the host society. This theory posits that individuals’ self-
concepts derive from knowing that they are members of particular social groups. It
implies that a strong link exists between group identiﬁcation and self-concept. People
strive to achieve or maintain a positive social identity which tends to boost their
self esteem (Phinney, 2001). Maintaining a positive identity derives from favorable
comparisons between one’s ingroup and relevant outgroups. Even when individuals
attempt to maintain a positive social identity, preconceived notions regarding the
social groups to which they belong can overwhelm their choice of group membership.
For this reason, researchers have focused on the perceptions that the dominant society
has of ethnic groups and on how these perceptions influence an individual’s decision
to identify with one group over another.

Researchers espousing social identity theory have tended to focus on self-
identiﬁcation. Individuals who internalize unfavorable opinions of their ingroup in
relation to the outgroup may seek to leave that group or ﬁnd ways of achieving more
positive distinctiveness. This is why the context of reception is crucial: newcomers
to American society are often viewed in derogatory terms, and this is reinforced in
American culture and in the media. In the face of devaluation of their ingroup, immi-
grants may adopt a variety of self-preservation methods. The various methods depend
on whether or not devaluation is attributed internally or to external circumstances
(Liebkind, 2001).

My theoretical contribution to political science in general, and the subﬁeld of
ethnic politics in particular, is that the process of immigrant re—socialization (Finifter

and Finifter, 1989) regarding ethnic identity fundamentally shifts the content and

57

meaning that immigrants attach to their own ethnic identity. The meaning of one’s
ethnic identity has consequences for the political incorporation of Latino immigrants.
These immigrants arrive in the US and are influenced by the context in which they
are received. Immigrant reception lies on a continuum from friendly and Open to
hostile and oppressive (Portes, 1984). The trajectories of immigrant assimilation are
in turn inﬂuenced and moderated to a large extent by this context. As Huddy (2001,
p. 130) so appropriately stated, “American identity does not mean the same thing
to all Americans. And it is the meaning of American identity, not its existence, that
determines its political consequences. Yet social identity researchers have tended
to ignore this subjective aspect of identities, paying considerable attention to the
existence of simple group boundaries while ignoring their internal meaning.” Social
identity and acculturation theory provide the causal mechanisms by which immigrants
attach meaning and value to their ethnic identities. The various theories of immigrant
assimilation outlined above serve to contextualize the processes of acculturation and
identiﬁcation. Taken together, students of Latino politics who use this perspective
stand to gain a more accurate and realistic understanding of how immigrants relate
to the American political system and adopt the political attitudes and behaviors we
observe empirically. As I discussed in the previous chapter, immigrant attitudes and

behaviors provide a ﬁrst step in articulating a theory of immigrant incorporation.

58

Chapter 4

Data, Research Methodology and

Hypotheses

This chapter discusses three predominant modes of thought as to the likely outcome
of Latino acculturation: the straightline model, the ethnic disadvantage model and
the segmented assimilation model. By generating hypotheses based on these three

theories, the intent is to focus on the processes that facilitate political assimilation.

4.1 Data

Data for this study comes from the several sources listed below.

4.1.1 The 2002 National Survey of Latinos in America

The 2002 National Survey of Latinos in America was conducted by telephone between

April and June, 2002 among a nationally representative sample of adults selected at

59

random who were 18 years and older. The survey was administered by The Wash-
ington Post/ Henry J. Kaiser Family Foundation/IIarvard University’. In order to
represent the opinions of Latinos living in the United States, interviews were con-
ducted with a statistically representative sample that could be examined nationally,
in target regions of high Latino concentration, and by country of origin. A highly
stratiﬁed, disproportionate random digit dialing sampling design of the 48 contiguous
United States was employed. One beneﬁt of using this particular survey pertains
to its inclusion of an oversampling of Latinos. This oversampling resulted in 2,929
observations on Latinos (1,284 were non—Latinos), 1744 of which were United States
citizens.

The survey examined how members of the Latino community identify themselves
as well as their (1) views of the United States; (2) experiences with discrimination; (3)
linguistic preferences; and (4) economic and ﬁnancial situations. Because of its broad
scope, these data allowed for the possibility of making conclusions about political
attitudes and behaviors that are far reaching rather than state speciﬁc (Baretto and
Munoz, 2003). In the analysis based on this data, results were weighted to reﬂect the

actual distribution of respondents throughout the United States.

 

1I would like to thank the Pew/ Kaiser Foundation for their generosity in allowing
me the use of these data.

60

4.1.2 The 2004 National Survey of Latinos: Politics and
Civic Life
Additional data utilized in this dissertation are drawn from the Pew Hispanic Cen—
ter/Kaiser Family Foundation 2004 National Survey of Latinos: Politics and Civic
Engagement. This survey was conducted between April 21 and June 9, 2004 among a
nationally representative sample of 2,288 Latinos adults, aged 18 or older, who were
randomly selected for inclusion in the survey. The survey included 1,166 registered
voters. Respondents’ views on a wide range of political issues and concerns, immi-
gration, identity, assimilation, context of reception in the host country, experiences
with discrimination and contact with one’s country of origin were examined. The
survey allowed for comparisons of characteristics, attitudes and civic participation
both among Latino subgroups and among registered voters, those who are eligible to
vote but have not registered and those Latinos who are not US. citizens. The sample
design employed was similar to the 2002 National Survey and again, the subsequent

analysis weighted the data to reﬂect the actual distribution of respondents throughout

the United States.

4.2 Components of the Straightline Model of As-
similation

In order to address the competing hypotheses pertaining to the straightline and ethnic

disadvantage models of immigrant assimilation, two classes of independent variables

61

were introduced into the analysis. The ﬁrst class considered variables related to assim-

ilation while the second class considered variables related to group-based resources.

4.2.1 Cultural Indicators of Assimilation

The classical perspective views assimilation as a process in which social, psycho-
logical, cultural and economic assimilation, whether occurring simultaneously or in
phases, eventually leads to the abandonment of old cultural practices and political
loyalties. According to classic assimilationist thought, ethnic perceptions among im-
migrant groups decline as assimilation into American society increases. Proponents of
the classical assimilation perspective such as Gordon (1964) and Fuchs (1990) suggest
that in the long run Latinos are likely to become less “minority” centered in their
political orientations (Kaufman, 2003). Accordingly structural assimilation, decreas-
ing perceptions of discrimination, lack of contact with dominant groups, and social
isolation will result in Latino political attitudes and behaviors that are increasingly
similar to Anglos. It is worth noting that, in what follows, the assimilationist po-
litical culture is assumed to be the norm for the “prototypical American” (Citrin et
al., 2003). In accordance with Citrin et a1. (2003), participation in American politics
and the adoption of American attitudes and values are taken as evidence of political

assimilation.

62

4.2.2 Operational Deﬁnitions, Exogenous Predictors and Re-
search Hypotheses

The measures of assimilation used in this dissertation are categorized under ﬁve sep-
arate headings corresponding to demographic factors as well as cultural, structural,

attitudinal / behavioral and identiﬁcational assimilation.

Operationalizing Cultural Assimilation

Cultural assimilation is a multidimensional construct that includes behavioral and
cognitive components. Adopting US cultural habits and believing in its core funda-
mental values while simultaneously rejecting other countries’ values can be considered
separate but related dimensions of cultural assimilation (Huntington, 2004). Linguis-
tic preference and language skills are common ways to gauge the extent to which
immigrants have adopted US cultural characteristics. Several studies (Pearson and
Citrin, 2002; Citrin, Reingold and Green, 1990) have shown that white Americans
overwhelmingly believe that speaking English is very important for being a “true
American.” Knowledge of English has consistently been identified as a factor that
facilitates the assimilation process (W'arner and Srole, 1945; Simpson and Yinger,
1972; Portes, 1984). Linguistic preference is also an indicator of political assimilation
because language mediates political responsiveness (Baretto, 2004) and as such is
an important predictor of political attitudes and behavior. Language has also been
shown to facilitate political communication because of its tendency to appeal to po-

tential voters and to be an expression of a politicized ethnic identity (Johnson, 2003).

63

For example, in the 2000 and 2004 presidential elections, George W. Bush spoke
Spanish on the cable network television station Univision for the purpose of appeal—
ing to Latino voters and it seemed to work since a large portion of Hispanics voted
for him (30% in 2000 and 40% in 2004). If ﬂuency in a nation’s language inﬂuences
a person’s ability and opportunities for immersion into the nation’s political culture,
then those who are more proﬁcient in the dominant language should be more likely
to be politically assimilated.

Language acquisition is a necessary but not sufﬁcient condition of cultural assimi—
lation. Beyond tangible indicators such as the development of linguistic proﬁciency, a
person’s attitudes may also reﬂect the extent to which they are adapted to American
culture. Some studies have found substantial differences between native- and foreign
born immigrants’ attitudes towards the appropriate gender roles for men and women
(Phinney, 2002); therefore, gender role attitudes provide a good measure of Latino
cultural assimilation. Attitudes towards appropriate gendered behaviors are likely to
vary as a function of immigrants’ contact with, or involvement in, US culture. For ex-
ample, gender roles in traditional Hispanic cultures are clearly deﬁned with authority
typically given to males (Gowan and Trevino, 1998; Mayo and Resnick, 1996; Soto and
Shaver, 1982). In traditional Hispanic families, males hold dominant positions and
are expected to assume responsibility for the family. In addition, males are granted
more freedom from an early age while females are encouraged to be submissive (Mayo
and Resnick, 1996) and responsible for housework and childcare. In contrast, studies
have shown that among Americans, attitudes towards women’s roles in society have

become increasingly egalitarian since the late 19705 and that the majority of indi-

64

viduals favor woman’s dual status as a homemaker and professional. For example,
Zou and Tang (2000) found that both men and women held egalitarian views towards
women’s roles as mother, worker and provider. In addition, the importance placed
on the family unit itself has been found to decrease with time spent in the United
States. Barboza and Williams (2004) found that the primacy that Latinos place on
family bonds as opposed to non-familial relationships declines with residency status
(Unpublished manuscript). Clearly, immigrants who come to America are exposed
to societal norms about family life and their attitudes towards it change over time
(Phinney, 2002). Therefore, lack of traditional attitudes about the proper roles for
men and women gauge the extent to which immigrants have internalized and adapted
to American cultural norms. Classical theories would suggest that individuals who
adopt egalitarian views are more culturally assimilated and are likely to be more
politically assimilated as well.

The permanency of cultural ties to the United States is reﬂective of the psycho-
logical commitment to become a full ﬂedged member of the American polity, which
includes becoming a citizen, adopting a partisan afﬁliation or otherwise participating
in politics (F inifter and Finifter, 1989). An additional measure of cultural assimila-
tion is therefore reﬂected in the importance immigrants place on creating strong ties
to the United States while simultaneously decreasing ties to their country of origin.

Finifter and Finifter (1989) found that among Australian migrants, partisanship
and overall ideology are important factors in helping migrants adapt to new envi-
ronments, and that new political learning is generally dependent on previously es-

tablished political attitudes. The propensity to be involved and committed to home

65

country politics may indicate an immigrant’s propensity to be engaged and active in
the Australian polity. In environments where there are strong pressures to assimilate,
for example, the new political environment may unconsciously exert an inﬂuence on
the adoption of new political attitudes and behaviors (Finifter and F inifter, 1989).
Consequently, newly acquired attitudes and behaviors might parallel those in the host
country. On the other hand, old patterns that inﬂuence the development of new po—
litical attitudes and behaviors may persist. The process of translation which speaks
to this phenomenon is thought to occur when individuals who are highly active in the
politics of their home country consciously or unconsciously seek out opportunities to
participate in Australian politics (Finifter and Finifter, 1989).

The maintenance of cultural practices may seem to interfere with the process of
adaptation to the environment. Some researchers have speculated that advances in
technology, increased economic and political ties between the United States and an
immigrants’ home country, and outreach efforts by home countries to their diasporas,
have enabled immigrants to forge and sustain transnational links to a greater degree
than ever before (Pantoja, 2005). While these ties have traditionally been viewed
as impediments to political incorporation (Dwyer, 1991; Torres-Saillant, 1989), more
recent studies have argued that transnational participation may co—exist and even pro-
mote participation in US politics (Levitt, 2000; Graham, 1997; Hernandez and Jacobs,
2001). Quantitative studies have been inconclusive as to the effects of transnational
networks on immigrant political incorporation, however. Jones-Correa (2001) found
that Latin American immigrants from countries offering dual nationality had a higher

propensity to naturalize while Yang (1994) found that it had a. prohibitive effect. With

66

respect to political participation, DeSipio (2006) observed that certain measures of
transnationalism are positively associated with an immigrant’s desire to remain in
the United States, participate in civic organizations, and naturalize. Finally, Pantoja
(2005) analyzed the relationship between transnational ties and immigrant politi—
cal incorporation among Dominicans residing in Washington Height, New York. He
found that individuals who are active in the politics of the Dominican Republic and
have most of their families there are less likely to pursue US citizenship. On the
other hand, belonging to an association concerned with the events in the Dominican
Republic and participation in the politics of the Dominican Republic exert a posi-
tive and powerful inﬂuence on US political participation. Still other researchers have
found that the adaptation to US culture and strong ties to one’s country of origin
vary independently (Keefe and Padilla, 1987; Kitano and Daniels, 1988; Hutchinson,
1988)

The straightline assimilation theory would predict that the assimilation process
will parallel a decline in attachments to an immigrant’s country of origin. The ex-
pectation is that decreasing contact with one’s country of origin should result in the
relinquishment of that country’s values and customs in favor of the host polity. Con—
versely, frequent contact with the host country’s institutions should provide sufficient
exposure to result in acceptance of American political norms (Citrin 2003). If clas-
sical assimilation theories provide an accurate depiction of Latino acculturation, the

following research hypothesis should be supported by the data:

The cultural assimilation hypothesis

67

H11: Latino immigrants who are culturally assimilated will be more ac-
tive participants in American politics and will show higher levels of
attachment to the American polity than those who are not culturally

assimilated.

Operationalizing Culture In this dissertation, cultural assimilation as measured
by linguistic skill is operationalized as English dominance. The 2002 national Latino
Survey asked respondents how well they could read a newspaper and carry on a con-
versation in English and Spanish, respectively. Two measures of linguistic competence
(English, Bilingual or Spanish) were derived from these variables. First, two indicator
variables were used to assess the differential impact of bilingualism (d = 1) or English
dominance (d = 1) on the one hand and Spanish dominance (d = 0) on the other. A
scale based on these questions was constructed with scores ranging from -3 (ﬂuent in
Spanish but not English) to 3 (ﬂuent in English but not Spanish). Those who scored
0 were considered equally proﬁcient in both languages (i.e., they are bilingual).

The measures of ethnic ties or attachment to home country cultures that are used

2, voting, and sending money or planning to

in this dissertation include being a citizen
move back to one’s country of origin. The 2002 National Survey asked respondents

who were born outside of the US or in Puerto Rico whether (1) they are a legal

citizen of their country of origin; (2) they have voted in their country of origin since

 

2Puerto Ricans are US citizens by birth. The Cuban Constitution provides that
Cuban citizenship is lost by becoming a citizen of a foreign country and dual citizen-
ship is precluded, Constitution of the Republic of Cuba, 1992, Article 32. Addition—
ally, the United States has bilateral agreements with a number of countries, among
them Cuba, that effectively preclude the possibility of dual citizenship.

68

moving to the US; (3) they regularly send money to someone in their home country;
or whether (4) they plan to someday move back to their home country. Responses
to these questions were summed to form a scale that indicates the intensity of home
country ties. The scale ranges from 4 to 8 with increasing values indicative of fewer
ties to one’s home country.

The degree to which the respondent subscribes to the US as a “melting pot” is
a measure of how psychologically committed he or she is to fundamental American
values. Both national surveys, in two separate questions, asked respondents whether
they believe that Latino subgroups should change in order to blend into the larger
society (d = 0), as in the idea of a “melting pot”, and how important it is for
Latino subgroups to maintain their distinct culture ((1 = 1). The dummy variable
captures the differential effect between an immigrants’ view that the US comprises
“one culture” to which they belong, or many different cultures to which they are only

one cultural group of many residing in the United States.

Identiﬁcational Assimilation

It is widely agreed that a primary outcome in the assimilation process is the tendency
to identify oneself as an American rather than a member of one’s country of origin.
Huntington (2004, p. 239) argued that the extent to which immigrants identify
with the United States as a country is an important indicator of their willingness
to assimilate. Similarly, Gordon (1964) believed that identiﬁcational assimilation
was a prerequisite to political assimilation. If these scholars are correct, then we

should expect that immigrants who self-identify as an American are more politically

69

assimilated than individuals who self-identify in terms of either a panethnic or ethnic

group.

The identiﬁcational assimilation hypothesis

H i2: Latino immigrants who adopt an American identiﬁcation will be more
active in American politics and will show higher levels of attachment

to the American polity than those who do not.

Two attitudes that provide direct measures of American identity are self-
identiﬁcation as American and identiﬁcation of the United States as the individual’s
“true homeland.” The 2002 survey asked respondents whether they self-identify in
national, national origin or panethnic terms. Respondent’s who said yes to one of
these terms were then asked the following question: “You have said that you describe
yourself as an American, a Latino/ Hispanic, and as a (Respondent’s / Parent’s country
of origin). In general, which of the terms that you use to describe yourself is the term
you use ﬁrst? Second? Third?” A respondent’s preference for identiﬁcation with their
own ethnic or panethnic group or as an American was indicated by their response
to this question. Responses were recoded into dummy variables using the following
coding scheme: d = 1 if American, d = 0 otherwise; d = 1 if from respondent’s
country of origin, d = 0 otherwise3. The reference category is Latino/ Hispanic unless
otherwise noted. Additionally, respondents were asked what country they consider

to be their “true homeland.” Responses were coded d = 1 if respondent considered

 

3If this variable was a dependent variable, it was left in its original form as 1 :—
country of origin; 2 = Latino/ Hispanic; and 3 = American.

70

their country of origin to be their true homeland and d = 0 if they considered the

United States to be their true homeland.

Structural Assimilation

Structural assimilation occurs when different cultural groups come into contact in
institutional settings (Aguirre et al., 1989). Institutional barriers include factors that
preclude access to education and prevent occupational mobility. Education is a pow-
erful assimilative force for two reasons: ﬁrst, educated immigrants are generally more
accepted in society; second, education facilitates overcoming linguistic and cultural
barriers which prevent access to societal institutions (Portes, 1984). As Alba and
Chamlin (1983) noted, “education is assumed widely to bear an important relation
to assimilation in that it introduces the individual to an ethnically widening circle
of classmates as he or she proceeds higher in the educational system” (p. 242). In this
dissertation, high income and education levels are taken as evidence of an immigrant’s
success in becoming structurally assimilated. Classical assimilation theory predicts

that structural assimilation is positively correlated with political assimilation.

The Structural Barrier Hypothesis

H i’ : Latino immigrants who have higher levels of education, income and
economic attachment are more likely to participate in politics and will
show higher levels of attachment to the American polity than those

who do not.

71

The Structural Assimilation/ Ethnic Competition Hypothesis

H i1: Latinos who live in highly concentrated Latino areas with large for-
eign born populations will be less likely to participate in politics and
will show higher levels of attachment to the American polity than those

who do not.

Overall, the models include several measures of structural assimilation. All three
data sets asked respondents to provide information about their income and educa-
tional status. Four contextual measures of structural assimilation were also included:
(1) percentage of native born persons in a neighborhood; (2) percentage of co-ethnics
in a neighborhood; (3) percentage of citizens in a neighborhood; and (4) ratio of in-
come to poverty in a neighborhood. These percentages were computed from the 2000
US. Census at the level of census tract and appended to the 2002 National Survey

4. Besides these percentages as continuous measures, categorical versions

of Latinos
of them were included in some models in order to focus on individuals who are not
living in highly concentrated immigrant/coethnic areas (Greenman and Xie, 2000).
When a dummy variable was created out of these measures, the cutpoints were set at
the group speciﬁc means. For example, if a dummy variable was included, the mean
of the variable was ﬁrst calculated and a dummy variable was created that equals 1
for all values above the mean and 0 for all values of the variable below the mean.

The mean was calculated only for respondents who were eligible for inclusion in the

analysis.

 

4The other data sets used in this dissertation did not have the appropriate geocodes
that are necessary to append the variables.

72

Attitudinal/Behavioral Receptional Assimilation and Context of Recep—

tion

The straightline and ethnic disadvantage perspectives have very different predictions
as to the likely outcome of unfavorable reception, including discrimination, on pre-
dicting political attitudes and behaviors. Gordon (1964), for example, argued that
assimilation would proceed only in the absence of discriminatory conditions in the
host country. More recently, researchers continuing in the tradition of classical as-
similation theory have argued that in the long—run, racial distinctiveness will not be
a major barrier to incorporation (Alba and Nee, 1997, 2003; Perlman and Waldinger,
1997) because racial boundaries have proven to be ﬂuid (Xie and Greenman, 2000).
Alba and Nee (2003) further argue that today’s immigrants have benefited from the
civil rights movement which has effectively increased the cost of discrimination. The
implication is that as overt discriminatory acts decrease barriers to incorporation are
minimized.

Portes (1989) very explicitly considered a context of reception characterized by
prejudice in the host society as a dimension along which contemporary immigrant
experiences vary (Portes, 1989). “The stance of host governments, employers, the
surrounding native population and the characteristics of the pre-existing native com-
munity, if any, are important aspects of the situation confronting new immigrants”
(1989, p. 618). The awareness of discrimination, according to Portes, is primarily re-
sponsible for the reconstitution of ethnic culture. According to Cuellar et al. (1995),

“signiﬁcant barriers (personal or otherwise) based on animosity and hatred toward

73

the new culture” are responsible for an immigrant’s desire to retain values of the
original culture.

Several studies have demonstrated that experience with discrimination fosters
naturalization and civic engagement among Latinos (DeSipio, 2003; Schkinderkruat,
2004, Wrinkle et al., 1996). Dawson (1994) and Uhlaner (1991) offered as explanations
that experiences with discrimination serve to enhance ethnic consciousness, which is
a critical resource for ethnic/racial political engagement. Lee and Ramakrishnan
(2002) found a positive relationship between having an ethnic name, perceiving that
discrimination among Latinos is a problem and having a heightened sense of ethnic
identity and certain forms of political participation. Discrimination may foster a sense
of vulnerability leading immigrants to seek access to the political system in order to
enhance their legal rights and social standing (Pantoja, 2005), both of which can
reduce overt acts of discrimination.

Discrimination against one’s own ethnic group as opposed to personal discrimina—
tion may play an important role in facilitating political assimilation among certain
groups of Latino immigrants. Nevertheless, no study to date has considered the role
that group versus individual discrimination plays in the political assimilation of Latino
subgroups. There exist, however, two studies that investigated whether perceptions
of group level discrimination have more of an impact on political engagement than
individual-level discrimination among Latinos generally (DeSipio, 2002; Schildkraut,
2004). DeSipio (2002), for example, found that the perception of individual-level
discrimination increases the likelihood of engaging in nonelectoral forms of political

activity while the perception of discrimination against onc’s national origin group

74

had no impact. In contrast, Schildkraut (2004) found that perceptions of discrimi—
nation against one’s self promote behavioral and attitudinal alienation from political
processes, speciﬁcally electoral participation and registration, that can be counter-
acted when an immigrant identiﬁes with his or her country of origin.

The analyses that follow include both group-based and individual level measures
of discrimination. To operationalize group—based discrimination, respondents in 2002
were asked three questions designed to measure whether they believed Latinos as
a group have been discriminated against: “In general, do you think discrimination
against Latinos is a major problem, a minor problem, or not a problem in (a) schools;
(b) the workplace; or (c) in preventing Latinos in general from succeeding in America?
Answers to these questions were combined to form a scale ranging from 3 (respondent
believes all three types of discrimination to be a major problem) to 9 (respondent
believes that none of these types are a problem). As a measure of individual based
discrimination, these same respondents were asked how often, in their day-to-day
life, the following things happen to them because of their race or ethnic background:
(a) they are treated with less respect than other people; (b) they receive poor ser-
vice compared to others at restaurants or stores; or (c) they are called names or
insulted. Possible responses included “Very often,” “Fairly often,” “Once in a while”
or “Never.” A scale was created from these responses that ranges from 3 (if all three
things happened very often) to 12 (if none of these things ever happened). Increas-
ing values correspond to more frequent acts of discrimination while decreasing values

indicate more favorable contexts5. An additional measure of personal discrimination,

 

5In some cases throughout this dissertation, this variable was reverse coded. I

75

included in both national surveys, included responses to the following question: “In
the past 5 years, have you or a family member experienced discrin’iination?” (Yes =
d = 1, N0 = d = 0). Both the 2002 and 2004 survey asked respondents to answer
the following question: “In general, do you think discrimination against Latinos is
a major problem, minor problem, or not a problem in preventing Latinos in general
from succeeding in America?” In both cases, responses “major problem” were coded
1 and response “minor problem” and “not a problem” were coded d = 0.

Besides discrimination, an additional measure of the context of reception is given
by respondents’ perception of the relative conditions of life in the United States
versus their country of origin. Both national surveys included measures designed to
assess the extent to which immigrants believe that conditions in the US are poor
relative to their home countries. Respondents were asked “Overall would you say
that (a) Treatment of the poor is better in the United States, better in country
you/ your parents/ ancestors came from/Puerto Rico, or about the same?; (b) The
moral values of the society is better in the United States, better in country you/ your
parents/ ancestors came from/Puerto Rico, or about the same? (c) The strength of
family ties is better in the United States, better in country you / your parents / ancestors
came from/Puerto Rico, or about the same? of (d) Opportunity to get ahead is better
in the United States, better in country you / your parents / ancestors came from / Puerto
Rico, or about the same?” Possible responses included “Better in the United States,”

“Better in the countr ou came from” or “the same.” Individual res onses were
’

 

have made every attempt to indicate when this is the case. For examples of each
case, see Appendix B and C.

76

recoded -1 for “Better in the country you came from,” 0 for “the same,” and 1 for
“Better in the United States” and then summed to create an index. The indexed
values ranged from -4 to 4 with - 4 indicating that all four conditions were better in
respondent’s home country, 0 indicating that respondent believed that there was no
difference in either of the conditions, and 4 indicating that all four conditions were
better in the United States.

If context of reception affects the course of Latino political assimilation, then the

following hypotheses should be supported by the data:

The Behavioral Receptional Assimilation Hypothesis

H15; Latinos who have not experienced discrimination or who do not per—
ceive discrimination to be a major problem are more likely to par-
ticipate in politics and will show higher levels of attachment to the

American polity than those who have.

The Context of Reception/ Life is “Better in the US” Hypothesis

H {’2 Latinos who have had favorable contexts of reception and believe
that life is better in the US are more likely to participate in politics
and will show higher levels of attachment to the American polity than

those who have not.

77

Ethnic Residential Concentration

Those who live in states or metropolitan areas with high concentrations of co—ethnics
are more likely to have contacts with or exposure to ethnic media and community orga-
nizations. Having higher concentrations of co—ethnics therefore lowers the per-capita
cost of ethnic mobilization by interest groups, party organizations and candidate orga-
nizations (Ramakrishnan and Espenshade, 2001). At the same time, previous studies
have shown that areas of high ethnic concentration can be areas of lower mobilization
and participation. This may be due to the effect of other social contextual factors
associated with higher Latino populations such as larger proportion of foreign born,
higher residential poverty and greater proportions of non—citizens (Ramakrishnan and
Espenshade, 2001).

There are multiple competing strands of thought as to the likely outcome of res-
idential concentration on the likelihood of political assimilation. One strand of the
ethnic disadvantage model holds that as immigrants desegregate out of their own
neighborhoods and ethnic enclaves in favor of those dominated by the majority group,
they begin to compete directly with natives. The entry of minorities into mainstream
occupations and other roles heightens ethnic awareness (Portes, 1984). Therefore,
ethnicity remains less salient among immigrants who live in areas characterized by
high ethnic populations. They are therefore less likely to become politically assimi-
lated. Both the classical assimilation model and the ethnic competition strand of the
ethnic disadvantage model predict the following political outcomes regarding Latino

acculturation in these contexts:

78

The Ethnic Competition Hypothesis

I117 : Latinos who live in highly concentrated Latino neighborhoods or
in neighborhoods characterized by large foreign born populations will
be less likely to participate in politics and will show lower levels of

attachment to the American polity than those who do not.

Generational Status and Years Lived in the US

In previous studies of Latino political behavior, exposure to US culture has been found
to be an important indicator of political assimilation, especially among foreign born
Latinos. In their study of political participation in California, Uhlaner et al. (1989)
noted that voting among immigrants is inﬂuenced, in part, by their length of stay
in the United States. In their study of Latino electoral participation, Highton and
Burris (2002) similarly found that nativity has a powerful effect on turnout but only
when considered in conjunction with how long foreign born citizens have lived in the
United States. Finally, in a well-cited study, Wong (2000) found a strong relationship
between the number of years an immigrant has lived in the United States and the
acquisition of partisanship. She concluded that a process of reinforcement through
exposure to the political system underlies the development of political attitudes across
diverse immigrant groups.

Here, as elsewhere, exposure to US culture is operationalized by generational
status and years lived in the United States. If classical assimilation theory is accurate,
the expectation is that longer residence in the United States will lead to greater

political assimilation among ﬁrst generation immigrants. Theoretically, immigrants

79

who have lived here longer are more ﬂuent in English and have greater contact with,
and commitment to, American political institutions.

The straightline assimilation model assumes that ﬁrst generation immigrants will
be less politically incorporated than second generation immigrants, who in turn will
be less politically incorporated than their ﬁrst generation counterparts. Similarly,
from an intra—generational perspective, this same model presumes that immigrants
will become more politically incorporated over time. The ethnic and segmented as—
similation perspectives also predict a positive relationship between years lived in the
United States and political behavior.

Classical theories of assimilation considered it to be a unilinear process whereby
the economic and social conditions of individuals and ethnic groups improve over each
succeeding generation. Consequently, for each subgroup of Latinos, the expectation is
that ﬁrst generation immigrants will be the least likely to vote because of linguistic and
cultural barriers that make it difﬁcult to access political information. First generation
immigrants are also more likely to have strong ties to their home country. According
to the straightline model, we should expect the following hypotheses to be supported

by the data:

The Generation Hypothesis

H {3: Third generation immigrants will be more likely to participate in
politics and will show higher levels of attachment to the American
polity than second generation immigrants who in turn will be more

likely to participate than ﬁrst generation immigrants.

80

The Exposure Hypothesis

Hi): Latino immigrants living in the United States the longest will be
more likely to participate in politics and will show higher levels of
attachment to the American polity than tl‘iose who have lived in the

United States for shorter periods.

All three primary data sets had measures of both generational status and years
lived in the United States. Generational status was recoded to reﬂect consistency
with current literature. Individuals born outside the United States were deemed
“First Generation” immigrants. Puerto Ricans born in Puerto Rico were recoded as
foreign born and therefore considered to be in this category. Individuals who were
born in the United States but who had at least one foreign born parent were coded
as “Second Generation” immigrants. Those born in the United States whose parents
were also born in the United States were coded as “Third Generation” immigrants.
The measure of generational status used here is binary and equal to unity if the
immigrant is in generation 2' = 1, 2, 3 and zero otherwise. Years lived in the US is a

continuous measure ranging from less than 1 year to 81 years.

4.3 The Ethnic Disadvantage Model of Assimila-
tion

Both segmented assimilation and ethnic disadvantage theories present opposing view-

points that suggest Latino political acculturation is derivative of (1) a greater sense

81

of affinity for ethnic-based issues; (2) more robust panethnic identiﬁcation; and (3)
heightened ethnic group consciousness (Greenley, 1971; Glazer and Moynihan, 1963;
Portes, Parker and Cobas, 1980; Portes, 1984; Portes and Bach, 1985; Portes and
Rumbaut, 1990; Jones—Correa and Leal, 1996; Kaufman, 2000) that are requisite for
their political mobilization. These theoretical perspectives argue that an immigrants’
increasing awareness of social inequalities and discrimination results in stronger eth-
nic identiﬁcation and uniﬁcation. If these later perspectives accurately portray how
Latinos adapt to the US political system, ethnic resilience should be an important
factor explaining the development of immigrant political attitudes and behaviors.
By reversing the sign of the coefficient pertaining to a hypothesis of straightline
assimilation we have the coefficient that serves to test a hypothesis of ethnic disad-
vantage (Portes, 1984). Both homeward political concerns and resilient ethnicity are
factors that prevent total assimilation. According to Portes and Rumbaut (1996),
“when immigrant communities ﬁnally turn to domestic issues and the vote, they
tend to mobilize along national rather than class lines” (Portes and Rumbaut, 1996,
p. 125). In fact, the major distinction between classical assimilation theory and
the ethnic disadvantage perspective of immigrant assimilation pertains to the role
that ethnicity and in particular cultural retention plays in promoting the adoption
of American political attitudes and behaviors. If the latter is valid then we should
expect, for example, cultural retention, ethnic or panethnic identiﬁcation and bilin—
gualism to be positively related to political participation in the US. This leads to the

following hypothesis regarding the ethnic disadvantage view of Latino acculturation:

82

The Ethnic Resilience hypothesis

H110: Latino immigrants who retain the cultural values and attitudes of
their country of origin will be more active participants in American
politics and will show higher levels of attachment to the American

polity than those who do not.

Portes and Rumbaut (2001) studied the phenomenon of “reactive ethnicity” as a
theoretical model for understanding reactions to inimical conditions in host societies.
For example, they found that reactive formation processes provided a basis for collec-
tive solidarity and political mobilization among ethnic groups to protect those groups’
interests in response to California’s Proposition 1876. As an explanation they offered
that reactive ethnicity is beneﬁcial when group cohesion and mobilization to confront
a threat are needed, and is ﬁrst and foremost a response to societal racism, either
against the individual or against the ethnic group. Finally, it is most intense among
immigrants with low socio—economic status and contributes heavily to the hardening
of oppositional attitudes. If Portes and Rumbaut (2001) are correct in their assess—
ment of reactive formation processes, we should expect that individuals who identify
with either their panethnic or ethnic origin group (i.e., have reactive ethnic identities)
are more highly politically assimilated. The following research hypothesis should be

supported by the data:

The reactive ethnicity hypothesis

 

6California Proposition 187 was a 1994 ballot initiative designed to deny illegal
immigrants social services, health care, and public education that was passed by
California voters but subsequently overturned by a federal court

83

H111: Latinos who adopt reactive ethnic identities are more likely to par-
ticipate in politics and will show higher levels of attachment to the

American polity than those who do not.

Reactive ethnicity was operationalized as identiﬁcation with one’s ethnic or

panethnic group (see section on Identiﬁcational Assimilation above).

Group-Based Resources

According to social identity and acculturation theories, perceptions of linked fate and
common culture, experiences of ethnic discrimination, and participation in ethnic and
church-based organizations may address various components of ethnic identiﬁcation
and are useful indicators for understanding the ethnic identity choices among certain
minority groups. Prior research in political science has identiﬁed these types of re~
sources as intangible attachments that motivate and shape participatory behaviors
(Baretto, 2003).

Group identiﬁcation and consciousness have been measured in a number of dif-
ferent ways in the research literature that are worthy of discussion. Tate (1993)
measured group identiﬁcation with two items taken from the 1984 National Black
Election Study (NBES): whether the respondent feels affected by what happens to
blacks in the US and the degree to which blacks think consciously about being black.
In a different study, Verba, Schlozman and Brady (1995) combined multiple items
designed to measure how feelings of group closeness, the perception of common prob-
lems, experience with discrimination and support for government programs impact

political participation. In her study of African American and Latino involvement

84

in Chicago area politics, Marschall (2001) used a combined measure of low political
trust/ high political efﬁcacy as a proxy for group consci01.1sness. Lien (2001) on the
other hand, included separate measures of group identity and consciousness. She
operationalized “group identiﬁcation” by one’s membership and concern about one’s
own ethnic group and “group consciousness” by the individual’s perception that one’s
own ethnic group has fewer opportunities than most Americans and by personal ex-
perience with discrimination.

If group based resources effect political adaptation, as proponents of ethnic dis-
advantage would suggest, then the following hypothesis should be supported by the

data:

The Group Consciousness Hypothesis

H 112: Individuals with heightened levels of ethnic group consciousness and
cultural retention are more likely to participate in American politics
and will show higher levels of attachment to the American polity than

those who do not.

The survey items used in this dissertation allowed me to build measures of two
broad forms of identiﬁcation: (a) a group identity based on acceptance of a common
fate with other Latinos; and (b) a more exclusive identity based on identiﬁcation with
one’s own ethnic group. Those who endorse the notion of a common fate are conscious
of sharing a panethnic identity and common interest with other Latinos. In contrast,
a lack of common interest is reﬂected in the tendency to think of one’s interests as

being distinct from that of other Latinos. To measure this, in 2002, respondents were

85

asked whether they believed that Hispanics / Latinos from different countries (1) share
one Hispanic/ Latino culture; or (2) have separate and distinct. cultures.

According to Chong and Rogers (2003), group consciousness augments group iden-
tiﬁcation by articulating collective discontents and strategies for improving the status
of Latinos. I examine two components of group consciousness: (a) a belief that group
disparities are produced by discrimination and are illegitimate; and (b) belief in the
political efficacy of group action. In addition to the measures of discrimination dis-
cussed above, the national surveys asked respondents whether they believed that
all Latinos are working together politically to achieve common goals and whether
the respondent believed they would be better off if Latino groups worked together
politically.

The theoretical predictions of each model are summarized in Table 4.1.

86

:oSeEOCEQ Reason :0 $33er iopscmaxm mo muoemm woﬁmoﬁoaa Ad oEsH.

+ +
+ +
IT I
IT I
+ I
+ I
[T I
+ I
+ I
- +
+ I
+ +
+ l
+ +
+ +
+ +
- +
- +
- +
+ I
- +
+ I
+ +

wmgegwemaQ 853m :BNENEwmmV E3330
momuzom m3. 5 somuwmmomaamm "20602 woummonuoahm 95.

25th
$5st 3332
mm<
mmzﬂuﬂr 35:00
m3 2: E uoﬁom
5:358:85 3:8qu
coswﬁﬁﬂoma vowmmdzohw
somewﬁmammmaa mwaomuamomm ~mhom>maom>wﬁvsﬁﬁ<
mmmamsomomcoo Rompsom
A8330 29 sea see:
3553 EmEO 3:033 Z
3553 2:53am
mmonmsomomnoO was humanovH 9.qu
3.85m o... “”5883 mo 053m
Eon 2328 958mm
:wﬁﬁo “smegma
0:53-00 9:8th
“58:033. 2quon
~838st
@885
:omumzemmm< Euseozpum
vcﬂmao: on: ma m3
@553 aeoEmE<
:omumzﬁmmﬂw ﬁmcomuwomﬁcmwm
EeEEoQ 3quan
3:35
wondEEoQ smzmcm
Emzsqowﬁweﬁ.
cosgmsow vacuum
:23thme pmhm
men—Sm USED 05 3 Bid 28>
:omummaﬁmma‘ Ease—:0 van... :omawusﬂsoo<

sees:

87

4.4 The Multiple Pathways of Assimilation

Classical assimilation theory articulates a single path of assimilation down which it
is assumed that all immigrants eventually travel. If this is true, then the following

hypothesis should be supported by the data:

The Similarity Hypothesis

H113: The unique life histories and life experiences of Mexicans, Puerto
Ricans and Cubans are irrelevant in determining the assimilation tra-
jectory they will follow and hence the path to political assimilation is

the same among subgroups of Latinos.

4.5 The Components of Political Assimilation:

The Dependent Variables

4.5.1 Extending the Assimilation Framework to Political
Outcomes

Dahl (1962), posited a model of immigrant political incorporation, describing it as a
function of mobilization in conventional party politics. Dahl views political incorpo-
ration in terms of a three-stage model of ethnic incorporation that closely parallels
classical theories of assimilation. In the ﬁrst stage of incorporation, immigrants are
viewed as unsophisticated political actors, who remain politically uninvolved because

they lack both knowledge and resources. Immigrants gradually learn how to mobilize

88

in party politics, a phenomenon that demarcates the beginning of the second stage
of his model. Immigrants are thought to possess a recognizable and mobilizable eth-
nic identity that provides sufficient. resources to be able to negotiate their collective
and individual interests. The third and ﬁnal stage is characterized by diverse eth-
nic interests which make brokerage no longer viable. Lower levels of discrimination
coupled with higher levels of education and wealth enable individuals to gain access
into mainstream social, economic and political institutions. He believed that ethnic
identity would become less politically relevant over time. Scholars have increasingly
come to doubt whether Dahl’s model of incorporation, which may have “worked”
for European immigrants, appropriately characterizes the assimilation trajectory of
post-1965 immigrants and if not, why?

Today political incorporation refers to levels of inﬂuence over political decision-
making via descriptive representation (Browning, Marshall and Tabb, 1984). For
example, Browning et al. (1984) measured the degree of incorporation using a com—
posite scale of the number or percentage of African American ofﬁceholders present in
the policy-making arena, their presence in important leadership positions, and their
active participation in dominant ruling coalitions. This implies that the degree of po-
litical incorporation is a function of the strength and ability of a group to effectuate
its policy preferences (Browning, Marshall, and Tabb, 1984). It is generally assumed
that high levels of incorporation are prerequisites to the actualization of political goals
and objectives (Haynie, 2001). Scholars in the ﬁeld of ethnic politics have either de-
ﬁned political incorporation as one type of political activity to the exclusion of others

or have deﬁned it too narrowly (Ong and Meyer, 2004) ignoring in the process issues

89

of identity and group interests (Andersen and Cohen, 2002).

As of late, there is a trend among political scientists to reconceptualize past no—
tions of political incorporation (Ong, 1999; Wong, 2002). One issue that has received
increasing attention is that immigrants’ political issues may be oriented towards their
country of origin and therefore do not reﬂect an immigrant’s proclivity towards po-
litical incorporation. Cubans, for example, have lobbied US. ofﬁcials to further their
efforts to denigrate the Castro regime (Ong and Meyer, 2004). There is a need, then,
to distinguish political activity pertaining to one’s home country from participation
based on domestic concerns whenever possible. Besides homeward looking concerns,
a theory of Latino political incorporation must include ethnic resiliency. The roots of
a pan-Latino movement built on ethnic pride and resiliency may be gaining a strong-
hold in this nation. On May 1, 2006, illegal immigrants protested across the country
to take part in “Un Dia Sin Inmigrantes” as a reaction to immigration laws proposed
by the federal government to restrict illegal immigration. Theories of political in-
corporation pertaining to new immigrants should incorporate this potentially newly
formed collective identity, ethnic collective interests and concern for their community.

Political participation is an important way for immigrants to assimilate into Amer-
ican life. Studies of political assimilation7 have focused primarily on the determinants
of naturalization and differences in citizenship acquisition across different national
origin groups (Ramakrishnan and Espenshade, 2001; Portes and Mozo, 1995; Lian,

1994). Voting and other forms of participation that occur after naturalizing have

 

7While I do not use the terms “incorporation” and “assimilation” interchangeably
here, the majority of other scholars do.

90

received very little attention (Ramakrishnan and Espenshade, 2001) despite the fact
that beliefs about the proper relationship between citizens and their government are
central features of American political culture. Additionally, since thirty-nine percent
of Latinos are ineligible to vote on account of their status as non-citizens (de La
Garza, 2004), it is important to have a broad deﬁnition of political incorporation
that includes nonelectoral activity (Wong, 2002), in addition to citizenship acquisi-
tion, party identiﬁcation and voting. In this vein, Wong (2002) challenges researchers
to move beyond these traditional notions of incorporation in order to provide a more
adequate conceptual framework for assessing political incorporation among immi-
grants. The main limitations of existing theories, according to Wong, are twofold.
First, incorporation among new immigrants is not necessarily linear and permanent;
second, political incorporation means more than empowerment in the Marshall/ Tabb
sense.

In this dissertation, I deﬁne political incorporation broadly enough to encompass
more than empowerment via elected officials. To do so, I expand the concept to
include (1) the acquisition of political attitudes and values, (2) levels of political par-
ticipation, and (3) naturalization. The motivation for having a broad deﬁnition of
incorporation comes from the fact that public opinion, participation and representa-
tion are all ﬂip sides of the same coin. For example, participatory integration and
representative integration are not mutually exclusive, rather the former is a precondi-
tion of the latter. Obviously, absent full participation among Hispanics, representa-
tives can remain unresponsive. In addition, democratic responsiveness is a function of

shared and distinct political views. Using the sociological theories presented above as

91

a framework for understanding the political incorporation of Latinos, I discuss how
immigrants relate to the American political system. I emphasize the role of iden-
tity and context in mediating this relationship and in fostering the development of

political attitudes and behaviors.

4.5.2 Electoral Activity

Participation in electoral politics, in the form of registering to vote and actually
voting, is the most obvious indicator of political assimilation.
The following items were asked of survey respondents in the 2002 Latino National

Survey:

1. Have you ever voted in a United States Election?

Two additional questions were asked of survey respondents in 2004:

1. We often ﬁnd out that a lot. of people aren’t able to vote because they were not
registered, or they were sick, or they just didn’t have time. Do you remember
for sure whether you voted in the November 2002 congressional election in your

district?

2. Do you remember for sure whether you voted in the November 2000 presidential

election when George W. Bush ran against Al Gore and Ralph Nader?

In the case of voting, the universe is citizens who are registered voters. In the
case of registration, the universe is citizens of the United States. Possible responses

included “Yes Voted” and “No, did not vote.” Unless otherwise indicated, a binary

92

variable equal to unity was included in the analysis to indicate that the respondent

in fact voted in any of these elections.

4.5.3 N onelectoral Participation

A second type of dependent variable measures the level to which an individual is
involved in participation in politics beyond voting. The following question was asked
of respondents in the 2004 National Survey: People express their opinions about
politics and current events in a number of ways. In the United States, in the past

year, have you:

1. Attended a public meeting or demonstration in the community in which you

live.
2. Contacted an elected ofﬁcial
3. Contributed money to a candidate running for public ofﬁce
4. Attended a political party meeting or function
5. Worked as a volunteer or for pay for a political candidate

Possible responses included “Yes” and “No.” These individual items were summed
to form a scale. The variable was considered an ordinal variable indicating the fre-
quency with which the respondent participated in nonelectoral types of activities.
A value of 0, for example, corresponds to “No Participation” and a value of 5 cor-
responds to “Frequent Participator.” Therefore, this variable does not distinguish

between each of the possible types of participation, but rather measures the quantity

93

of the involvement. Additional analyses were done on these variables that preserved
the speciﬁc type of participation. Those analyses required each variable to be coded

as 1 if the behavior occurred and 2 if it (lid not occur.

4.5.4 Citizenship

United States citizenship is an important indicator of political assimilation among
immigrant populations because citizenship is a prerequisite for full participation in
American politics. Therefore, a model was estimated which had “time to citizenship
acquisition” as its dependent variable. In 2002, ﬁrst generation immigrants were
asked how many years they have lived in the United States. These respondents
were additionally asked to provide information about the year they became a citizen.
Subtracting these two values yielded a positive count variable that indicated length of
time it took to acquire citizenship status. The universe was foreign born immigrants

only.

4.5.5 Political Culture and American National Identity
American Liberalism and Civic Incorporation

Civic incorporation is deﬁned as support for economic individualism. Most researchers
deﬁne economic individualism as a set of “core values encompassing the dual beliefs
that” (de la Garza et al., 1996, p. 341) central to American political culture is the
notion that 1) individuals should be self-reliant, meaning that they are responsible

for providing their own basic needs (de Tocqueville, 1945; Fuchs, 1990; Lipset, 1964;

94

Huntington, 1981; Sniderman and Hagan, 1985; McClosky and Zaller, 1984; de la
Garza et al., 1996); and 2) people should get ahead on their own through hard work
(Feldman, 1988). Economic individualism is a core element in accounts of American
values and beliefs (Feldman, 1988) as evidenced by the widespread belief in the work
ethic apparent in mass opinion surveys and in-depth interviews (Sennet and Cobb,
1972; Lamb, 1974; Feagin, 1975; Lewis, 1978; Sniderman and Brody, 1977; Feldman,
1983; McClosky and Zaller, 1984). It is noteworthy that researchers have failed to ﬁnd
any significant differences between Mexican-Americans and whites in their acceptance
of and support for economic individualism (de la Garza et al., 1996).

To measure support for economic individualism, all respondents were asked to

indicate which of the following statements comes closer to their own view?

1. The government in Washington should do everything possible to improve the

standard of living of all Americans. OR

2. This is not the government’s responsibility; each person should take care of

themselves.

3. Neither.

Importantly, the fact that neither was an Option allows for the possibility that some
other entity apart from the individual or the government, should be held responsible
for individual welfare.

The extent to which Latinos support the major ideals and premises behind the
American work ethic is yet another measure used to operationalize civic incorpora-

tion. Respondents were asked to indicate their level of agreement with the following

95

questions: “Hard work offers LITTLE guarantee of success” and “It doesn’t do any
good to plan for the future because you don’t have control over it.” Possible responses
included 4=Agree strongly; 3==Agree somewhat; 2=Disagree somewhat; and 1=Dis—
agree strongly. Responses were recoded as 1 = Agree strongly/Agree somewhat; and

2 = Disagree somewhat/Disagree strongly.

Free Enterprise

Americans have always given strong support to the free enterprise system (Lipset,
1979; McClosky and Zaller, 1984). The free enterprise system can be seen as the
economic side of the individualistic social system, and support for the free enterprise
system has typically been accompanied by a distrust of big government (Devine,
1972). McClosky and Zaller (1984) argued that support for capitalism and free en-
terprise forms one of the basic elements of political culture. The following item was
designed to measure support for free enterprise: Would you say you favor a (smaller
federal government with fewer services), or (a larger government with many services)?

Possible choices included 1=Smaller; and 2 2 Larger.

Melting Pot

In the American context, the society envisioned by hard multiculturalism is gener-
ally contrasted with the symbolism embodied in the “melting pot” metaphor. This
popular phrase projects the image and process by which diverse cultures come to
shed their heritage in favor of a single American identity. To measure support for

the idea of the United States as a melting pot, the following question was asked of

96

survey respondents: How important is it for Latinos to change so that they blend
into the larger society as in the idea of a melting pot? Possible responses included:
1: Very important, 2 2 Somewhat important, 3=N0t too important and 4: Not at
all important. Responses were recoded 1 2 Very important/Somewhat important and

2 = Not too important/Not important at all.

National Identity

De la Garza (2005) argued that an additional measure of the linkage between Latino
values and the “national interest” is the difference between Latino perspectives of
Latin America versus the United States. Two items pertaining to the relative beneﬁt
of life in the US versus one’s country of origin were included in the analysis. In 2002,
respondents were asked the following question: “Do you feel conﬁdent that life for:
your children will be better than it has been for you, or don’t you feel this way?”
Possible choices included 1 = Life will be better and 2 = No, don’t feel this way.
Additionally, respondents were asked, “Do you think there is more opportunity to
get ahead in the United States, more in the country you came from, or is it about
the same?” Possible choices included 1 = More in the US; 2 = More in the country
you came from; and 3 2 Same.

To explore the meaning and value of identiﬁcation as an American, an additional
ﬁve items were selected from the 2004 National Survey of Latinos: Politics and Civic
Participation to measure Latinos views regarding American national and political
identity: In order for an immigrant to say that they are part of American society, do

they have to do each of the following, or not?

97

1. Do immigrants have to Speak English to say they are part of American society,

or not? (1 = Yes, have to; 2 = No, do not have to);

2. Do immigrants have to believe in the US. constitution, to say they are part of

American society, or not? (1 = Yes, have to; 2 = No, do not have to);

3. Do immigrants have to Be a US. Citizen or not? (1 = Yes, have to; 2 = No,

do not have to);

4. Do immigrants have to Vote in US. elections to say they are part of American

society, or not? (1 = Yes, have to; 2 = No, do not have to);

5. Which comes closer to your views? :

a. The United States has a single core Anglo—Protestant culture; or

b. The United States is made up of many cultures

These ﬁve questions provide the basis for the discussion of ethnocultural Ameri-

canism versus multiculturalism that is discussed in Chapter 5.

4.5.6 Control Variables

The analysis included several control variables. First, age is included to mitigate any
confounding effect with years lived in the United States and also because research
has shown that older individuals participate in politics at higher levels. Gender
differences in participation rates highlight the need to include a binary variable equal

to unity if the respondent is female. Finally, some studies have found that married

98

individuals are more active in politics than unmarried individuals, and therefore a
dummy variable was included in the analyses. In some cases, differences among
ethnic groups were estimated by allowing slope coefficients to differ among Mexicans,

Puerto Ricans and Cubans.

4.6 An Empirical Model of Political Incorporation

Berry’s (1990, 1997) model of acculturation is a useful starting point for understand-
ing variation in Latinos’ identification with the American polity. Similar to the two—
dimensional model of acculturation (Phinney, 2001), identiﬁcation with the American
polity can be viewed as a two-step process. The extent to which an immigrant’s po—
litical activities are ethnically based on the one hand, and the extent to which they
adopt American political attitudes and behaviors on the other, can be considered as
two dimensions of their incorporation that vary independently (Table 4.2). Accord-
ing to Jones-Correa (1998), immigrants choose to maintain their positions outside the
traditional political sphere as a way of balancing the demands by the host country
and the country of origin. He described this process as the practice of the “politics of
in—between,” whereby the political participation is negotiated on the immigrants’ own
terms, which is done by keeping some distance and ties to both polities. Individual
political attitudes and behaviors lie on a spectrum ranging from political alienation
to ethnically based attitudes and behaviors to purely American attitudes and behav—
iors. A purely ethnic orientation includes being more concerned about government

and politics in one’s country of origin and participating only for ethnic based causes

99

or issues. Political engagement in American politics includes voting in US elections
exclusively, participating in nonelectoral politics for reasons unrelated to ethnicity,
and identifying with a US political party. In between are activities such as voting in
both one’s country of origin and in US elections and dual-citizenship. Individuals who
lie towards the “ethnic” side of the spectrum are politically separated, whereas indi-
viduals who lie towards the “American” side are politically assimilated. Individuals
who are “in-between” are expected to be politically integrated. Finally, individuals
who are not concerned with politics in either country are politically alienated. This
presents a broad theoretical view of possible incorporation categories that may be
evident among immigrants.

Table 4.2: Possible Modes of Political Incorporation

 

Participation in Home Country Politics

 

 

High Low
Participation in High Politically Integrated Politically Assimilated
US Politics Low Politically Separated Politically Alienated

 

Political incorporation is deﬁned as “the development of the capacity to mobilize
effective political action in response to perceived political opportunities in a host coun-
try” (Ong and Meyer, 2004, p. 4; emphasis in original) irrespective of the nature of
the action. To predict actual placement in the four cells is dependent on knowing the

characteristics of immigrant groups and the context of reception (Table 4.2).

100

 

x83? 28 smotoﬁ< ﬂows 8
A823, 28 583an Boss 833 5035903ch Laoszom swam:
8 23:3 .moﬂsamumﬁma Mo ﬂog/2 .QEEmQEmE macawuso mo mmosousas
Bod 2.363.895 broad? no women 23:5 08an mo

23.8.2: 2062?:in BSBEEQ cosgsmcgmu was 85:62 085m

.Amogg Eco
sonmE< mo so???“
was :onQCSEQ Eggnog
:wmmv AMEN—«$282354.

3 252 compamﬁmmm/x

 

A83? p.80 csoEoE< Hope 3
236% was :oﬁwmmomﬁsm #85209
I :wmmv .QEanEoE 922950 MO
838st no woman 55:62
was mmocoumas 253m

mcoﬁmmo: :ossmmﬁoo £53m—

.Amoig whoa smotmad. ﬂows

3 232$ was :oSwQBSBQ
Bod .mmocmusam 253m.
mmcﬁooﬁ 3:on 3:95:23

5:» 83:8 go 435

 

”stem mszsa

A853, 28
525:5. mo :osmovm
95 :oEwQBSEQ EUSSOQ
away mosesmqsotm8<
3 mes“: comaszammma‘.

Hasa

comaszszmﬁx

peucmﬁwom

5023.
wwsﬁsgpsma 225m

:osﬂmammma‘

oczﬁwmwbm

$2383.29 :oﬁazﬁwmmaw modal .Hom $80850 38*: was EmEdAooE R960 ”we 28%.

101

Table 4.3 shows the three potential pathways (Path 1, Path 2 and Path 3) of im-
migrant political incorporation and their causal mechanisms. The straightline model
suggests that all immigrants travel only one path; therefore, no other path corre—
sponding to this theory is shown in Table 4.3. According to the straightline model,
assimilation leads to Americanization which in turn implies high levels of political
participation and the adOption of American core values such as American national
identity. The ethnic disadvantage theory presupposes two pathways of assimilation
(Path 1 and Path 2). Immigrants who follow Path 1 lack contact with dominant
groups in society, for example, whites; therefore, they do not possess the group-based
resources that mobilize them to participate in the American polity. These immigrants
are characterized by low levels of participation and do not adopt American core val-
ues. Immigrants who follow Path 2, however, possess heightened ethnic awareness of
their membership in an outgroup and participate in politics as a means of breaking
down structural barriers to inequality in the United States. The segmented model
assumes three possible pathways of assimilation. Path one is the same path that im-
migrants travel under straightline theory. Path 2 is characterized by ethnic resilience
and the reconstitution of ethnic culture based on awareness of outgroup membership,
and is the same as Path 2 of the ethnic disadvantage theory. There is an additional
assimilatory path, Path 3, in which political incorporation is characterized by low
levels of participation and the failure to adopt American core values, similar to the
“minority underclass” living in the United States. In Table 4.3, the columns reﬂect
the fact that America is a stratiﬁed society and that immigrants reside in a variety

of community contexts.

102

Each path is assumed to be conditional on the community context in which the
immigrant resides. More speciﬁcally, the general models of assimilation and incor~
poration play out differently for different immigrant groups, as a function of timing,
context and culture. In the chapters that follow, I focus on Mexican, Puerto Rican
and Cuban immigrants to examine how adequately these theories capture their ex-
periences in the United States. In many of the models presented, the assumption is
that immigrants have a latent propensity to participate in political life. The task is

to understand how identity and context of reception inﬂuence this latent propensity.

4.7 Statistical Methodology

A variety of standard multivariate analyses using multiple regression techniques is
appropriate for all of the research questions I proposed above. The nature of the
dependent variable dictates the appropriate model to use in each case. For example,
when the variable is ordinal, ordinal logistic regression is used. Treating the measure
as continuous and using any other method would result in bias and provide inconsis—
tent estimators of the effect of each independent variable on the dependent variable.

I elaborate on this below, in the appropriate subsection.

4.7 .1 Binary, Ordinal and Multinomial Logistic Regression
Binary Logit Models

Logistic regression is a widely used technique in political science. The binary nature

of the voting models suggests the use of logistic regression where success is deﬁned as

103

voting and failure is deﬁned as non—voting. The logistic: transformation is interpreted
as the logarithm of success versus failure. The logistic transformation of the success

probability 7r is given by

 

logit(7r)=log( 7r ).

1-7r

The above equation is then used as a link function to obtain the logit model:

K
71'
109 (I?) = 7h? = Zﬂkicz'k-
k=0

Logistic regression models the logit of the odds of voting, which is not intuitively
appealing. Therefore, I often convert the log odds to a probability to ease interpreta-
tion. Several times throughout this dissertation, for example, I estimate the following

probability:

K
€$Pf2k=o ﬁkxik)

' = = AW)
2 2 :

 

where

( 1,): 62310072“)
i l1+5$Pf7li)l.

 

For example, the model used to estimate registration and voting in US elections
is
logit[7r(:c)] = a + ﬁlxl + X32332 + - - - + +ﬁprrp,

alternatively

104

_ 61717081531 + 52172 + ‘ ‘ ‘ 'l‘ +r8p17p)
1+ 8:13p(/31$1+ X32332 + - ° - + ﬂpasp),

 

where 7r(a:) = P(Y :2 1) and a: = (51:1,...,:1:p) each 17,-,2' = 1,...,p represents a
predictor variable. Inferences are made based on the estimated parameter ﬂi, which
is the effect of (1:,- on the log odds that Y = 1 after controlling for the other 233.3 in the
model.

This model is useful for describing the odds of one outcome (“success”) versus
another (“failure”). Odds-ratios are a useful way to describe and discuss the impact of
an independent variable on the dependent variable. To illustrate the transformation
of beta coefﬁcients to odds ratios, suppose there are two groups of binary data,
with probability of success p1, and 102, respectively. Let logit(7r1) = {30 + ’7 and

logit(7r2) = B0, where ﬁg and 7 are parameter estimates. The odds ratio of success

versus failure is given by

This quantity expresses the odds of success for the ﬁrst group relative to the
second. The odds ratio is used to explore the strength of association between the
independent and outcome variables. Odds ratios are strictly positive. If the odds
ratio is greater than one, the odds are increased; if the odds ratio is less than one,
the odds decrease. Finally, if the odds ratio is equal to one, the odds of one outcome
compared to another are the same.

Throughout this dissertation, I computed the percent change in the odds. The

105

percent change in the odds for a (5 unit change in :rk, holding other variables constant

is calculated by:

A($,.’I‘k + 6) — A(:I:, ark)

100
X Mining)

 

= 100 X [ezrp(,l3k x 6) — 1].

It is useful to consider the extent to which change in one variable affects Pr(y = 1).
In this dissertation, I do so by allowing one variable to vary from its minimum to
its maximum, while the other variables remain ﬁxed, usually at their means. Sym-
bolically, Pr(y = 1|T, xk) is the probability computed when all variables except :rk
equal some speciﬁed variable. If Pr(y = 1|3,min(:rk)) is the probability when :ck is
equal to its minimum, and Pr(y = 1|E,maa:(:1:k)) is the probability when :rk is equal
to its maximum, then Pr(y = 1|E,ma:r(xk)) -- Pr(y = 1|E,min(:ck)) is the predicted
change in the probability when ark changes over its range.

Oftentimes, too, it is useful to consider the effect of certain combinations of vari-
ables while the remaining variables are held constant. For example, let :30 contain
the mean of all variables except let 11:1 = 2:, a: = (0, 1) and let an vary over its range.

Then

157" = (y =1|~T21$1l= ©0803)-

This strategy was used to compute predicted probabilities that correspond to ideal

types in the pOpulation.

106

Logistic models as Latent Variable Models

The binary response model can be extended to consider a latent variable approach
to estimation. Since the notion of a latent variable will be discussed throughout this
dissertation, it is essential to understand what the term means and the context in
which it is used. Consider an individual’s voting behavior as the observed y. The
variable can only be observed in two states: the person either voted or did not vote.
However, not all persons vote with the same level of certainty. One person may have
been more “skeptical” in his decision to vote whereas another person may have been
very ﬁrm in his decision. In both cases, we observe empirically the same y = 1. The
idea of a latent y* is that there is an underlying propensity to vote that generates the
observed state. Although we do not directly observe y*, at some point a change in
ya: results empirically in the observation that the individual voted. For example, as
individuals become increasing assimilated into American society, it is reasonable that
the propensity to vote would increase as well. At some point, the propensity crosses
a threshold that leads to participation.

For estimation purposes, let y; be a linear function of the 1'22]: plus some residual
error, 62', as in

K
y; = miﬂ + 62: = Z 55% + 62'-
k=0

The latent variable y* is linked to the observed dichotomous variable by the fol-

lowing measurement equation:

0ifyg‘g'r
yi= , *
llfyi>T

107

If y* S 7’, then y = 0. If y* crosses the threshold then y = 1.

The nature of some of the models suggests the presence of an interaction effect, for
example between perceived discrimination and generational status. A methodology
called inteﬁS is used here to estimate the interaction effect in nonlinear models.
The magnitude of the interaction effect is computed by taking the cross derivative
of the expected value of the dependent variable for each model. Consequently, any
assessment of statistical signiﬁcance of the interaction term is based on the estimated

cross partial derivative and not on the coefficient from the interaction term (Chunrong

and Norton, 2001).

4.7.2 Ordinal Logit Models: Proportional Odds Models

Ordinal variables are ranked such that the categories they contain lie on a continuum
ranging from high to low. Ordinal logistic regression is the most appropriate statistical
model to use when the dependent variable is ranked from high to low. Nonelectoral
participation is graded on a scale that ranges from “high” to “low” and therefore
ordinal regression incorporates the nature of the dependent variable.

Ordinal variables are characterized by unknown distances between adjacent cate-
gories. The methodological preference of using statistical techniques designed to han-
dle situations where an ordinal variable is ranked from low to high but the distance
between adjacent categories is unknown has several advantages. First, by preserving

the dependent variable in its current form we do not need to infer a lower bound

 

8At the time of this writing, this estimation technique is available only for inter-
actions between two continuous variables

108

classiﬁcation cut—off point for estimation purposes. Second, treating the outcome
measure as if it were interval involves the implicit assumption that the intervals be-
tween adjacent categories are equal (Long, 1997). Finally, estimating simultaneous
equations using prOportional odds models takes advantage of the natural ordering of
the outcome variable, hence remaining “faithful” to the nature of the response.

The cumulative probability of the ordered logit model is written as:

expo. + mm)
1 + exp(aj + (Diﬂ)

 

P7“(yz' > jlfBi) =

Therefore, the parameterization used here is equivalent to that associated with
regression-type estimates. In other words, for ﬁk > 0, higher values of :L‘k are as-
sociated with higher levels of y.

For the ordered logit model, the odds of an outcome being less than or equal to

m versus being greater than m have the following equation:

_ de S mlw) _

Am(m) —— PM}! > mlm) — exp(7'm — 33,8)

 

In the ordinal regression model, a latent variable, y* ranging from —00 to +00
is mapped onto an observed variable y. Based on this relationship, a latent variable
model for y* can be written as a cumulative probability model. The measurement

equation is given by:

92' = m if Tm_1 S y; < Tm for m = 1, ..., J

where the T’s are called thresholds or cutpoints. In the population, y* is equal to:

109

A.
21* = Zﬂkai‘sz' = Zi+€i
k=1

of which the ordinal regression model estimates part. For a model with K 0s,

A! -— 1 5’5 are computed by:

K
Zr = Z ﬁkxkz'
k=1

and the following threshold parameters are estimated:

lif T0=—00Sy;<7'1

2if71£y;<72

y2_ 3if ngyg<r3

EJif TJ_ISy;<TJ=OO

The main assumption of this model emphasizes an interpretation of the odds
that are the same across categories of the dependent. variable. This assumption,
which has been coined the proportional odds assumption, is often violated in practice.
A Wald test proposed by Brant (1990) allows both an overall test that all of the
coefficients are equal and a test of the equality of coefﬁcients for individual variables.
In the circumstance where this test revealed strong evidence that the assumption was

violated, another estimation technique was preferred (see below).

Multinomial Logit Models

The multinomial logistic model is similar to estimating simultaneous binary logits
for all possible combinations of the dependent variable. This is the most appropri—
ate model to use when the dependent variable is categorical. In this dissertation,

multinomial logistic regression is used to model primary identiﬁcation as a measure

110

of national identity because the levels of this variable are categorical in nature. For ex-

3

ample, the model predicts the probability of identiﬁcation as an “American,’ member
of an ethnic group, or as a “Latino/ Hispanic.”

This model is commonly written as:

€1‘P(93i5m)
1 + 231:2 e$meiﬁjY

Like the binary response model, this model can be parameterized in terms of odds.

m>1

 

Prf'yi = mixi) =

For example the odds of outcome m versus outcome n, given :13, is given by

exam/Gm)

Amln(mi) : €$meiﬁn)

It can be shown that taking the log of this equation yields the following equation

that is useful in computing contrasts for all combinations of dependent variables:

ln(Am|n)$i : xii/3m — ﬁn]-

When the multinomial model is the preferred estimation technique in this disser-
tation, contrasts are computed for every combination of the dependent variable.

Finally, both partial changes and discrete changes were used to summarize the
effects of each variable on the probability of an event occurring. The partial change
is deﬁned as the unit change in the independent variable that causes y to change by
0;, units, holding other variables constant. The discrete change is the change in the
independent variable from one value to another value that causes a change in the

predicted probability of the event occurring.

111

In order to test model ﬁt, I used the likelihood ratio test as follows. The con-
strained model with parameters ,BC was nested in the unconstrained model with
parameters 01]. In each case, the null hypothesis is that the constraints imposed to
create the constrained model are true.

As an alternative parameterization considered here, the multinomial model is
conceptualized as a discrete choice model based on the idea that individuals choose
outcomes that maximize the utility gained from that choice.

To concretize this notion, let the utility for choice 2' be u,. A person chooses utility
2' when u,- > usz' aé m, and the utility derived from choice i for individual j equals

uz-j = uz-j + 6,3. The probability of choosing alternative 2' is

Pr(y = 2') = Pr(u, > uk‘v’i 75 k)

These probabilities are derived from the MN LM.

A crucial assumption of multinomial regression models is the independence of
irrelevant alternatives. This assumption implies that the multinomial logit model
should only be used in cases where the outcome can plausibly be assumed to be dis-
tinct and weighted independently (McFadden, 1973). In this dissertation, a Hausman
test of the IIA property was implemented by estimating the model on a restricted set
of outcomes. The test is based on the assumption that if the other alternatives are
irrelevant in computing subsets of outcomes, then their omission should not affect the

parameter estimates that affect those outcomes.

112

Sequential Logistic Models: Continuation Ratio Logistic Regression

lV‘Iultinomial response problems can be sin‘iplificd by taking into account the temporal
ordering of the responses. This approach takes advantage of the sequential nature of
decision-making. This model is useful for addressing the question of what sustains
participation over time because it models the probability of participating in more
than one political act given participation in at least one.

In this case, the interest is in conditional probabilities in modeling Pr(y,; = j) for

j > 1 given that y, = j — 1 denoted P737. More concretely,

221-4 Pit
J
21:34 Pm

A continuation ratio logit model is then given by:

PJL‘PTWiZle/iZj—l): ,j>1-

Although the variable can be considered ordinal, proportional odds models are
restrictive in not allowing separate structural mechanisms across different categories.
If different mechanisms are at work for different levels of participatory transitions (i.e.
going from none to one) the process can be broken down into a sequence of transitions.
In this case, conditional probabilities model the probability that y,- 2 j for j > 1,
given that y, = j -— 1, denoted as P137. As an illustration of this relatively unknown
technique, consider the process of participating in nonelectoral politics, which can be
viewed as a series of binary choices. This is illustrated diagrammatically below.

To estimate this model it is necessary to condition on appropriate subsamples

113

of the data. First, the entire sample is used to model the probability of 1 or more
participatory act (Pl-‘5). In the model for the ﬁrst transition, the response variable is
coded as 1 if the respondent engaged in at least one participatory act, 0 otherwise.
Note that P“ = 1 — P5. The conditional probability of having participated in more
than one act 1323 was modeled next using the continuation logit model using only
the subset of respondents with at least one participatory act. In a model for the
second transition, the response variable is coded 1 if the respondent has more than
one participatory act and 0 otherwise such that

19,2 = P1211 _ P75)

2 ’l

and

,_ + +
1313”Pi2pi3’

which shows that the sequence of binary logit models completely describe the

multinomial process.

4.7 .3 Path Analysis and Structural Equation Modeling

Structural equation modeling techniques were used to asses whether the indicators of
identity, behavioral and attitudinal assimilation mediated the relationship between
socioeconomic status and political participation. Finally, model parameters were
constrained to be equal to test whether assimilation trajectories that shape political

behavior are the same for Mexican and Cuban immigrants.

114

Data for the structural equation and path analysis models were analyzed using
AMOS 6.0 and EQS 1.0 software. These programs use the method of maximum
likelihood estimation in order to obtain parameter estimates. The method is robust
when data are missing (Arbuckle and Worthke, 1999; Byrne, 2002). The overall
ﬁt is generally determined by the chi-square statistic, which tests for comparability
between the proposed and independence model where constructs are assumed to be
unrelated (Bollen, 1989). Because the chi—square statistic is sensitive to large samples,
alternative ﬁt measures have been suggested (Hu and Bentler, 1999). These are the x2
ratio statistic (2&2), which adjusts for sample size, the Comparative Fit Index (CF I)
and the root mean square error of approximation (RMSEA). In accordance with most
researchers, a x2 ratio of between 1 and 3 (Arbuckle and Worthke, 1999)and a value
of CF I 2 0.95 and RMSEA g .05 was taken to indicate an acceptable ﬁt (McDonald
and Ho, 2002). The signiﬁcance level criterion for all statistical tests was a = .05,

unless otherwise noted.

4.7 .4 Latent Class and Logistic Latent Class Analysis
Importance of Latent Class Analysis

Latent class analysis allows researchers to identify typologies based on a set of latent
characteristics or traits. Hagenaars and Halman (1988) explain that basic to ideal
types is the notion that overt behavior can be conceived of as belonging or closeness
to an underlying pure type. Latent class analysis analyzes data from a multiway con-

tingency table describing aggregate responses and decomposes the tabular frequencies

115

into a set of latent classes that display certain characteristics. This type of analysis
identiﬁes a clustering of individuals who display certain traits and is used here to
identify individuals who adopt multicultural or ethnocultural Americanism, Ameri—
can liberalism and electoral responsiveness versus non-responsiveness. Importantly,
it is possible to statistically test for the structure of each underlying trait instead of
imposing the structure on the data. This tends to negate any ambiguity associated
with the number of dimensions associated with a latent construct, such as national
identity. Instead, the number of classes emerges from the analysis.

In addition to obtaining the number of classes associated with a latent construct,
latent class analysis has additional beneﬁts. For example, it is possible to estimate
the prevalence of each class or type. For example, if we identify a highly participa-
tory class of voters, we can estimate the percentage of respondents who actually fall
into this class. The conditional response probabilities generated by the model allow
researchers to assess the likelihood of a particular response given that the respon-
dent has been associated with a particular subtype. This model offers the additional
beneﬁt of reducing several indicators into fewer categories, and in this way resembles
factor analytic techniques.

An exciting new program created by researchers at The Pennsylvania State Uni-
versity predicts the probability that a given member of the population will fall into
one class versus another. By adding covariates to the latent class models, it is possible
to explore how the measures of assimilation, acculturation, group—based resources and
context of reception act to facilitate Latino political assimilation. In this respect, it

is no different than adding predictors to a logistic or regression model to assess their

116

impact on a given dependent variable.

Typically, multivariate logistic regression analysis has provided political scientists
with a typical portrait of voters. This type of analysis has several shortcomings that
are addressed, in part, with latent class analysis.

Logistic regression assumes that predictor variables are exogenous and that the
only endogenous variable is voting behavior. While some variables that enter into
the model can legitimately be considered exogenous (for example, age and marital
status), other variables, such as trust in government and political interest, cannot.
Second, logistic regression assumes that variables are measured without error. This is
highly unlikely to be the case for many variables, especially in surveys with immigrant
respondents. If this error is ignored, seriously biased estimates result. Third, the
presence of multicollinearity potentially masks significant effects due to the presence of
correlations between variables. Fourth, unmeasured heterogeneity is usually ignored
in logistic regression analysis and when it is not, it is often treated as a nuisance
rather than as part of the structural information of the sample.

Latent class analysis attempts to explain the relationships among a set of ob-
served measures by means of discrete but latent groups. Similar to cluster analysis,
latent class analysis is designed to categorize cases into a smaller number of rela-
tively homogenous clusters based on a set of manifest characteristics. In latent class
analysis, however, the classes are based upon a probability model and does not re-
quire an arbitrarily chosen metric. The appropriateness of an LCA model assumes
a statistical basis by comparing the characteristics of the observed data with those

that would have been observed based upon the speciﬁed probabilistic model. Finally,

117

by accounting for the unobserved heterogeneity, LCA accommodates interactions of
observed variables by accounting for the fact that different clusters can be similar
with respect to certain variables but different with respect to others.

The €EM program was used for this analysis because of its ability to deal not
only with manifest variables, but also with latent variables. This makes it possible
to specify factor analytic models for categorical latent variables with categorical in-
dicators (KEM program manual, 2005), which is the deﬁning characteristic of latent

class analysis.

Unrestricted Latent Class Models

The classical formulation of the latent class model is defined as a probability model

(Lazarsfeld and Henry, 1968; Goodman, 1974):

7r:mbcd = 7r$7ra|$7rb|$7rc|177rd|r

where 7r]; is called the latent probability while walx, 7Tb|rv “clan and 7rd“. are called
the conditional response probabilities. The same unrestricted LCM can be formulated

as a log-linear model for the incomplete frequency table m .mbcd- That is,

logmmbcd = u + uf + uf + u)? + u? + 11,11) + ufaA + ube + 11$.ch + ude.

The relationship between the two formulations of the LCM can be illustrated

by writing the conditional probabilities in equation 1 as a function of the log-linear

118

parameters appearing in equation 2, such as,

e:1:p('u{;1 + ‘U.§d4)

all _ Z e.’1;‘.p(u{,4 + ui‘a‘l),

 

71'

which is a saturated logit model for the probability on AIX.
The simultaneous latent structure model is a natural extension of the single group

LCA with G > 1 groups.

7rgacabcd = ﬂx|gwalxgﬂbhgﬂclngdIIga

meaning that for individuals in group 9, the probability of being at level (i, j, k, l, 2:)
on the latent and observed variables is equal to the product of (1) the conditional
probabilities for each of the observed measures for latent class :1: in group 9; and (2)
the conditional latent class probability of being in latent class 2: for members of group

g. Within each group, the conditional response probabilities sum to 1.0.

Restricted Latent Class Models

To test for homogeneity in the latent class conditional response probabilities, equal-
ity restrictions were imposed. Across -group equality constraints on the conditional
probabilities enable the hypothesis of similarity of the groups’ latent structures to be
tested. The simplest type of equality restriction in the context of latent class analysis
is to set these probabilities to be equal to each other across groups. For example, the
following equality constraint restricts the response for category A of class X to be

equal across G = 1, 2, 3.

119

ala‘g _ alrg _ alrg
2:11: — 75:23: — ”1.3.1:

7r

Where applicable, completely homogenous latent structures were tested incremen-
tally, beginning with the heterogeneous T class model and testing partially homoge—
nous restrictions thereafter. The decision to retain the partially homogenous model

was based on the x2 test statistic. In other words, an improvement of ﬁt was deemed

acceptable if the increase did not differ signiﬁcantly from the data.

4.7 .5 Poisson and Negative Binomial Regression

The measure of citizenship used in this dissertation is by nature a count variable,
since it is the difference between an immigrant’s year of entry and year of citizenship
acquisition. This measure, then, provides a “waiting time” to citizenship acquisition.
The Poisson model is the most appropriate model to use when the dependent variable
is countably inﬁnite, such as this. The Poisson and negative binomial regression
models are applicable with count data, where Y 6 {0,1,2,....} Count data can be
seen as the result of an event generating process. If the event rate, u, is constant, the
number of events follows a Poisson distribution, where

euuy

P(Y = y) = —'—,y = 0,1,2,...Vu > 0.
y.
The property of equidispersion characteristic of the Poisson distribution is that
E (u) = Var(u). A regression model is obtained by specifying
,I
W = Efyilmi) = €5FP(K31137:)-

120

Overdispersion requires the use of the negative binomial distribution, which sim-

ilarly models count data. This model adds unobserved heterogeneity by specifying

I
w = 617:003 93062“-

The underlying assumption is that the 6, follow a gamma distribution. Absolute
effects on u are calculated as discrete or multiplicative effects (see above), 63, which
depend on the values of the independent variables.

We are now ready to proceed with the analysis of the contexts and characteris-
tics that deﬁne an irmnigrant’s assimilation into the United States polity. The next
chapter discusses the relevance of ethnic identity and identification in understanding

how immigrants relate to US politics.

121

Chapter 5

The Meaning of Americanism and
National Attachment from an

Immigrants Perspective

It must talk the language of its native-born fellow-citizens, it must pos-
sess American citizenship and American ideals. -— Theodore Roosevelt

Advocates Americanism, 1915

5. 1 Overview

In this chapter, I present the results of a latent class analysis on several categorical
indicators of national identity among Latino subgroups in the United States. In
accordance with previous literature, the latent class analysis revealed an interpretable

solution of Latino national identity into three classes that can be loosely characterized

122

as ethnocultural Americanism, multiculturalism and modiﬁed multiculturalism. The
simultaneous latent structure of national identity among Mexicans, Puerto Ricans and
Cubans further revealed homogeneity across each national origin subgroup, indicating
a substantially similar structure between them. A multinomial latent class regression
was also performed on several covariates related to immigration status and group
identity. The analysis revealed that context of reception and group consciousness
signiﬁcantly predicts membership in each class. A major finding of this chapter is that
Latinos show overwhelming support for the contours of ethnocultural Americanism as
evidenced by the large prevalence of this class. Overall, I found a high level of support
for American values among Latinos an indication that Latinos are assimilating both

politically and culturally, despite popular belief to the contrary.

5.2 Review of Research

The ability of Americans to deﬁne who they are culturally, politically and economi-
cally is important for understanding the dynamics of immigrant incorporation. Some
researchers have even suggested that Latino political assimilation depends crucially
on the ability of Americans to clearly deﬁne the contours of American national iden—
tity (Citrin et al., 2003). The very meaning behind the components of American
identiﬁcation among immigrant populations, especially Latinos, is currently at issue
in political discourse. Samuel Huntington (1982, 1996, 2004), for one, has become
an outspoken critic of the role that immigrants are playing in destroying American

national identity and interests.

123

For example, in 1997 he argued that deﬁning American national interests pre-
supposes agreement on the nature of the country whose interests are to be deﬁned.
National interest, according to Huntington, derives from national identity. We have
to know who we are before we can know what our interests are. In his book ap-
propriately entitled Who Are we? The ChaIIenges to America ’3 National Identity,
Samuel Huntington (2004) reiterated his understanding of the origins and dimen-
sions of American national identity and reasserted that American national identity is
threatened by growth among native born and immigrant populations of Latin Amer-
ican origin (Fraga and Segura, 2006). No sooner did Huntington pose the question,
“Will the United States remain a country with a single national language and a core
Anglo—Protestant culture?” than he provided an answer “[b]y ignoring this question,
Americans acquiesce to their eventual transformation into two peoples with two cul-
tures (Anglo and Hispanic) and two languages (English and Spanish)” In addition to
their preference for cultural retention, Huntington argues that immigrants maintain
strong ties to their home countries, are poorly educated, fail to become citizens, and
develop ethnic enclaves apart from the rest of society.

The notion that Latinos, in particular, Mexican-Americans, support values that
are central to American political culture such as economic self-reliance, limited govern-
ment, and a commitment to speak English has been further undermined by American
stereotypes depicting them as lazy and reliant on governmental services. Moreover,
political pundits have argued that instead of promoting “Americanization” as in the

3

past, today’s dominant political culture encourages “Imilticulturalism,’ with negative

implications for American politics. Yet another basis for claims that post-1965 immi-

124

grants are weakening American national identity is the common belief that today’s
immigrants are moving into a mature welfare state in which government assistance is
aggressively advocated rather than into a society that encourages newcomers to make
it on their own. It is not difﬁcult to conclude that part. of the reason we do not know
“who we are” is due to the fact that newcomers are subversive of US culture because
they are not attitudinally assimilating into American society.

As noted by Fraga and Segura (2006), it is difﬁcult if not impossible to dispute that
Anglo-Protestant culture has played a central role in the development of American
national identity. Ethno-linguistic and religious ties are central aspects of Anglo-
Protestant culture. Since its founding, this nation has worried about its ability to
assimilate immigrants and their ethnic descendents into the political mainstream
(de la Garza et al., 1996). Thomas Jefferson, for example, argued that immigrants
would “bring with them the principles of the government they leave” and would then
infuse them into the polity, making it a “heterogeneous, incoherent, distracted mass”
(Fuchs, 1990, p. 12—13). Surprisingly, however, there is little empirical evidence with
respect to how speciﬁc subcultures in the US feel with regard to these, as well as
other, American values. Scholars have tended to focus either on Americans’ beliefs
regarding what “being American” means (Citrin, 1994) or on how Latinos and other
ethnic groups view their own ethnicity and ethnic identity (Phinney et al., 1996). In
contrast, Latinos’ views regarding what it means to be an American and American
identiﬁcation have been largely overlooked (but see de la Garza et al., 1996).

In response to Huntington’s theoretical claims, researchers Fraga and Segura

(2006) insist that immigrants are not threatening the fabric of society, they are merely

125

transforming it. In fact, the ’threatening versus transforming’ debate regarding Amer-
ican national identity is proving to be a source of cleavage between scholars of ethnic
politics and mainstream political scientists. To what extent do Latinos endorse the
liberal conception of national identity? How much support is there for an exclusion-
ary deﬁnition of Americanism based on ethnicity? In 1997, Citrin and his colleagues
posed these same questions but they were directed towards Anglos. In contrast, in
this dissertation I examine the contours of support for and opposition to ethnocul-
tural Americanism, American liberalism and national identity among Latinos living
in the United States. After reviewing some of the factors that contribute to a sense

of Americanism and national identity, I address the following empirical questions:

1. Are deﬁnitions of political and national identity similar among Latino sub-
groups? If so, then this provides evidence against the “multiple pathways of
assimilation” hypothesis because it implies that the process of identiﬁcation

with the American polity is similar across subgroups.

2. Do Mexican, Puerto Rican and Cuban immigrants believe that in order to
claim membership in the American political community one must endorse more
restrictive notions of American values and attitudes? If so, this is evidence that

Latinos are assimilating into American political culture.

3. Is political identity unidimensional or multidimensional and to what extent are
the indicators of political identity homogenous across Latino subgroups? This
question addresses the components of and strength of attachment to American

attitudes and values.

126

4. What is the likely trend in support for adopting an American identity over
an ethnic identity? The preference for an American identity over an ethnic or
panethnic identity is taken as an indication that Latinos are becoming politically

assimilated, as evidenced by a strong American national identity.

5. How does exposure to American culture and the context of reception change
fundamental attitudes towards Americanism and “being American?” In other
words, do immigrants identify with the American polity more if they are cultur-
ally assimilated and have had favorable experiences and contact with American

society?

A quick statement of the rationale for this chapter is warranted. Since political
assimilation has both attitudinal and behavioral components, any study of immigrant
political assimilation would be lacking without an assessment of the degree to which
they accept core political and American values. If Latinos are not adopting US beliefs,
behavioral manifestations of political assimilation may mean that they are pursuing
objectives that are incompatible with mainstream political culture (de la Garza et
al., 1996).

This chapter is concerned with the attitudinal components of political assimila-
tion while later chapters will be concerned with behavioral manifestations. Taking
the most restrictive deﬁnition of “American creed,” this chapter ﬁrst examines the
extent to which Latinos have internalized American liberal traditions to empirically
determine whether past practices are changing due to the presence of Latinos. Latent

class analysis was performed on several indicators of A111ericanism derived from both

127

the 2004 National Survey of Latinos: Politics and Civic Participation and the 2002
National Survey of Latinos. Turning next to a more objective measure of Ameri—
can identity, namely American self—identiﬁcation, I explore the factors that motivate
identiﬁcation with the American political system and culture. This chapter sets the
tone for future chapters by arguing that conditions in the host country fundamen-
tally transform not only how people view themselves but also the meaning given to
their own identity. Additionally, I show how these processes vary to some extent
by ethnic origin and generational status. What follows also provides a context in
which to understand how assimilation may be facilitated or impeded by exploring
how Latinos identify with the American polity. The main results of this chapter
are that (1) Latinos show overwhelming support for the contours of ethnocultural
Americanism and American values as evidenced by the large prevalence of an “eth-
nocultural American” class; (2) membership in both the ethnocultural Americanism
and the American liberal class changes with sociodemographic characteristics, expo-
sure to US culture, perceptions of group—level discrimination and context of reception;
(3) individual level discrimination has no effect on primary self-identiﬁcation among
either Mexicans or Cubans; (4) perceptions of group-level discrimination signiﬁcantly
increase the likelihood of selecting both panethnic and national origin identities over
an American identity; (5) third generation Latinos are signiﬁcantly more likely to
identify as Americans than are ﬁrst or second generation Latinos, whose identities

are primarily oriented towards their country of origin.

128

5.2.1 American National and Political Identity

While Americans share much in common, the persistence of certain sub—cultural iden-
tiﬁcations suggests that individuals prefer at least some continuity of the values spe-
ciﬁc to their own culture. There is evidence that individuals tend to identify pri-
marily as members of their own ethnic group or subculture than as Americans. For
example, in studies asking people to report nationality, they typically report white,
Euro—American, African-American or Asian-American rather than US or “American”
(Peppas, 2001). But the contours of national identity run much deeper than self-
identiﬁcation. Does identiﬁcation with one’s ethnic group preclude the ability to
identify with American values? In one of the very few empirical studies in this area
(Sidanius, 1997), de la Garza and his collaborators studied the relationship between
patriotism and “ethnic consciousness” using a national survey of Mexican Americans.
The authors found that (1) Mexican Americans were no less patriotic than Euro-
Americans, but rather there was a positive relationship between attachment to one’s
Mexican heritage and attachment to the nation as a whole. They concluded that in
the United States, commitment to ethnicity is not an instrument for the “disuniting
of America” but rather “American ethnics use ethnicity to create resources such as
group solidarity and political organizations to facilitate their full participation into
American society” (p. 337). In sum, the two indicators used in their study strongly
suggest that Mexicans have a heightened awareness of American national identity.
The term “national identity” is deﬁned as the clmracteristics that objectively de-

ﬁne membership in a political community or that underlie a shared sense of distinctive

129

peoplehood. In most countries, national identity is centered on a common language,
religion or ethnic heritage. In the United States, however, there are multiple compo—
nents of national identity. The current debate over national identity stems from these
multiple ideological conceptions of American nationality (Schildkraut, 2002) that have
widespread agreement (de Tocqueville, Lipset, 1964; Williams, 1960; Huntington,
1981; Pole, 1967; Kohn, 1957). The speciﬁc content of national identity was ﬁrst
articulated by Smith (1993) and was later expanded by Schildkraut (2005). Schild-
kraut’s model included “liberalism,’ “ethnoculturalism,” and “incorporationism.” Ir-
respective of the content of national identity, it has a psychological component that
is encapsulated by the strength of emotional attachments to American symbols and

to the institutions and practices that embody them.

Ethnoculturalism Ethnoculturalism has been an important deﬁning element of
American national identity since its inception (Schildkraut, 2005). Unlike liberalism,
the ethnocultural tradition sets strong boundaries on? group membership by main-
taining that American identity is deﬁned by white Protestantism rooted in Northern
European heritage and ancestry (Smith, 1993, 1997). At the core of the ethnocultural
conception of identity (Smith, 1988), is the belief in ascriptive criteria for citizenship,
embodied in the idea that only some racial groups, religions or cultures are truly
“American.” This stereotypical image of an American as a white Christian has been
internalized and shapes the thinking of many people toward non-whites in the United
States. Despite the fact that the primacy of ethnocultural beliefs and practices has

waned, there currently exists a resurgence of ascriptive notions regarding what it

130

means to be an American among individuals who assert that immigrants do not ﬁt
into ethnoculturalism’s ideal types (Schildkraut, 2002). Opinion data still reveal that
many people are willing to delineate American identity along ascriptive lines. For
example, in the 1996 General Social Survey, 55% of the respondents said that being
Christian is either somewhat or very important in making someone a “true Ameri-
can,” and 70% said the same about being born in America. In addition, in a 1988
California poll analyzed by Citrin et al. (2002), 76% of respondents endorsed the

idea that this was “very important” in making someone a “true American” while

61% stated that believing in God is a necessary ingredient.

American Liberalism According to Schildkraut (2005, p. 4), liberalism “is the
image of America that comes most easily to mind when people think about what it
means to be American” and is the deﬁning essence of American political culture. The
liberalist tradition is typically deﬁned in terms of beliefs about universal human rights.
As a philosophy, liberalism places individualism over governmental intervention and
maintains that individuals should be afforded certain constitutional protections such
as freedom and equality. Most authorities, in fact, view democracy, liberty, equal-
ity and individual achievement as essential components of American identity. The
right to privacy translates into a belief system that includes tolerance, privacy and
individual liberties. It also leads to a preference for minimal governmental oversight
and a free market economy. According to liberalist philosophy, American national-
ism is ideological or political in nature, meaning one’s national origin is irrelevant

to claim membership in the American political community. For example, according

131

to proponents of the liberal view of national identity, in order to consider oneself
“an American”, one has only to endorse the national “creed” (Huntington, 1981).
In short, membership in the American community is said to be open to all, irre-
spective of ancestry or background, as long as group members endorse principles of

individualism, equality of opportunity, democracy and strong work ethic.

Multiculturalism Sparked by the civil rights movement and fueled by the in-
ﬂux of immigrants from Latin American and Asia, multiculturalism has emerged to
challenge liberalism as an ideological solution for balancing unity and diversity in
America. Schildkraut (2002) used the term “incorporationism” (p. 515) while oth-
ers have preferred the term “multiculturalism” or “ethnic pluralism” to describe this
immigrant-based conception of American identity.

Multiculturalism is generally deﬁned by ethnic distinctiveness whereby no one
group assumes dominance over any other group in society and where individuals can
simultaneously maintain a positive commitment both to an ethnic particularism and
to the larger political community. Jack Citrin (2001), for example, deﬁned multicul-
turalism as a normative conception of political identity and national community with
a derivative policy agenda. As a political formula, multiculturalism is an ideological
political response which assumes that differences in culture, in the sense of a co—
herent cluster of beliefs, values, habits and Observances, accompany ethnic diversity.
Citrin and his colleagues further clarify the distinction between “hard” and “soft”
dimensions of multiculturalism, which fall on a continuum depending on the extent

to which they are committed to preserving different ways of life within a political

132

system. “Hard” multiculturalism, in particular, is an encompassing ideology main-
taining that the purpose of politics is to afﬁrm group difference. The multicultural
component to American identity is not focused on the similarities between minorities
and whites in terms of their rights to citizenship, but rather it is about promoting
ethnic differences and diversity. This viewpoint is intended as a balance between
complete cultural divisions and the notion that America is a homogenous society in
which all members “look, sound and worship alike (Schildkraut, 2002, p. 515).”
Just as there is support for the more exclusive deﬁnition of American identity,
scholars have also found support for an immigrant-based conception of national iden-
tity that is based on multiculturalism. For example, in Citrin’s analysis of the 1994
General Social Survey, he found that when respondents were asked to place them-
selves on a multicultural continuum, their responses were evenly divided among those
who favored the melting pot position, those who opted for the cultural maintenance
option and those who disagreed with both (Citrin, 2001; Schildkraut, 2002). As Cit-
rin noted, this pattern reveals that “many Americans may not consider maintaining

one’s ethnic heritage and blending into the larger society as mutually exclusive” (p.

260).

133

5.3 Social Identity Theory, Symbolic Politics and

Group Attachment

The conceptual analysis of this chapter ﬁts within the larger framework of social
identity theory and symbolic politics. The later theory focuses on the relationship
between symbolic predispositions towards national identity and political outcomes.
The development of a strong sense of identiﬁcation with and attachment to the nation
is viewed as a mechanism through which people come to understand and navigate the
political world (Schildkraut, 2002). A prerequisite for understanding Latino political
behavior is understanding the meaning given to American symbols and the values
attached to them. Of course, this begs the question of how Latinos come to identify
as Americans from the start.

Social identity theory provides some insight as to a possible mechanism by which
individuals would gain from the internalization of American symbols. W hile ethnic
identity has been conceptualized as a complex construct including a commitment
and sense of belonging to the group, positive evaluation of the group, interest in
and knowledge about the group and involvement in social activities of the group
(Phinney, 1990, 1996), ethnic identity is not the only group identity for members of
immigrant groups, who have the option to identify as Americans. Insofar as group
membership is important to immigrants, identifying as an American contributes to a
sense of belonging. According to Tajfel and Turner (1986), social identity “consists
of those aspects of an individual’s self-image that derive from the social categories

to which he perceives himself as belonging” (p. 16). Some researchers have explored

134

the psychological implications of national identity and pointed out that one’s nation
plays an important role in the psychological task of self-deﬁnition (Scheibe, 1983).
Identity derives from knowledge about one’s self that is sensitive to changes in the
environment, and is vulnerable to inconsistent feedback about the self. The process
of assimilation creates the possibility that immigrants are internalizing valued aspects

of American culture as part of their own identity.

5.3. 1 Hypotheses

Since it is difﬁcult to fault immigrants for not assimilating if they do not understand
what they are assimilating to, the ﬁrst objective of this chapter is to deﬁne what
it is to be a part of American society from an immigrant ’3 point of view. Given
that the literature delineates three primary subtypes of national identity, it is hy-
pothesized that the nature of the construct measuring American identity will be best
described by a three class latent structure. Given the discussion above, it is difﬁcult,
a priori, to articulate a directional hypothesis as to the likely effect of acculturation
on an immigrants understanding of the components of American identity. On the
one hand, the ethnic disadvantage model posits that immigrants bring with them
idealistic views about the nature of American society and that these views become
increasingly cynical over time. The ethnic disadvantage model further predicts that
the latent structures will be heterogeneous across subgroups. On the other hand,
assimilation theory predicts that the pressure to assimilate and the forces of political

and cultural socialization are likely to become internalized and latent affinity towards

135

American norms may become manifest. In contrast to the ethnic disadvantage theory,
assimilation theory predicts that the nature of the latent structure will be homoge-
nous across subgroups. Therefore, a reasonable hypothesis is that discrimination is
inversely related to Americanism but that, in the absence of discrimination, accultur-
ation will have a strong positive relationship with the adoption of American political

and cultural values and norms.

5.4 Results

Table 5.1 summarizes Latinos’ views about what it means to be part of American soci-
ety among various demographic and political groups. As revealed by the tabulations,
there is a high level of consensus among Latinos regarding these particular indicators
of Americanism. There is almost complete agreement, for example, that before claim-
ing membership in the American community, one must believe in the United States
Constitution. The primacy of constitutional principles is a universalistic criteria in
liberal conceptualizations of American nationality, and the overwhelming majority
of Latinos deem such beliefs important as well (82%). Speaking English is a more
restrictive basis of American identity that some believe contradicts egalitarian prin-
ciples embodied in the constitution. Nevertheless, 57% of Latinos believe that to
be part of American society, one must speak English. There is very little variation
between Latino subgroups with respect to speaking English as a criterion for mak-
ing someone an American. Cubans are more likely to endorse this view (66%) while

Mexicans are least likely (55%). Puerto Ricans are the group most likely to believe

136

that US citizenship is an important component of American identity (66%) while
Mexicans are least likely to endorse this view. Cubans (70%) are more likely than
either Mexicans (64%) or Puerto Ricans (68%) to believe that voting in US elections
is an ingredient for making someone an American. As expected, Republicans were
more likely to endorse the indicators of Americanism across all categories, but the
variation tends to stay within 20 percentage points. The college educated and the
relatively young were more likely to endorse these categories of Americanism, which
contradicts Citrin et al.’s (1990) ﬁndings with respect to whites; however those in the
highest income brackets were less likely to endorse these values. Some interesting and
counterintuitive ﬁndings are revealed in the cross classiﬁcation as well. For example,
foreign born Latinos are more likely to endorse notions regarding what it means to be
an American, and this holds across all categories except citizenship. Additionally, the
Spanish dominant are more likely to give affirmative responses than are the English
dominant or bilingual with respect to all measures except citizenship and multicul-
turalism. Nevertheless, Table 5.1 shows that the symbols of American nationality are

generally agreed upon irreSpective of ethnic origin and sociodemographic status.

137

 

om 3 S as as 8 B 80% 8893.

so. 0 at. at mm 3 m2 32920 as=00

om S 8 mm. mm as mom m.0260 seem

a w E 8 mm mm. ”8 as .8 as 328 as:
GOmuNUQUm

no. w a. E .8 we as +3

8 a 8 as a a 0%.. $3

mm 2 E as as as as 3.8

mm 2 on a . E we as $3:
0w<

a a as 8 mm 8 Br. 252:5 seem
8:80:82

a a E 3 ma 8 E :85

mo. m we 8 mm as «2 sea 0:80

mm 2 8 mm mm mm 3... 52x02
ssmomzasm
am 2 3 9.. mm S no: 352830 3..

00330 03:00 000300Hm 00$me 0:00
-552 .282 30> m0 m0 €230
0>0:0m Axe 0>0z0m 0% 0>0:0m “N0 0>0:0m ca 0>0:0m 2K. x00am R Z

$0: .8 $53020“ 0:0 m0 £000 00 00 030: $5 00

200600 0.002084 0 0.25 08 >05 ”:23 ham 00 00.00380: 0.0 H8 00000 5 ”0050003

 

35:03 :00E0E< 000 0000093sz 660m #6 0308

138

 

8 E E 3 0w 8 3m 02

8 w E 8 a S as 8>
53:0

mm 2 E 0... mm 8 was :80 5080

mm m mm ow mm «.m. 9% 50m m3
@532

am : E. s. we as s... 020 seesaw

3 as as mm mm s. m: 88:8er

5 a. 8 .3 am 2 as. 5830

aw : E as a as man 523390
:03005303 been

3 a as as am 3 as use 8 25.8

mm 2 a8 3. mm m... a: 8st as: as 8 25.8

as S 8 mm mm as 0% 83m 52: a3
@EOUQH

00530 00330 0005005 00305 0:00
-032 .252 30> m0 m0 €205
0>0z0m 0x0 0>0:0m ck. 0>0z0m Rs 0>020m AK 0>020m 0A0 x00am “R Z

$0: 00 “$005028 0% m0 £000 00 00 0;: a2: 00

«$0600 :00E0E< m0 0.80 03 m2: was: 000 0a ESmaEE 00 000 000.8 5 ”0030000

 

33003 000€0E< 000 0000093me 309.0. “30:03:08 HA». 0508

139

 

E 2: E 8 .5 8 new 5558 595%
a S 3 mm mm mm, 2% Esmcam
a o E 8 3w 8 2m 5585 ﬁzwsm
Exm mwmswamq
8:“st 2.22:0 95385 =onqu 250
-222 .252 20> m2 m3 imam
383m X «323mm ox). m>mzwm 05 326m 05 m>mzmm 0% xwmmm o5 Z

So: Ho $33028 9: go some ow 3 96: x2: ow

$668 5828;“ Mo ﬂag Ba 35 99: zmm 3 83385“ cm “om $20 5 “coﬁmozg

 

3353 cautm8< was casewxowm Emoom ”€25quan ﬁn onH.

140

The ﬁve dichotomous response items from the 2004 National survey were used to
measure attitudes towards being American, and the cross-classiﬁcation of these items
was obtained for Mexicans (group 1), Puerto Ricans (Group 2) and Cubans (Group
3) (Table 5.2). In applying the standard latent class model to the multidimensional
table formed by these indicators, the ﬁrst step is to decide the number of categories
of the latent variable. Table 5.3 provides the model ﬁt statistics for the T-class
model, for T = 1,2,3. As can be seen from the table, the T = 3 class model
provides the best ﬁt to these data (X32 = 5344,19 = .11, see Table 5.3)1. The
three latent class model yielded an interpretable solution in terms of the division of
Latino public opinion into regarding the ﬁve indicators of Americanism derived from
the 2004 survey2, “Ethnocultural Americanism,” “l\~’Iulticulturalism,” and “Modiﬁed

Mnlticulturalism3.”

 

1When assessing model ﬁt, the null hypothesis is that the model “ﬁts” the data.
Therefore, a non-signiﬁcant p value is sufﬁcient to retain the model. Obviously,
claiming that one model provides a good ﬁt does not preclude numerous other models
from also providing a good ﬁt to the same data.

2The question asked of respondents was “In order for an immigrant to say they
are part of American society, do they have to do each of the following, or not?” (See
Table 5.1). The answers to this question provided the basis for this analysis.

3Since the Ethnocultural Americanism and Multicultural classes comprised al-
most 90% of respondents, and because there are very little differences between the
Multicultural and Modiﬁed Multicultural classes, the analysis focuses on differences
between Ethnocultural Americanism and Multiculturalism. Any analysis of the Mod-
iﬁed Multicultural class would be substantially similar to that of the Multicultural
class presented in the text.

141

 

 

Ht m HH m m H H m
ow mm @m H m H H m
m Ht NH m H H H m
34H mm mom H H H H m
H m HH m m m m H
H H mH H m m m H
m o m m H m m H
m o m H H m m H
H o H m m H m H
m o m. H m H m H
o o o m H H m H
H m m H H H m H
o o w m m m H H
o o w H m m H H
o o m m H m H H
o o cH H H m H H
o o H m m H H H
o m a H m H H H
v H m m H H H H
mH oH um H H H H H
2:350 ,3on 33an 250382 5533980 mD nmzwam :35an 9830on m3 3330
E o>mzmm xsmmm m3 E 30> 3.389th
VH mm 4 @580 mmcoammm 83H

omaoammm 9583935 o>HnH 3 wEEooo< msg#32854 Ho qoSwomHHmwwHO mmoHO ”Gm @358

142

 

355m 0534 chosw Z voom “oopsom .mocmeHEEH 5530 was 58$ 38:.» :8onon MESH

 

an

em

NH

mm

3
wH
m
0H
m
vH

mH

Hum
moH
mH
hm
HH
cm
@
HHV
0H
mm
w
an

HNHNHNHNHNHC‘I

m

r—lr—INNr—CHNNHHN

m

NNNNv—ir—dv—«HNNN

m

v—IHHHNNNNNNN

m

NNNNNNNNNNN

143

 

82 3a 23%
2.0 3% 5%: ”£8: Saws- mm assesses; 80358 £333-».
23 3:: 2st: names: $83. N: 365383 3238 £533
8.0 ”38 24me 39K: 38%.. E assesses
a ax 2m 02 3 382

 

moizm 05qu Hacosa Z Vega 2: 8 8295‘ 2252 2305a econ—ma
3033385 mEom H8 Ammuwsdm Ho macaw HHH 2:99 mosmsgm ”Fm H2302 5m 2nt

 

144

An important question is whether identification as American is similar among the
three Latino subgroups considered here. It is typical to group all Latinos together for
purposes of the analysis or to assume they are different given vast cultural differences.
Table 5.3 provides a statistical test as to whether the analysis should more appro-
priately assume that Mexicans, Puerto Ricans or Cubans are similar with respect to
the various indicators under consideration or whether they are inherently different.
The analysis starts by assuming that there are subgroup differences (i.e. Mexicans,
Puerto Ricans and Cubans are completely heterogeneous) and lets the data assess
whether the homogeneity constraints ﬁt the data. If so, then the nature of American
identiﬁcation is similar across the three groups.

Table 5.4 gives chi—squared values for some simultaneous latent structure models
applied to these data, using the T = 3 class model that provided the most acceptable
fit. First consider model H3 in Table 5.3. This model was obtained by positing a
three class unrestricted latent structure for each group (See Table 5.3). Since this
model does not apply homogeneity constraints, it is a model that allows for complete
heterogeneity between groups. Since L2(H3) = 55.58 on 42 df, H3 provides a rather
good fit to the data. Additional homogeneity constraints were applied in order to
explore whether enforcing additional constraints tended to maximize model ﬁt. In
this vein, models H;,H;’,H;” and H; were considered. Models H; and Hg impose
partial homogeneity but the nature of the restrictions are different. These models

I
are nested within H3 since H3 is obtained from H3 by imposing homogeneity on the

145

latent class proportions, in other words, the following constraints were enforced

53X = «EX = ngX,t=1,2,3.

 

Table 5.4: Chi-Squared Values for Some Alternative
Simultaneous Hypotheses Applied to the Simultaneous
Latent Structure Models From the T = 3 Class Model
of the 2004 National Survey Data

 

Model
H3
H;

II

(if
42
46
72
78
49

x221, XZGF 1905*
55.58 53.44 0.11
51.30 55.01 0.17
115.11 118.23 0.00
154.82 143.17 0.00
58.25 56.83 0.21

 

On the other hand, Hg is obtained from H3 by imposing homogeneity on all

conditional probabilities, namely

MGX _

7r2'11

7911

fax _
7rkit

EGX

7r11::

EGX
mlt

71'

4,721,“ = WEGX,2'=1,2,t= 1,2,3; (5.1)
= 372%: E?X,k=1,2,t=1,2,3; (5.2)
ﬁg?“ =7rgfx,k= 1,2,t= 1,2,3; (5.3)
=WEEX=WEEXJ=L2¢= 1,2,3; (5.4)
= «295‘ = 7.5ng =1,2,t= 1,2,3; (5.5)

where M = melting pot, V = voting, C = citizenship, E = English and B =

believe in Constitution.

In .
Model H3 imposes homogeneity on both the conditional probabilities as well as

the latent class probabilities. In other words, this model imposes complete homo—

146

geneity on the parameter estimates for the latent class model. With a chi-squared
value L2(H;”) = 154.82 on 78 (if, 11:3” provides an unacceptable fit to the data. The
unacceptability of an across-group equality constraint on the latent class conditional
probabilities indicates a significant. difference in the distribution of Mexicans, Puerto
Ricans and Cubans in these categories. Finally, model Hg is a model nested between
H3 and Hg” with the particular homogeneity constraints determined from the original
unrestricted model. W ith a chi-squared value L2(H§) = 58.25 on 49 df, H; provides
the best fit to the data (Table 5.4). Parameter estimates for this model appear in

Table 5.5.

147

 

Table 5.5: Estimated Parameters for the Final Latent Class Model (Hg).

 

Group Latent

 

Model .9 Class t ﬁgx 3‘1“!th 77K?“ aggx affix afﬁx
H; 1 1 .57 .10 .95“ .79 .75a .95a
2 .31 .18 .30 .13 .15 .61
3 .12 .00 .001) .39 .52 .84
2 1 .48 .13 .95“ 100’) .75a .95“
2 .32 .05 .00” .34 .36 .62
3 .21 .00 1.00 .27 .82“ .81
3 1 .54 .11 .99 100’) .82a .95
2 .38 .00 .43 .20 .52 .95a
3 .08 .36 .00?) .28 .41 .56

 

Notes: Only “Yes” answers are shown for each class.
a The parameter value is subject to an across-group homogeneity constraint
b The parameter value is subject to within-group restrictions.

(The entries of 0.00 for H; were fixed restrictions.)

ng 2 percentage of each subgroup that are members of each class.

«ﬂax = conditional probability of “Yes” to US as Melting Pot given
class membership.

an?“ 2 conditional probability of “Yes” to Voting in US elections given

class membership.

afﬁx = conditional probability of “Yes” to Citizenship given
class membership.
«{2th = conditional probability of “Yes” to speaking English given
class membership.
«ﬁgx = conditional probability of “Yes” to believing in US Constitution given

class membership.

 

A similar analysis was performed on the various indicators derived from the 2002
National survey. In contrast, a 2—class restricted model with full homogeneity on the
conditional latent class probabilities emerged to provide the best fit to the data (see

Table 5.6).

148

 

we: 8% 23$

:8 $.33 3....me 88% 38%... at 355855: 82858 £5 533m

8.0 2.8.8 8.0% 2.5. 8.23- M: 558882 85358 5.23 3.29m

88 $.QO 3.92: $88 2.8%.. E 558858; 58358 5:» 3.29m
3 mx 2m 02 3 882

333m 05qu 8:03st mgm mg 8 comma/x £2502 ouzuoaum E333

msowcaﬁsamm maom pom AmmuwSUm wo wasm H: omkrﬁv woﬁmmﬁwum ﬁrm EUOE “ch 336R.

 

149

The conditional probabilities represent a measure of the degree of association be-
tween each of the manifest variables and each of the latent classes. What, then, is the
overall pattern that emerges from the analysis? The latent structures of Americanism

for Mexicans, Puerto Ricans and Cubans are elaborated below.

5.4.1 Ethnocultural Americanism (Class I)

As can be seen from Table 5.5, there is a strong ethnocultural component to Amer-
ican identity reflected by the clustering of responses. For example, among Mexican
respondents who claim membership in class I, more than 7 out of 10 believe that in
order to be an American one must speak English (.75 in Table 5.5), more than 9 out of
10 believe that to be American one must believe in the US constitution (.95), almost
8 of 10 believe that to be an American one must be a US citizen (.79) and more than
9 out of 10 believe that to be an American one must vote in US elections (.95) (See
Table 5.5). In addition, the conditional probability of accepting the notion that the
US is made up of a single core Anglo-Protestant culture given membership in this
class is .10 for Mexicans, .13 for Puerto Ricans, and .11 for Cubans. The fact that
these are high can be seen only in relation to whites’ views with respect to this mea-
sure. In all cases, Mexicans, Puerto Ricans and Cubans in class I are twice as likely
than whites (.05) to agree that the US has a single core Anglo—Protestant culture (de
la Garza, 2005). On this basis, this class is considered the class of respondents who
hold prescriptive conceptions of what it means to claim membership in the American

community. The conditional latent class probabilities indicate that Mexicans (.57)

150

(Table 5.5) are more likely than either Puerto Ricans (.48) (Table 5.5) or Cubans
(.54) (Table 5.5) to be in the ethnocultural class. In total, this class comprises the

majority of the total population (.55)(not shown in Table).

5.4.2 Multiculturalism (Class II)

The multicultural class, which accounts for approximately 26% of the entire sample,
represents respondents who adopt multicultural conceptions of American identity.
This is best seen by the fact that Latinos in this class overwhelming believe that
speaking English is not necessary in order to claim membership in the American
community. For example, 85% of Mexicans in this class believe that it is not nec-
essary to speak English in order to be an American (Table 5.5)4. While a large
proportion of respondents in this class feel that belief in the US Constitution is a nec-
essary prerequisite for “being an American,” presumably because it guarantees equal .
rights and freedom, the overwhelming majority of Mexican, Puerto Rican and Cuban
respondents in this class do not believe that a prerequisite for claiming membership
in the American community is either voting in US elections (70% of Mexicans, 100%
of Puerto Ricans and 57% of Cubans)5 or acquiring American citizenship (87% for
Mexicans, 66% for Puerto Ricans and 80% for Cubans). For example, the conditional

probability of believing that it is necessary to vote in US elections given member-

 

4Since possible responses are “Yes” or “No,” the probabilities must sum to one.
The ﬁgure quoted in the text of 85% comes from Table 5.5 and is computed as com-
puted as (1 - .15) Only afﬁrmative responses are provided in the table for conciseness

5These percentages are taken from Table 5.5 indirectly. For example, the percent-
age for Mexicans was computed as 1 - .30 = .70. The percentages were similarly
derived for Puerto Ricans and Cubans.

151

ship in this class is only .30 for Mexicans, .00 for Puerto Ricans and .43 for Cubans
(see Table 5.5). Additionally, the conditional probability of believing that to be an
American one must be a citizen given membership in this class is .13 for Mexicans,
.34 for Puerto Ricans and .20 for Cubans. The larger estimate among Puerto Ricans
presumably reflects the fact that they are granted citizenship status by birth. Finally,
the majority of respondents in this class believe in the multicultural conception of
American identity as reflected by their acceptance that the US is made up of many
cultures. For example, all Cubans in this class view America as being a diverse place.
A similar proﬁle with respect to this measure emerged among Mexicans (82%) and
Puerto Ricans (95%).

The same analysis performed on the 2002 data with respect to the measures of
American liberalism revealed that a two-class solution provided the best ﬁt, with
differences between classes lying primarily with the role of individual self-reliance
and governmental paternalism. Parameter estimates for the two class model are
shown in Table 5.7 and Table 5.86. As can be seen from both tables, one class
adheres to the strict interpretation of American liberalist philosophy while one class
favors American liberalism but with governmental paternalism. The structure of

these classes is elaborated on below.

 

6Table 5.7 provides a more intuitive and straightforward way of presenting the
data. Table 5.7 is the ﬁnal restricted model whereas Table 5.8 provides the hetero—
geneous model. The interpretation of the classes and their meaning is largely the
same.

152

 

mm. mm.

so. no.

mo. co.

Hm. mm.

vm. em.

co. 8..

00. mm.

OH H H.

we. we.

vm. mm.

me. ms.
Emzmﬁopwm 8ngan
35885950 swoﬂmﬁ<

 

: 536 533

H mmﬂv page;

985 8 33893
8m were:
mﬁem
EQSOO Show E Hwﬁmm
mb 5 Sﬁmm
humnsupoaao £anon
m9 5 Siam 82
m3 5 Siam
hymnsﬁoano Enemumm
9,6:stme
8.385
useEHEQEU mo 20m
.5562
moqszvuﬁmm 33:3qu
Emzwsumpam EEmEEQEU

moswzmnubmm

 

a
283:0 was. 285 0.225 23“on

US$335 “Becca/a mo ESE 8.30-038 88.03%

28.23 msomcmﬁsﬁﬁm wmuoﬂamwm 2: 8m Eouoagam voamﬁﬁmm «an @3st

153

mocswq m0 racism #00038 Z meow ”08:0m

ﬁm 0.5me E 00303 000 800800 wwgom ”302

 

 

z... 8. 23:0

mm. mm. £52m oﬁosm

om. ow. 28632
000% med—U 000.23 3:036:00

S. 8. oz

om. em. spa
msomwzmm

3.. 8. pea 5B cam

8. am. .5 82 860 53 83:
3:50 5:63

MS. 3. 8330 £30m

154

8:504 m0 303:6 :32me meow ”00.50m

 

 

H3 3 H03 moo 5.0 H30 moo m
H30 8d 8.0 m3 H20 :5 H; H m
m; a; woo 33 H: 0.0 oHd m
So :3 m3 Ed m3 HHS 8.0 H m
Hmo Ed 30 moo o H 8.0 2.0 m
:8 SS 30 mg 2.0 HS 93 H H Hm
XMW? X03: 5%.“: xmw: xmw: 5%? kW: H $20 m 352

30pr 0:000

 

Haﬁmswmohmuwm 303800 :55

H0002 mmﬂo m H B 05 How 800088me vmpmﬁsmm ”wb 0308

 

155

5.4.3 American Liberalism (Class I)

A consequence of the partial homogeneity constraints is that the components of Amer—
ican liberalism are the same across ethnic subgroups. In other words, a statistical test
of the structure of American liberalism revealed that it is substantially comparable
across Latino subgroups. The latent class probabilities, however, were allowed to vary
among ethnic origin groups (hence, partial as opposed to full homogeneity). In this
class the majority of respondents favor individual self-reliance, or the belief that it is
each person’s responsibility to take care of him- or herself (.54) as opposed to govern-
mental paternalism (.42) or neither (.04) (Table 5.7). All of the respondents in this
class favor a smaller government with fewer services as opposed to a large government
with many services. \Nith respect to conditions in the US relative to one’s country of
origin, 66% of respondents believe that personal opportunity is better in the US while
91% believe that economic opportunities are better (Table 5.7). Eighty-two percent
of Latinos in this class accept the melting pot conception of American identity while
74% have accepted the Protestant work ethnic, namely that hard work offers the best
guarantee of success (Table 5.7). Finally, 90% of respondents in this class manifests
a belief in God. Overall, 20% of Mexicans, 22% of Puerto Ricans and 32% of Cubans

are members of this class (Table 5.7).

5.4.4 Governmental Paternalism (Class II)

As shown in Figure 5.1, the main differences between the American liberal class and

the governmental paternalism class pertains to the fact that 73% of respondents in

156

this class believe that the government in Washington should do everything possible
to improve the standard of living of all Americans (See also Table 5.7). In addition,
89% of respondents favor an expansive government that provides many services. The

majority of Mexicans, Puerto Ricans and Cubans all claimed membership in this class

(80%, 78% and 69% respectively).

. WWWWMWWMW
l 1.2 r—-~ ~ --~ — _.__-___.__,,_",

 

 

 

  

 

 

 

 

  

 

 

 

_'.\
.....__J

 

    

.____T_. _

Paved Enamic “but (Hid llelthot WM)
VLGOV) W W W “bdrm (Omar

WW (BdhrhUSor (BahthBor 81mm moan)

EXP-MW) 00 00

’ HMW'WW

 

 

Figure 5.1. Components of American Liberalism

157

5.5 How Does Acceptance of American Liberal-
ism change with Acculturation, Exposure and

Context of Reception in US Society

In contemporary politics, divisions between multiculturalism and liberalism on the
one hand and ethnocultural Americanism on the other, are further delineated by
race, ideological self-definitions and partisan affiliation (Citrin, 2001). For example,
Democrats and self-identiﬁed liberals are more likely than Republicans and conserva-
tives to support multiculturalism. Therefore, to explore how other factors change the
distribution of latent classes, I re—ran the latent class models adding covariates. The
latent class model with covariates allowed me to test the following propositions: ( 1)
American identification increases Latino support for American liberalism; (2) Expo-
sure and acculturation increases the probability that Latinos will endorse American
attitudes and values; (3) Latinos who are structurally assimilated will adopt more
favorable attitudes regarding American political culture; and (4) Context of recep-
tion, particularly discriminatory context and favorable community context changes
the distribution of responses for each class. Beta coefficients and odds ratios for the
relative effect of the covariates on membership in each class (Ethnocultural Ameri-
canism, Multiculturalism or American Liberalism) are presented separately for the

two analyses below.

158

5.5.1 A Multinomial Logistic Latent Class Analysis of Amer-
icanism

American Liberalism versus Governmental Paternalism Among l\-’Iexicans
and Cubans, the odds of being in the American liberal class versus the governmental
paternalism class increase with age, education (Mexican and Cubans only), income
and English proﬁciency (Table 5.9). First. generation Mexicans and Puerto Ricans
are more likely to be in the American liberal class but first generation Cubans are
more likely to be in the multicultural class. Nativity status appears to distinguish
membership in each class. For example, Mexican citizens are less likely to be in the
American liberal class while Cuban citizens are less likely to be in the class favoring
governmental paternalism. Mexicans who believe that life in the US is “better”
relative to Mexico are significantly less likely to be in the American liberal class. For
Cubans, however, the odds ratio is greater than one, indicating that Cubans who
believe that life in the US is “better” relative to Cuba are significantly more likely
to be in the American liberal class. While the unadjusted latent class probabilities
were low, the adjusted estimates are much higher. Importantly, after controlling for
other factors, membership in the American Liberal class increases by .22, .12 and .37
percentage points for Mexicans, Puerto Ricans and Cubans, respectively (see Figure
5.2). These figures represent the difference of the class prevalence that is reported
in Figure 5.2 (i.e., .4186 - .2022 m .22, etc). Substantively, this means that age,
acculturation, education, nativity status and context of reception are associated with

the relative proportion of individuals who favor the components that are the basis for

159

each class. Importantly, the measures of identificational assimilation were not found

to be related to class membership.

160

0032555me

 

 

as Home. m3 mg. was Had- owe m3- HazeEHaH

EOEGCHEEQEQ

moo Es 3H and Ed Hoe- w; Ed. 880 a=20
comaaoomd m0 $8.250
U20 wmmSmsomomcoO @520

8.0 H3: O: 8.0 I 1 8.0 93- m0 5 .5st
coHuaHHEHmma. HaaoHESmm 383260 H2633?

m3. wen SH a; - I moo m3- Ame Hexagon 8:.

S .0 Se «3 3H and 8H 858952 555.
ﬁOSGEE~mm< ~5GOENU¢$Q®©H

was NEH 8H 8.0 8H 8.0 SH HHS 2:85

8.0 3.8 92 8.0 was :3. was 53 858350
ﬂowed—mammmd. ~5h§uoﬂhum

8.0 93 SH was - I 8.0 moo- 850

8.0 2% m3 ”3... Es H.H.H on mos 832250 $5

8.0 8.8. 8H 8.0 8H 8.0 3.0 Be. EHEBHEE

25 Sam m3 H2- go 8.0- SH m3 @282
:Omum~mgmmm< ~mhﬂu—ﬂo Uﬁw ﬂomumhﬂu—ﬂou<

a .3 x m 4 0.0 0 £0 a £0 n
953:0 mﬂﬂomm Cayman“ 250682

539580me 130055950 0:29,. 0520.825 aeotoﬁxx ”moﬂwm 3000 00m mucowomwooo 30m ”mm 035.

161

..Emsmemnﬁ $0208<e $20 0:83 mm @0330 8:230: 23. ”mos—oz

 

cod

mmd

owe.»

#0.:

2: mod

3.0 end.

m.

04

wed 2: v0.0 0m<
03:20:? 35:00
mm: omd 28395353 Egamwmbﬁﬁmq

162

Ethnocultural Americanism versus Multiculturalism The odds of being in
the Ethnocultural American class versus the Multicultural class increase with educa—
tion and years lived in the United States (Table 5.10). Latinos who hold traditional
values and who are proﬁcient with English7 are more likely to be in the ethnocul-
tural American class. For all three groups, increasing perceptions of group—based

discrimination are associated with less support for ethnocultural Americanism.

\ Prevalence am airliners" —isrnaftieracoour 7' Jag” j
forldeology,Acculh1raﬁonandConteﬁofMoepﬂon 1

 

 
  
 

  
  

0.2022 0.2167 Vlﬁthout i
Covariates
means Puerto Ricans cm

Figure 5.2. Distribution of American Liberalism with Covariates

 

7Linguistic proficiency i 3 reverse coded here such that increasing values are asso—
ciated with Spanish dominance.

163

 

20000000004 00000000000: $30 0002 mm 3000000 000000000 2:. ”00002

 

00.0

0.00

00.0

0.0

00.0

00

00.0

00.0

00

:0 00A
004 00.0
002 00.0
mm: 04
00.0 mm:
0.3 N00
0502 000
000m 0mg

00.0

30-

N00-

00

000

00.0

mm0-
00.0

qumd £0 0

moa N00 . 00A 0.0

0H; :0 00 00.0-
30 R:- 000. 0.0.0-
000 «0.0 00.0 00.0-
050 00.0- wm.m 0.0.0
I I 004 00.0

3.0 30- 0.00 00-
mﬁﬁ :0 :0 R0

000 Q 050 m

0m<
033.:5/ 35:00
0030000085
000305
0030000800
0.0QO 00000
:oﬁmooom no 008000
0:0 mmo:m:0mom:o0 000:0
00:00::
0030000m
0050:5330. 30:00:00
m3 0H 23>
00000004
05000030008

:omummammmdx 3.50:0 0:0 :omuwu:u_:0o<

 

 

memo-:0

maxim 3.80m m:00€82

Emzdsuwudnm awaﬂoesuokwow m5mu®> amzdhmﬂwg Gdowhma< 00“me mUUO Uﬂd mucwmoawoo wuom .Oﬁh oﬁﬂmn—L

164

5.6 The Nature of Ethnic Identity Choices Among

Immigrant POpulations

I next turn to a more objective measure of Latinos’ degree of attachment to the
nation, namely primary mode of self-identification. The primary question I consider
is whether individuals of Latin American descent tend to think of themselves as

“Americans” rather than as Latinos or co—ethnics.

5.7 The Distribution of Identity Choices Among

Latino Immigrants

The 2002 National Latino Survey asked the following question to survey respondents:

1. People choose different terms to describe themselves. I’m going to read you
a few different descriptions. Please tell me whether you have ever described

yourself as any of the following
a. Respondent / parent Country of origin?
b. A Latino/Hispanic
c. An American

Tables 5.11, 5.12 and 5.13 report the distribution of ethnic self-identity modes
by ethnicity and generational status. As can be seen from the tables, there exists
tremendous variation in how Latinos view themselves despite the fact that they are

viewed by outsiders as being relatively homogenous.

165

 

:3000 00003000?“ :5300

:3000 .00? 00:0 00:09-00 .02, 0:0 h00>r :00 00:00:00005 00 0000 mx :0 00000 8:020:00 00 00.0 ”0030 Z

 

8.0 a 00 E 8.0 S 5 S 8.0 3 00 cm $2.05. 3.
So 0 S 00 80 a 00 8 80 8 2 0 06805803
8.0 m S 00 3.0 S a E 8.0 E 2 g 50:0 0 00:60

GompdommuzovH

 

& :00 0:0 :00 0:0 :00 003 a :00 0:0 :00 0:0 :00 303 a :00 0:0 :00 0:0 :00 00H
:0000 :008 03.00% :00082

A:030003:00H 5:353 :030:0:00 0:0 00000000 00 00:30am 0:080. 0033503000 03:30 00 :03:£3me 0000:000m 030 030.0

 

:3000 0003:0890 5:300

:3000 00>h 3:0 00:00.00: 02. 0:0 B00%F :00 00:00:0000E 00 0003 mx :0 00000 8:000:00 00 300.0 ”00302

 

00.0 00 00 00 00.0 003 ma 00 00.0 00 30 mm :0030E<
00 00 00 E. 00.0 E. 00 3 00.0 B. 00. 0.0 0:303
00.0 00 30 00 00.0 00 00 00 00.0 00 00 00 £030 00 NA3:000

:030003:00H

 

a :00 0:0 :00 0:0 :00 003 a :00 0:0 :00 0:0 :00 003 a :00 0:0 :00 0:0 :00 303
:0000 :003 03000. :00002

:030:0:00 0:0 00000000 am 00:30.0 0:084 0033503000 00:50 00 :03003305 0000:0000m ”:0 030B

166

 

:3000 00m00:00:0m 5:200 “0002

 

 

8.0 0000 8.0 00.: 8.0 .303 00x
02 02 2: 3 02 2: 02 02 m 089
E 00 E 3 00 S E 00 0 08:08.4. :2
E 0: 0: S E 0: a: 00 mm 200005800:
3 00 E 8 00 m0 «0 00 E 50:0 00 00:50

”83000353

 

:00 0.5 :00 0:m :00 00H

:0000

:00 0.5 :00 0:m :00 00H

:00E 3025

:00002

:00 0.5 :00 0:m :00 00H

:030:0:00 0:0 0:0:wasm :5 09504 w:0::< :03005303 00 0002 000:0:m 00 0030005000 0w00:00:0m ”2.0 0309

167

As shown in the responses to this question, there are differences both within
each national origin group and between groups. The overwhelming majority of first
generation Latinos regardless of national origin, reported having identiﬁed in terms
of their own ethnic group. As shown by Table 5.11, first generation immigrants are
more likely than either second or third generation immigrants to identify as a member
of their own ethnic group. This pattern is reversed by the third generation, where
the plurality of Mexicans prefer to self-identify as an American. Only 25% of ﬁrst
generation immigrants reported identiﬁcation as American but by the third generation
this percentage increased to 96%. Of all subgroups, ﬁrst generation Cubans are more
likely than ﬁrst generation Mexicans or Puerto Ricans to report self-identifying as an
American. With respect to identifying as a member of one’s ethnic group, responses
vary. Among the ﬁrst generation, for example, 98% of Mexicans, 93% of Puerto
Ricans and 82% of Cubans indicated that they had done so. This pattern also changes
with exposure to American culture as indicated by drop in the percentage of third
generation Mexican, Puerto Rican and Cuban immigrants who reported identifying in
national origin terms. By the second and third generation, however, Mexicans, Puerto
Ricans and Cubans are all less likely to have identiﬁed themselves as members of their
own ethnic group with Cubans being the most likely to change and Puerto Ricans
being the least likely to change. Table 5.12 reports the distribution of responses
within identiﬁcation. Of all Mexicans who identified themselves as a member of their
ethnic origin group, 67% were ﬁrst generation immigrants while only 14% were third
generation immigrants. With respect to Puerto Ricans, 57% were ﬁrst generation

compared to the 12% who were third generation. Finally, of all Cubans who gave

168

afﬁrmative responses to this question, 82% were ﬁrst generation immigrants and only
2% were third generation Cubans.

Just as ﬁrst generation immigrants are more likely to self-identify as a as a mem-
ber of their own ethnic group, they are also likely to identify as a panethnic. The
percentage of respondents who reported having identiﬁed as Latino/ Hispanic was
higher among ﬁrst generation immigrants across all categories. However, by the third
generation, Latinos were more likely to identify both in terms of a panethnic group

and as an American than they were their own ethnic group.

5.7.1 A Multinomial Logistic Model of Identity Among
Latino Immigrants

The coefﬁcients in Table 5.14 are the standard output from a program that esti-
mates the multinomial logit model. The minimal set of contrasts involves all possi-
ble comparisons with outcome American. This means that identiﬁcation with one’s
country of origin is compared to American identiﬁcation (CIA) and identiﬁcation as
Latino / Hispanic is also compared to American identiﬁcation (L|A). These coefﬁcients

correspond to the equations:

9
(WOW?) = 30,C|A + Z ﬁnd/4173+
i=1

9
[TIQLIAGIV = /30,L|A + Z ﬂan/1?

2:1

where :1: IS a vector of independent varlables including group-based dzscnmzna-

169

tion, individual-level discrimination, ﬁrst generation, second generation, language

8. As can be seen from Table 5.14, the

proﬁciency, education, age, gender, income
hypothesis that group-based discrimination does not affect primary identification can
be rejected at the .00 level. The LR and Wald tests indicate that all the variables in

the model are statistically signiﬁcant at the p = .05 level except for individual-based

discrimination and the income dummy variables.

 

8With J = 3 dependent categories, there are J - 1 = 2 parameters, ﬁkmllr asso-
ciated with each variable 23k. The hypothesis that :rk does not affect the dependent
variable can be written as H0 : mull? = Bkﬁlr = 6&3” = 0. Since ﬁkn‘lr is necessarily
0, the hypothesis imposes constraints on J — 1 parameters. This hypothesis can be
tested with either a Wald or LR test. The interpretation of the LR test is similar
to the Wald test statistic: If the hypothesis that group-based discrimination has no
effect on primary identiﬁcation is true, a 02 statistic of 20.09 or larger would occur
0% (I? < 0.001) of the time due to sampling variation.

170

Table 5.14: Logit Coefficients and z—scores for

Multinomial model of National Identity

 

Country of Origin vs. American Latino vs. American

 

 

Variable 6 Z 6 Z
Acculturation
English Dominant -.684*** —4.08 -.790*** -4.21
First Generation 2.29*** 10.83 1.27*** 5.24
Second Generation 1.02*** 5.20 .567*** 2.60
Structural Assimilation
Less than HS .918*** 3.44 .259 0.85
High School .588*** 2.82 .233 0.98
Some College -.107 -0.52 —.706*** —2.83
Group Consciousness
Group-Based Disc .719*** 2.86 1.31*** 4.39
Individual Disc. .042 0.10 .426 0.92
Control Variables .
Female .373*** 2.53 .471*** 2.72
Age -.020*** -3.59 -.031*** -4.58

171

 

 

 

wzr mam. 2;. 30.- m5. moo. m2. wmﬁ. 0mm. m :00
30. 0mm. m3. moo. wwo. mac. moo. owe. w:. H :00
$0 HHQU OHQU .05 HHQU OHQU ~00 HHQU oHQU

0:504 00 :00E0::<

 

0:00pm 0:200:00 0:0 :050582005 0000m000:0 00

:0~E00:0& .3 :03000553 00 ..350000:& 03200.5 ”man 0308

172

It is possible to also test the hypothesis that the two outcomes American and
panethnicity can be combined. This is done by ﬁrst selecting the individuals who
prefer identification as either an American or as a co—ethnic, estimating the following

binary logit model

9
("Q/1M?) = ﬁO,A|L + Z ﬁz’AlL?
i=1

and then computing the LR test of H0 : 61 = = 69 = 0. For these data,
G2A|L = 162.5, (if = 13, p < .0019. The hypothesis that American and Latino
identities are indistinguishable with respect to the independent variables in the model
is rejected at the .01 level. Similarly, a test of the hypothesis that country of origin
and Latino identities are indistinguishable with respect to the independent variables
in the model is also rejected at the .01 level.

I next turn to the predicted probabilities generated by the model (Table 5.15).
Table 5.15 shows primary mode of identiﬁcation by perceptions of group-based dis-
crimination and generational status. The results suggest that ﬁrst and second gener-
ation Latinos who perceive discrimination to be a problem are less likely to identify
as Americans than if they do not perceive discrimination to be a problem. Second
generation immigrants are more likely to identify as a member of their home country
if they perceive discrimination to be a problem. Finally, ﬁrst and second generation

Latinos are more likely to identify as Latino/ Hispanic if they perceive discrimination

to be problematic.

 

9I‘he Wald test gives a similar result: WA] L = 132.8.

173

To examine the effect of age on the probability of falling into category J all
variables were held at their means while age was allowed to vary over its range (Figure
5.3). The results indicate that as age increases the probability of identifying both with
one’s country of origin and along panethnic lines decreases while the probability of

identifying as an American increases.

 

Predcted Probdilityof Mode of ldentiﬁdion as a Function or
Rasponderfsﬁge

+AnAnen‘cm +Ooumy dOig'n ﬁ—Latino/Hisp‘anic
J

 

 

 

 

0.60
0.50 -
0.40 - ..\l
0.30 -
0.20 .

0.10 .

 

 

 

o-m l T l l l l r T T
18 26 34 41 49 57 65 72 80 88

 

 

 

Figure 5.3. Mode of Identiﬁcation By Age

Primary interest is in the variable measuring the effect of one’s perception of group-
level discrimination on the mode of identiﬁcation. Table 5.16 shows the effect of the

perception of group-based discrimination on primary mode of identiﬁcation. The odds

174

of identifying as an ethnic or panethnic group member versus an American increase
signiﬁcantly with increasing perceptions of group—based discrimination. Figure 5.4
shows how the predicted probability of group identiﬁcation varies with increasing
levels of discrimination. As can be seen from the ﬁgure, the probability of identifying
as American decreases while the probability of identifying either with one’s country
of origin or along panethnic lines increases with increasing perceptions of group—level
discrimination.

Table 5.16: Factor Change in the Odds for Group-Based Discrimination

 

 

 

Outcome n
Factor Change in the Odds of m vs. n C L A
Outcome m C Country of Origin - .554 2.05
L Latino/ Hispanic 1.81 - 3.70
A American .487 .270 -

175

Assam amwohomv messed
:osﬁocov HES muoE< :oﬁwﬁaﬁoﬂv womandsgv mo £5595th 53w :osmocscog m0 @543on wouowwoum «an 8sz

8820;
SEEBme 38883 889618 83:88 .90. 92m 3 2 as:
8853;85535332638382

2 a a b e m v m N _

V3 W 8.:
£32m

”.86

gigcoisooigi .8...

8.—
0.59::

176

Personal Discrimination

Gen 1
Gen 2

Language

Less than HS
HS
Some College

Age

Female

Less than $20,000
Less than $40,000
Less than $75,000

Probability at Mean

Change
A1
A0

ARange
A1
A0

ARange

O —+ 1
0 ——> 1
A1
A0
ARange
0 —+ 1
O —+ 1
0 —> 1
A1
A0
ARange
0 —» 1
0 —> 1
0 -—> 1
0 —> 1

177

Z

.123
.120
.038
.050
.046
.009
.268
.123
.153
.096
.070
.128
.080
.067
.234
.003
.047
.055
.038
.017
.016

Table 5.17: Discrete Change in Odds of Primary Self-Identification
Variable

Group Discrimination

C' L .A
.043 .142 a185
.034 .147 a181
.011 .046 =057
2041 .075 m034
a037 .069 2033
a007 .013 a006
.403 '3031 12372
.185 2021 «164
m118 a111 .230
m083 2061 .144
a060 2045 .105
.192 a060 a132
.120 -.026 -.093
.044 '3101 .057
a154 a197 .351
m002 a003 .005
a023 a047 .070
.042 .040 m082
m058 .014 .044
.008 .018 2026

.008 -.024 .017
.495 .224 .281

Table 5.17 contains estimates of discrete change from the model of primary iden-
tiﬁcation. First, consider the dummy variable ﬁrst generation. Holding all other
variables at their means, being a ﬁrst generation Latino decreases the probability of
identifying as an American by 37.2% and increases the probability of identifying with
one’s country of origin by 40.3%. By comparison, the average absolute change for a

standard deviation change in the perception of group-level discrimination is 12% and

is 5% for the perception of individual-level discrimination. The effect of group-level
discrimination exerts the largest effect on the probability of identifying as American:
the probability of identiﬁcation as an American per standard deviation change in
group discrimination decreases by 18%.

I now turn to the question of whether the effect of discrimination on identity choice
is the same regardless of one’s generational status. To test the hypothesis of no effect
requires testing the effect of discriminatory treatment, group and individual- based,
at the intergenerational level. This suggests adding an interaction effect between
discrimination and generational status to the model. The results are presented in

Table 5.18.

Table 5.18: Interaction Effect: Generational Status and Discrimination

 

 

Interaction Effect Country of Origin Hispanic/ Latino
Gen 1 * Group Discrimination 1.37** -.117
Gen 2 * Group Discrimination 1.60** .337
Gen 1 * Individual Discrimination -1.09 .480
Gen 2 * Individual Discrimination -.968 -.789

The results indicate that the perception of group-level discrimination on identity
choice is signiﬁcantly different for ﬁrst and second generation Latinos than it is for
third generation Latinos (Table 5.18). More speciﬁcally, individuals in both the ﬁrst
and second generation who perceive group-level discrimination to be a problem are

more likely than third generation Latinos to identify with their country of origin.

178

 

Table 5.19: Selected Multinomial Logit Contrasts

 

 

 

US Born Foreign Born
Variable 1 vs 3 2 vs 3 1 vs 3 2 vs 3
Group—Based Discrimination
b .480 1.32*** 1.11** 1.46***
p .160 .00 .013 .007
Second Generation**
b 1.02*** .512** .000 -.029**
p 0.00 .021 .978 .018
Language
0 -2.29** -2.44** -.390* -.588**
p .022 .015 .070 .019
Less than College Education
0 .695” .217 .719* .442
p .015 .471 .073 .345
Linked Fate (Latinos share one culture)
0 .236 .824** -.218 .370
p .480 .013 .629 .459

 

Notes: 1=CO; 2=Latino;3=American

** Years in US for foreign born

 

Separate analyses from two different regressions of US born and foreign born re-
spondents are shown in Table 5.19. As can be seen from Table 5.19, group-based
discrimination is a predictor of “American” identity for three of the four contrasts
reported. First generation immigrants (i.e., foreign born) who perceive discrimina-
tion among Latinos to be a problem prefer both ethnic and panethnic identities to
an American identity. Holding other factors constant, US born immigrants of second

generation status are more likely than third generation immigrants to prefer both

179

Latino and panethnic identities to an American identity. With respect to first gener-
ation immigrants, exposure to American culture significantly increases the likelihood
of preferring an American identity as opposed to a panethnic identity (result not
shown). English dominant and highly educated Latinos also prefer American self-
identification. First generation Latinos who believe that Latinos share one culture
are significantly more likely to prefer panethnic identiﬁcation to an American one.
Finally, Table 5.20 shows the results by ethnic originl.0 The table shows odds ratios
and standard deviation changes for each variable. The context of reception and group
consciousness variables proved to be the most consistent predictors of primary iden-
tiﬁcation. The results show that l\v*Iexicans who perceive group-based discrimination
to be problematic are significantly more likely to prefer to identify with their country
of origin than as Latino but the opposite is true among Puerto Ricans. On the other
hand, perceptions of group-based discrimination are significantly associated with an
American identity over a panethnic one among individuals of Cuban descent. With
respect to individual level discrimination the analysis revealed that Puerto Ricans
who have experienced personal discrimination are more likely to identify as Puerto
Ricans than as Latinos. Context of reception predicts identiﬁcation in other ways
among Mexicans as well. Mexicans who perceive conditions to be better in America

are more likely to identify as a member of a panethnic group than as Mexicans.

 

1Only the results that were significant are presented. There were 6 possible con-
trasts for each independent variable for each national origin group.

180

 

amoroﬁzwummoqsequm MOOHH ”8qu

 

 

 

 

a; $2 9:. I I .. 33 and $3 5.2%
£3 £2 34 I - I was memo mm; a
:33- 1&3- was . 1 .. ...EE- :37 ammo- s
8E. mb
3.3 was $3 08.0 was as 3w; 3d awe 5%
$3 :3 as; 33 was m2 sumo $3 memo a
ago. 84.0- ammo- two- 33. $3 $3- ammo- raid- s
m: 5 exam
mg m3 .9: was was 22 :3. 8.8 «3, as
is; it”? ”was *8; items 83 :33; time.” 1.1.33 s
GOMHNHQGOU Hmhmm
News 83 a: 33 Q: a: a: as. $3 5.2%
”as 83 Ba 53 :2 m3 «3 came Sec a...
33. as; «So 1.33- was imﬁd ES 83. 23- s
Gomuwﬂmgmhummﬂ— awﬁomhmm
08.0 mm; was 2: ammo $3 Sm. a: can ages
and some :2 ms: ....de some at. m: a; a
:53. $3- $3 was 98.0- ES- :53. 83 time s
:OmaﬁﬂmEthme nun—Ohm
”Em me; am: me; am: am: new 3.: mm: 0333/
Cdﬁab 2605* Ouhwznm. QQUNXQE

 

 

$335?» acovcommvﬁ wopoﬂom Mom EwescoU ﬁwoq RwanSBE Noah 2259

 

181

Several of the assimilation measures were strongly related to identification (results
are not shown). Within each national origin subgroup, first generation immigrants
are overwhelming more likely to self-identify as members of their ethnic group than
as members of a panethnic group or as an American. Second and third generation
Mexicans are more likely to prefer a panethnic identity than are ﬁrst generation
immigrants. Mexicans and Cubans with decreasing ties to their country of origin are
signiﬁcantly more likely to identify as Americans. In addition, the relative conditions
of Latin America versus the United States are associated with mode of identiﬁcation
among Mexicans. In contrast, Puerto Ricans were less likely to identify with their
country of origin than as Latino over time (b = —0.041,p = .076). All Latinos,
regardless of national origin, are signiﬁcantly more likely to identify as Americans
if they are English dominant. With respect to structural assimilation, education
and income were signiﬁcantly related to patterns of identiﬁcation among Cubans.
Cubans are signiﬁcantly more likely to identify with their country of origin than
as an American with increasing income level (b = .079,p = .075) and less likely
to identify with their country of origin than as Latino with higher education (b =
—.668,p = .012). Age was also a signiﬁcant predictor among Cubans, who are less
likely to identify as a Latino/ Hispanic than with their country of origin as they get
older (b = 046,}? = .034).

Several interaction terms were included in each model. Findings indicate that
ﬁrst generation Mexicans who perceive that social conditions in the United States
are better than they are in their country of origin are more likely to prefer either an

American identity (b = —-0.143,p = 0.04) or a panethnic identity (b = —0.353,p =

182

0.06) versus a Mexican identity than are second or third generation immigrants with
similar beliefs. First generation Cubans, on the other hand, are more likely than
second or third generation Cubans to prefer an American identity over a Cuban
identity conditional on perceiving social conditions to be better in the United States
than in their country of origin (b = -—0.497,p = 0.000). On the other hand, ﬁrst
generation Cubans who perceive group-based discrimination to be a problem are more
likely than second and third generation Cubans to prefer a panethnic identiﬁcation

over an American one (b = 0.557,p = 0.077).

5.8 Conclusion: The Integration of Identity

Huntington’s assumptions regarding the impact of Latino political and cultural incor-
poration are questionable (Alba, 2006) at best and, at worst, they are simply inaccu-
rate. The one of interest here is that the social cleavage between immigrants, mostly
of Mexican origin, and mainstream society are so deep and difficult to cross that they
will remain a distinct and identiﬁable group. Huntington claims that “the choices
Americans make will shape their future as a nation and the future of the world.” In
his analysis, he suggests that the range of choices begins with a reiﬁcation of and
recommitment to our nation’s Anglo-Protestant cultural origins. In contrast, in this
chapter, I have shown that Latinos are very committed to Anglo-Protestant ideals
and generally hold American values in high esteem almost to the exclusion of their
own. The analysis further revealed that the structure of beliefs about being American

and American values are partially homogeneous across national origin groups with

183

some heterogeneous components. Conditional latent class probabilities revealed that
Mexicans are more likely than Cubans or Puerto Ricans to internalize ethnocultural
conceptions of Americanism, but that this class comprised the largest prevalence of
Latinos across each ethnic group. This ﬁnding clearly demonstrates that notions of
what it means to be part of American society are similar among Latinos. If reiﬁcation
of and commitment by Latino immigrants to this nation’s Anglo—Protestant cultural
origins is important for the future of American society, as Huntington suggests, these
ﬁndings suggest that there is little to be concerned about.

Previous research has shown that, when viewed as a theory of national identity,
multiculturalism fails to resonate strongly in American public opinion (Citrin et al.,
2001). This chapter extends previous research by demonstrating that similar re-
sults hold among Latinos as well, who tend to support stricter deﬁnitions of what
it means to “be an American” rather than adopting “lenient” boundaries regarding
who can claim membership in the American community. In this regard, differences
in racial/ ethnic identiﬁcation did not signiﬁcantly effect the likelihood of being in the
class favoring the “American” point of view. However, exposure to American culture
in terms of both years lived in the United States and generational status did change
opinion towards the meaning of “being an American.” With respect to the measures
of acculturation, language, generation status and years lived in the United States also
consistently increased the likelihood of an American identiﬁcation over other modes
of identiﬁcations. The strong effect of discrimination coupled with the positive effect
of acculturation suggests the need to reconsider how current theories of assimilation

apply to post-1965 immigrants from Latin America.

184

The most often-cited evidence of Latinos’ failure to assimilate into American cul-
tural life pertains to their “preference” to remain “linguistically separate” and to
retain their “home country values” (de la Garza, 2005, p. 7). Consequently, much
speculation revolves around the idea that immigrants remain more faithful to their
countries of origin, while at the same time disregarding what is best for American
national interests. Preferring Spanish over English is a primary way to retain their
cultural origins. Despite claims to the contrary, several scholars have insisted that
English is the dominant language of native born Hispanics and that immigrants, in
general, learn English. Stevens (1994), for example, argued that immigrants from
Latin American “become proﬁcient in English at a more rapid pace than immigrants
from other non-English-language countries.” As noted by de la Garza (2005), allega-
tions regarding anti-English attitudes and behavior among Hispanics are not empir—
ically grounded. He argues that rather than threaten a core American value, their
linguistic patterns are more easily seen as supporting the centrality of English to
American life. The analyses presented in this chapter tend to support this latter
argument. The unequivocalness regarding Latinos’ acceptance that part of what it
means to claim membership in the American community means speaking English is
particularly noteworthy in this regard. This coupled with the large proportion of
immigrants in the sample who are either English dominant or bilingual contradict
any claim that Latinos prefer to remain linguistically separate or retain their home
country values.

In addition, context of reception emerged as a signiﬁcant predictor of class mem-

bership. With only one exception the perception of a common linked fate did not

185

predict identiﬁcation with any group. Nevertheless, the analysis revealed that. percep-
tions of group-based discrimination seem to matter a great deal in both structuring
beliefs about what it means to be an American and in distinguishing between the
modes of identiﬁcation considered here. Notably, the perception of group-based dis—
crimination consistently predicted attitudes towards Americanism and national iden-
tity. Interestingly, in a similar study undertaken by Citrin et. al. (2001) on whites’
attitudes towards multiculturalism, he found that animosity towards minority groups
was a source of opposition to multiculturalism. The present study turns this ﬁnd-
ing on its head, by demonstrating that Latinos perceptions of this animosity is a
source of opposition to being American and Americanism. According to Citrin et
al. (2001), a symbolic meaning of multiculturalism is that minority groups receive
special treatment which they feel is unjust. Again, the current research considered
and found support for the opposite statement, namely that for many Latinos, one
symbolic meaning of Americanism seems to be exclusionary treatment that they also
perceive to be unjust. For example, Latinos are less likely to be in the latent class
favoring ethnocultural conceptions of American national identity if they believe that
discrimination against Latinos is problematic. Common sense would dictate that
anti-immigrant sentiment reinforces these perceptions and fundamentally alters the
symbolic nature of Latinos’ attachments to American ideals.

The results are in line with existing research on identiﬁcational assimilation among
Latinos in general. Similar to Schilderkraut (2004), I did not ﬁnd that direct expe-
riences with discrimination due to one’s race or ethnicity impacts self-identiﬁcation

among Latinos. In contrast, however, I ﬁnd that the perception that Latinos are mis-

186

treated as a group increases the likelihood of selecting both pan-ethnic and national
origin identities as opposed to an American identity. In addition, I ﬁnd that gener-
ational status is a strong predictor of Opting for an American identity over either a
panethnic or ethnic identity. Speciﬁcally, third generation Latinos are signiﬁcantly
more likely to identify as Americans than are either second or ﬁrst generation Latinos.
Finally, the ability to speak English is also strongly related to one’s identiﬁcational
preference: Latinos who speak English well are signiﬁcantly less likely to identify with
either their country of origin or as a Latino, all else being equal. Taken as a whole,
these results are important for several reasons. First, they support the straightline
model of immigrant assimilation by suggesting that national identity increases with
increasing levels of acculturation and exposure to the United States. In addition,
they tend to support the hypothesis that minority groups who perceive that their
ethnic origin group is discriminated against by the host society tend to form reactive
identities as a consequence of the discrimination. Importantly, the discrimination
need not be directed at the individual personally for this reactive ethnic identity to
emerge.

The fact that a two—class structure was deemed the best ﬁtting model with respect
to the measures of American liberalism, and the fact that the two classes differed only
with respect to attitudes towards governmental paternalism suggest that Latinos un-
equivocally accept the fundamental tenets of American liberalism. Claims that Latino
social values undercut other aspects of today’s sociopolitical mainstream are equally
unsupported. Indicative of this is the overwhelming support among Latinos for eco-

nomic individualism, their belief that hard work leads to success, their preference to

187

blend rather than retain culture and their tendency to be more positively oriented
toward life in the United States than in their country of origin. The results demon-
strate the importance of the context of immigrant reception, such as the presence or
absence of discrimination, to shape ethnic identity choices, particularly the preference
for opting for an American identity.

In conclusion, with respect to national identity and national interests, Latino val-
ues are not radically transforming the American polity. In contrast, the ﬁndings
presented in this chapter indicate that Latinos overwhelmingly accept core political
values instead of undermine them but that assimilation into mainstream society de-
pends on opportunities to assimilate which are effective only when they are in fact

perceived to exist (Alba, 2006).

188

Chapter 6

A Test of the Straightline, Ethnic
Disadvantage and Segmented

Assimilation Theories

6.1 Introduction

Little is known about how alternative theories of acculturation explain differences in
electoral participation among Latinos (de la Garza, 2004) and much of what is known
ﬁnds its origin from data that is now more than 16 years old (LN PS, 1989). Given the
recent surge in immigration to this country, it is unclear whether ﬁndings from earlier
studies, even those done as recently as 10 years ago, continue to hold empirically.
This chapter comprehensively evaluates the predictions of three competing theories
derived from the sociological literature on immigrant assimilation as they are applied

to patterns of Latino political participation. Classical or “straightline” theory pre—

189

dicts a positive relationship between acculturation and political behavior while the
ethnic disadvantage and segmented frameworks predict an inverse relationship. This
research expands the current literature on Latino political participation by examining
three generations of Cuban, Mexican and Puerto Rican immigrants and by evaluating
whether the predictions of these competing models hold to empirical scrutiny. Using
logistic regression models based on data from the 2002 and 2004 National Survey
of Latinos: Politics and Civic Participation, I ﬁnd substantial support for the hy-
potheses of the straightline assimilation model, limited support for hypotheses of the
ethnic disadvantage model and virtually no support for the segmented assimilation
framework. As in the last chapter, inconsistencies suggest that extant conceptual
frameworks regarding post-1965 immigrant pOpulations should be revised to include
the role of identiﬁcation, particularly as an American, as a mitigating force for po-
litical alienation. Overall, the degree to which context of reception, acculturation
and group-based resource variables explain the variance in political behavior differs
by national origin and generational status and is contingent on the type of resource
under consideration (Wong et al., 2002). A latent class analysis of Latino attitudes
and behaviors is also presented in this chapter. The latent class analysis revealed
two typologies of Latino political participation, namely individuals who are highly
responsive to the electoral system and those who are highly unresponsive.

This chapter is divided into four main sections that proceed as follows. In section
one, I test the plausibility of a multidimensional model of straightline assimilation
as originally proposed by Gordon and his predecessors. In section two, I test the

three theories of assimilation as they pertain to voting and registering to vote in

190

US elections. To do so, I consider how linguistic, structural, identiﬁcational and
attitudinal/ behavioral assimilation affect the likelihood of participating in electoral
politics, speciﬁcally voting and registering to vote. Next, I consider the role of group-
based resources and the mitigating effects of identiﬁcation and cultural retention on
non-participation with an eye towards examining the ethnic disadvantage model of
immigrant assimilation. Finally, I test the segmented assimilation theory by consid-
ering the impact of socialization and context of reception on an individual’s decision
to participate in politics. Section three discusses the results of a latent class analysis
that shows how the assimilation, group-based resources and identity variables shift
an individual’s underlying propensity from nonparticipant to participant in electoral
politics. The fourth and ﬁnal section turns to citizenship as an indicator of polit-
ical assimilation, and studies the various factors that decrease the waiting time to

citizenship acquisition once an individual becomes eligible for citizenship status.

6.2 Review of Research

As noted by de la Garza in 2004, Latinos have historically had to overcome numerous
institutionalized discriminatory barriers before they could exercise their voting rights.
The Voting Rights Act of 1975 coupled with the Southwest Voter Registration Edu-
cation Project and the Puerto Rican Legal Defense and Education Fund was largely
responsible for removing these barriers. The environment produced by these changes
coupled with increasing levels of Hispanic immigration has created the “new Latino

electorate,” which is a term used to describe the pool of potential Latino voters.

191

Despite these efforts to equalize access to political participation, less obvious and
overt barriers remain. For example, minority voters still encounter barriers to voting
in states where voters have approved English-only laws and ballot initiatives. Ac-
cording to a recent reportl, eighteen discriminatory voting procedure changes and
redistricting plans propsed in Arizona, a heavily populated Latino state, have been
overturned by the Department of Justice since 1982. In fact, protecting the voting
rights of residents in states as diverse as Arizona has proven, in general, to be difﬁ-
cult. Despite discrimination and language barriers, however, the last two presidential
elections recorded an unprecedented Latino voter turnout.

Due to the perceived impact that numerical strength will have on the electorate,
many scholars have sought to identify the various reasons Hispanics tend to participate
in politics. The traditional model of voter participation, which views participation
as being conditioned on income and other individual characteristics (Verba and Nie
1972; Abramson and Aldrich 1982; Cassel and Luskin 1988), has been the model of
choice to explain patterns of Latino political behavior (Marschall, 2001; de la Garza
2004). Early studies predicting participation based on socioeconomic indicators such
as income, education, age, gender, marital status, ethnicity and citizenship (Montoya,
2002) have repeatedly shown that Latinos participate at levels lower than Anglos
(Michelson, 2001). Later studies have found that higher levels of political cynicism,
lower rates of naturalization, structural barriers to registration and voting, and low

levels of information and interest in the Unites States political system are largely

 

1Reauthorization of the Voting Rights Act: Policy Perspectives and Views from
the Field, available online at http://www.renewthevra.org/view_from_hill/Debo.pdf,
June

192

responsible for low turnout rates (McClain and Garcia, 1993). The empirical accuracy
of the socio-econornic model has recently been shown to be lacking as a comprehensive
model of Latino political behavior. Consequently, researchers have looked to other
factors such as organizational involvement, and various dimensions of ethnicity, to
provide richer explanations of Latino political participation.

Nevertheless, the standard socioeconomic status model has proven to be a useful
starting point for explaining Latino political participation, despite the fact that it
often provides us with conflicting results (de la Garza, 2004). For example, Arvisu and
Garcia (1996) and DeSipio (1996) have shown that age is positively associated with
Latino voting while Hritzuk and Park (2000) found no association. In addition, Arvisu
and Garcia found that education does not consistently result in higher turnout rates,
and this is true especially among Cubans (de la Garza, 2004). DeSipio (1996) reports
that national origin is not a signiﬁcant factor distinguishing voters from nonvoters.
There is disagreement regarding the effect of marital status, gender, employment
status, home ownership, and metropolitan residence on the immigrant’s propensity
to vote (Bass and Casper 2001, DeSipio and Pachon, 2002). These inconsistencies
demonstrate the overall complexity of Latino voting patterns, with potential subgroup
differences further complicating models of their participatory behavior. For example,
the consistently low turnout of Puerto Ricans compared with other groups has been
attributed to their status as sojourners. Puerto Ricans are less politically involved in
the United States than they are in their country of origin. Explanations have focused
on lack of mobilization (Dominguez, 2002) and on the relative political alienation

of Puerto Ricans in the United States as compared with Puerto Rico (McClain and

193

Garcia, 1993).

Due to the growing influx of immigrants, scholars in recent years have turned
to the question of whether turnout is a function of the continuous incorporation of
citizens who are not fully socialized into electoral politics. Increasing levels of political
socialization have been shown to increase political participation, particularly voting
(Leal, 1999). Given the importance of religion in Hispanic life, several researchers have
also turned their attention to the role of the church as a key actor in the political
socialization of Latinos. On this point, as well, the evidence is conﬂicted. In one of
the ﬁrst studies on this topic, Verba et al. (1995) found no evidence to support the
claim that religious organizations serve a socialization function. This nonﬁnding was
later supported by Hritzuk and Park (2000). Recent work challenges these ﬁndings by
arguing that church attendance, regardless of denomination, is positively associated
with higher turnout, but that differences in denomination explain little about Latino
political participation (Jones-Correa and Leal, 2000). Jones—Correa and Leal (2000)
further argue that in the absence of other institutions, churches are disproportionately
important to Latino civic life and, given their ethnic nature, Latino parishes have the
potential to serve as centers for political mobilizing. A study undertaken in 2002
further complicates the role of religion on Latino political behavior. This study found
that neither skills learned through participation in church activities nor church-based
mobilization signiﬁcantly affects turnout. The only religious variable that had a
positive effect was experience as a. born-again Christian.

Montoya (2002) demonstrated signiﬁcant differences in voting behavior between

men and women between and within Hispanic subgroups. For example, she found

194

that voting patterns among Mexican men and women are more similar to each other
than are Puerto Rican or Cuban men and women. Mexican women are more likely to
participate when they have additional monetary resources, as they grow older or as
their proﬁciency of English increases. In contrast, these factors are not as important
for Mexican men or Puerto Rican and Cuban women, whose voting behavior gener-
ally depends on the extent and quality of work skills. Organizational involvement was
the only statistically signiﬁcant explanatory variable that consistently explained the
behavior of all Latinos and other researchers have substantiated the role of organiza—
tional involvement in motivating participation. Diaz (1996), for example, found that
Puerto Ricans and Mexicans involved with organizations vote at signiﬁcantly higher
rates than those who are not members of any group. Organizational involvment had
no effect on Cuban participation, however.

Paradoxically, Latinos on the whole seem to be more concerned about issues in
the United States than they are about political issues related to their country of ori-
gin (Dominguez, 2002). Nevertheless, many Latinos are less politically involved in
the United States than they are in their country of origin due to a lack of efforts to
mobilize them and to their relative alienation in the United States as compared with
their home country (McClain and Garcia, 1993). Substantial differences in educa-
tional attainment and income explain, in part, these differences in political behavior.
Generally speaking, Latinos are economically disadvantaged relative to other groups,
with poverty rates being at least three times higher among Hispanics than whites
(Wildsmith, 2005; Ramirez, 2000). Partial and/or incomplete socialization into the

electoral system may be an additional factor in explaining low turn out rates and this

195

lack of socialization is exacerbated by the continuous inﬂux of immigrants (Hritzuk
and Park, 2000). Feeling politically efficacious, viewing voting as a duty or symbolic
act rather than as an instrument of change and interest in and knowledge of politi-
cal processes have all been shown to be signiﬁcantly related to voting among Latino
immigrants (de la Garza, 2004).

Much of the literature on minority political participation emphasizes the impact of
socioeconomic variables on political participation and the link between mobilization
and participation but relatively few studies have explored the relationship between
group consciousness and political behavior (Stokes, 2003). Additionally, most of the
research in this area is based on African American participation. Shingles (1981) was
one of the ﬁrst scholars to provide an explanation for the effect of group conscious-
ness on political participation among Blacks. He argued that black consciousness
contributed to mistrust and a sense of internal political efficacy which encourages
policy related participation. More speciﬁcally, he found that individuals who feel
strongly efficacious and do not trust the government are likely to take part in activi-
ties requiring a high degree of personal initiative. Later studies on black group con-
sciousness have doubted the relationship between group—based resources and political
participation and question the relevance of group consciousness. Bobo and Gilliam
(1990), for example, suggested that given changing race relations in contemporary
society, group consciousness has taken a back seat to local representation as a critical
component of Black political participation. Tate (1993), accepting the role of group
consciousness as a relatively important predictor of participation, noted that some

types of group-based resources may be more important than others. She suggested,

196

for example, that participation in an ethnic organization may be more important to
Black participation than group consciousness alone.

Building on the work of Tate (1993), scholars have begun to examine how group-
based resources matter for political activity. For example, Chong and Rogers found
that their multidimensional measure of group solidarity was modestly related to elec-
toral participation but had a powerful relationship to participation beyond voting.
Both Chong and Rogers (2003) and Stokes (2003) found that different aspects of
group consciousness matter to different degrees for distinct types of political activ-
ities. In a similar study pertaining to Asians, Wong et al. (2005) found that the
effects of group consciousness, group identity, sense of common culture, membership
in an Asian American organization, and experience with discrimination on registra-
tion and voting appear to be relatively weak and inconsistent. Nevertheless, similar
to Chong and Rogers (2003), they found that having a sense of ethnic group con-
sciousness and being a member of an ethnic organization were positively associated
with participation beyond voting. They reasoned that Asian American organizations
provide opportunities to participate and are powerful agents of Asian political par-
ticipation. Notably, adopting a panethnic or ethnic identity had no effect on political
participation while having strong ethnic ties to one’s country of origin increased the
likelihood that Asians would participate in activities related to politics in the United
States.

The literature is conflicted as to how group consciousness effects minority political
participation. Miller et al. (1981) argued that a perception of self-location within a

particular stratum and the psychological feeling of belonging to that stratum fail to

197

capture the psychological dimensions of participation (Stokes, 2003). Using a multidi—
mensional concept of group consciousness that includes feelings of power deprivation
(polar power), relative dislike for the outgroup (polar affect), and the belief that
inequities in the social system are responsible for a group’s disadvantaged status in
society (system blame), they found a strong relationship between group consciousness
and political participation. Uhlaner and her collaborators similarly found that ethnic-
ity does matter for political participation and suspected that culturally speciﬁc factors
may be responsible for observed differences in Asian American participation. On the
other hand, Wilcox and Gomez (1990) did not ﬁnd a relationship between group con-
sciousness and participation among African Americans. Leighley and Vedlitz (1999)
found that group consciousnes fails to account for political engagement among racial
and ethnic groups. Finally, as I discussed above, other researchers have found that
although group consciousness is unrelated to electoral participation signiﬁcantly pre-
dicts participation beyond voting (Wong, 2005; Chong and Rogers, 2003).

Scholars generally assume that the dimensions of ethnicity among distinct sub—
groups of Latinos are signiﬁcantly different due to socioeconomic, historical and cul-
tural differences. Lien (1994) developed several dimensions of ethnicity including
acculturation, attachment to ethnic culture, and discrimination which had two com-
ponents alienation and deprivation. He found that Asians and Mexicans, two groups
with different socioeconomic, historical and cultural characteristics produced similar
psychological proﬁles when ethnicity was viewed as an “emerging pheonomena formed
through a multidimensional process — variously called ethnicization (Sarna, 1978),

racialization (Omi and Winant, 1986), or ethnic Americanization (Fuchs, 1990).”

198

Might hr'lexicans, Puerto Ricans and Cubans have similar psychological proﬁles de-
spite their significant differences? Might these proﬁles impact political participation
in similar ways? These are some of the research question that are addressed in this
chapter.

Several additional limitations can be identiﬁed in the existing scholarship and
are addressed in this chapter. First, Portes and Rumbaut posited the phenomenon
of reactive ethnic identity as a generational effect, yet the existing studies do not
consider generation as an explanatory variable. A second limitation is that identity
involves more than ethnic identiﬁcation yet studies have largely overlooked how the
awareness of one’s status as an oppressed group affects political behaviors (Wong et
al., 2005). Third, since ethnic origin groups have different historical contexts, it is
important to perform separate analyses wherever possible, yet, due mostly to data
limitations, no study to date does. Finally, although Schildkraut’s (2004) study is
noteworthy, her model does not speciﬁcally test the hypothesis she posits, namely that
reactive ethnicity mitigates the political alienation that results from the perception
of discrimination. In order to test this hypothesis, one needs to examine the effect
of individual and group level discrimination on political behavior and attitudes for
those who assume a reactive ethnic identity and those who do not. However, this
suggests the presence of an interaction term: any construct measuring identity and
discrimination in level form merely measures the independent effects of these variables
on political behavior and not the joint effects of both variables on political behavior.
In the analysis that follows, I address these limitations by: (1) adding generational

status, years lived in the US and measures of group consciousness and resources to

199

the models; (2) developing ethnic identity more comprehensively; and (3) adding

interaction terms where appropriate.

6.3 Assimilation as a multidimensional concept

This section tests the plausibility of a multidimensional concept of assimilation for
Latino immigrants as suggested by straightline assimilation theorists. Therefore, the
original hypothesis that assimilation is a four-factor structure comprising: ( 1) identi-
ﬁcational; (2) attitudinal receptional and behavioral; (3) cultural; (4) structural and
(5) political components is tested against the alternative hypothesis that assimilation
is best conceptualized as a unidimensional construct. See Appendix A for further
details regarding the variables used, the model, and the ﬁt statistics.

Overall, the model supports the hypothesis that assimilation is a multidimen-
sional concept comprised of the ﬁve facets of structural, attitudinal/ behavior, identi-
ﬁcational, cultural and political factors. While this ﬁnding supports the straightline
model, it does not rule out the possibility that as Latinos interact with the dynamic
social environments in which acculturation takes place, differential patterns of asso-
ciation may lead to segmented outcomes (Portes and Zhou, 1993). In other words,
the results do not rule out segmented assimilation theory as a possible outcome of
the assimilation process but a failure to conﬁrm the multidimensional nature of as—
similation would a priori rule out the straightline model as an adequate theory of
immigrant assimilation. In what follows, I focus on examining several key measures

of acculturation and on delineating the patterns of Latino acculturation with an eye

200

towards explaining political participation.

6.4 Results

Before proceeding to the multivariate results, it is instructive to examine the distri-
bution of the independent and dependent variables used in the subsequent models.
Table 6.1 shows the group means for the measures of assimilation, acculturation,
group—based resources, contextual effects and control variables used in the analysis.
As can be seen from the table, Cubans have established stronger ties to America rela-
tive to Mexicans as reflected by the longer number of years they report having lived in
the United States as well as their stronger ties to the United States, on average. The
majority of Mexicans, Puerto Ricans and Cubans are foreign born, ﬁrst generation
immigrants. Slightly under half of the sample of Puerto Ricans report proﬁciency
in two languages (.48) while Mexicans (.49) and Cubans (.51) are more likely to be
Spanish dominant. The plurality of Mexicans, Puerto Ricans and Cubans all prefer
to identify with their own ethnic group. Nevertheless, a fairly substantial propor-
tion of Latinos prefer other modes of identiﬁcation. Puerto Ricans (.26) and Cubans
(.22) are more likely than Mexicans (.19) to prefer an American identity over either
a panethnic or ethnic group identity. Mexicans (.13) are more likely than Puerto
Ricans (.09) or Cubans (.10) to believe that Latinos share one culture, the basis for
the measure of linked fate. Cubans are less likely than Mexicans or Puerto Ricans
to believe that group based discrimination is a problem but all three ethnic groups

report similar average levels of individual—level discrimination. Almost one-third of

201

Puerto Ricans report being “born-again” Christians while only 13% of Mexicans and
19% of Cubans report this type of religious association. About half of all l\"‘lexicans,
Puerto Ricans and Cubans earn less than $35,000 per year. Cubans more than Mex—
icans or Puerto Ricans are structurally integrated in the US economy as indicated by
their higher reported levels of economic attachment. Cubans and Mexicans report
that conditions in the US are better than in their country of origin. Additionally,
an overwhelming majority of Latinos believe that racial discrimination is a major or
minor problem in the United States. Mexicans and Puerto Ricans are signiﬁcantly
more likely to be a victim of discrimination (meaning themselves or their family) than
are Cubans. Cubans tend to be slightly older on average than Mexicans or Cubans.
Exactly half of Mexican and Cubans are married compared to approximately 48% of
Puerto Ricans. Finally, a slight majority of Mexicans, Cubans and Puerto Ricans are

female.

202

 

 

as 8.: :m 3.2 $2 $2 885555 354-3223
as was :m was as: Es 883855 889880
mach—Omega magnum-3:20
mam mos :m a: E: a: sameﬁssa. eeoaoom
2: so” a: was an» as 323 85 Esaeoo
:. was a s2 as as. 85m swam 52: a3
a; was s: as. aw uses 833 52: s3
:Omuw—mammmdw ~dhq—uoﬁhum
omm cam in was as was 8589953
:domnvadw
8:3?me 38389303
mg as I .. 8m. 8e 3 2: 8 ea.
new as Sm 052 32 as? season €8an
a...” 5.... 3.. came E: oak 335$
mam $2 Sm oak” E: Rem Essence saga
2m :2 so” we: 82 9:: 882863
EN 3mm x - m2 8.: m: a: s 85 28>
a gm 3 s2 as was m 8:880
8 was as can” as $8 a 85550
:m came a: 058 man came a 83550
Gomuw~mammm< ~dh3350 Vim GOSQHSH—SUU4
Z Gamma/H 2. Cde 2 C602
ﬁdﬁﬁo Gmomm Ouumﬁm CdomeE m~£d€d>

Emmao EcoSaZ am :oSQooom mo 38250 was 2035:8834 mo $333.4 .coswusﬁsoo/x Ad 2nt

203

mam:
mama
mam:
mag

mam

aha

mom

omm

omm

mmm

mac

Roam
oat
oamﬁ
3A

mam:

0&3

Rune

cam:

$3

$2
mama
momﬁ
$3

omm

Sm

cam

3m

Em

mam

sac
oamm
“Raﬁ
mama
me.

oaaa

came;

Kmm

ea:

Rea

mama
mama
mama
mama

mna

moa

Ea

Hmoﬁ

Hmoﬂ

503

Kb
058
$3
Room
mag

“Roma

053

mag

oama

Xmﬁ

355m 3 2:85 mo 23m
3356 38.8%
Eom swatch 388m

0283mm.“ 385m

5 e 5:3

nomuaooem mo 88»on

cwsmtso nﬁw<é8m
mmocmzomomcoo Eomazom
9650 085m
:oﬁmowsﬂsg magnum
2:33ch
cosswowsamg basin

sea Essa

204

Table 6.2 shows participation across a range of measures of political incorporation
by Latino subgroup. While previous research ﬁnds that Latinos register and vote at
rates much lower than other groups, these data do not support this contention. In fact,
despite the large percentage of individuals who are not citizens and therefore ineligible
to vote, a comparison between foreign—born and US. citizens reveal that naturalized
Latinos register and vote at rates higher than their US. born counterparts who are
citizens by birthright. Eighty-two percent of foreign born Latinos are registered to
vote, compared to 78% of US. born individuals. Similarly, 76% of foreign born
Latinos reported having previously voted in a US. election compared to 72% of US.
born Latinos. While this result could possibly be due to measurement error in the
data, other researchers have observed a similar discrepancy between foreign born and
native born populations. For example, Wong et al. (2004) found that foreign born

Asians participate at higher rates than their native born counterparts.

205

:oﬁwbmwmom wad wcﬂo> mo mmusmmoﬁ E UmUEoE 0.8 2336 Eom swatch ch H

 

 

 

 

 

 

5, cm as E mam a 2: Q, 3, E. 33 S 88 88>
E. 5 So a mam a 2,: S. we, mm Ea; mm 88 88>
one E 3% E E mm 0mm 8 2e 9. $3 E 88>§m
.03 mm 9% wk. mg a R we 5% E 2,3 cm Beaaé
53 S wow 2: Sm 8 ma 2: :5 mm 23 3 856
2 ca 2 05 2. a Z 05 Z 05 2 05
1:03 Gwmmuorm Guom m3 633:0 Gmomm 09.85% Qdomxwz 1305

 

 

85an wcoﬁ< aosmawosnem Reason “am 3nt

206

Electoral Participation Among First, Second and Third Generation Immi-

grants

The first set of questions to be addressed in "the multivariate analysis pertains to how
acculturation and exposure effect Latinos’ participation in electoral politics. Table 6.3
presents the logistic regression coefficients pertaining to these variables for Mexicans,
Puerto Ricans and Cubans separately. As can be seen from the table, the results
strongly support many of the predictions of straightline assimilation theory. For
example, exposure to US culture across generational status has a powerful effect on
turnout and registration for Mexicans and Cubans. Second and third generation
Mexican and Cuban immigrants are signiﬁcantly more likely to have voted in a US
election than ﬁrst generation immigrants, holding other factors constant. In addition,
linguistic assimilation is a strong predictor of electoral participation. For example,
the odds of voting and registering for English dominant and bilingual Mexicans and
Cubans are signiﬁcantly higher than for Spanish dominant Mexicans and Cubans.
Both economic attachment and income are signiﬁcant predictors of Latino political
behavior but their effects differ by subgroup. Among Cubans, income is signiﬁcantly
and positively associated with voting and registering to vote. On the other hand,
economic attachment has a relatively larger effect than income among both Puerto
Ricans and Mexicans. Among Mexicans, for example, each additional increase in
economic attachment increases the odds of registering by 58% ((errp(.459) — 1) x 100)
and the odds of voting by 75% ((exp(.564) - 1) X 100). Educational attainment

was only an important factor in explaining voting behavior among Puerto Ricans.

207

Turning to the effect. of identiﬁcational assimilation, primary identiﬁcation as an
American had the largest effect on the probability of both registering and voting
among Cubans. The odds of voting and registering to vote for Cubans whose primary
mode of identification is American are signiﬁcantly higher than they are for Cubans
who prefer a panethnic idenitity. On the other hand, there is no signiﬁcant effect
for either Mexicans or Puerto Ricans. Age is signiﬁcantly related to the probability
of registering and voting among all three Latino subgroups. With respect to marital
status, the results show that married Mexicans and Puerto Ricans register to vote at
higher levels than do those that are unmarried. Marital status has the largest effect
on Puerto Rican and Mexican voting behavior. Puerto Ricans who are married are
almost 4 times as likely to register to vote as similarly situated Puerto Ricans who are
not married. The effect for Mexican electoral participation is also positive but smaller
in magnitude. Among Puerto Ricans, too, being female was found to signiﬁcantly

increase the likelihood of both voting and registering to vote.

208

 

 

 

Ev. 13a. cg. amc. vaa. ..cImam. Hcm. ......caA c3. ccv. amm. ccH. Baamm
a3. cma: a3. mac... amm. ..chm. mam. ......KcmA a3. mmv: :m. Imam. BEES
ch. *icmc. aac. ***m3. m8. ***cac. ch. iavc. cch. inmaﬁ. mac. *imac. mm<
mozwmuﬂr 35:00
a3. va. cmH. :c. mac. a3: c2. vmc. mmH. awc. SH. 1.2.- m3 5 :36m
Gomommooﬂ .5 38:80
vac 13v. mHa. *mnm. aHa. *ivcn. ma. aca. am. Fwd aaH. ***amw. Emanomﬁxw 2:5:oum
«ma. *mcm. aca. ***mmc. maH. ....Iacm. mma. *Iaaa. nca. mama. amH. ammo. :oSaoscm
wmc. vmmc. mmc. ***ccac. vac. 12:. amc. :cr mamc. *ccc. mac. ***cac. 9:85
:omumzﬁmmmd.
Ryan—055m
mmw. mmar cmca imﬁa mac. ccc. Hmv. *cE..- ch **aa.a mHm. cmc. DH :doﬂo::<
Gown—w—mammmd.
Ecomuwucmacma
c5. amH. mwa. ﬁlm»; awm. 1.72.4 cmc. EA. cam. *iamA amm. ***m.c.H Eswﬁzm
ecu. amc. ccc. 13a; ca. 1.34 mm. Em. 2c. *mHA wan. ***cc.a .EeEEoQ ﬁzwﬁm
cﬁc. *cma Hm; mag mcv. ....ILAA mmm. acc. cwc. HS. ccm. iccm. :oﬁauogmv BEE
mav. mca. SA *mma amm. ***mm..H a3. ama. cmvn. *mca. ccm. *3.ch 9532080 988m
somewzammmd‘
ﬂaw 9555.3930ko
.w.m Q .m.m Q .o.m Q .m.m Q .m.w Q .m.m Q
mm 53:0 53382 mm 55:0 :52on
UBO> cgmummmmm

moEES> 35:00 98 53:2: dosswzammma‘ :0 :oSwSmmwom cg“ w:€o> .«o :ommmoﬁom osmmwoq m.c 2an

209

 

a. Var... .3. var. .8. Va...

$520: v.5 Echo 335QO “53.8%: ”$qu

 

E. mm. mm. ow. 5. mm. @8285 .8850 285
amm. m3. Es. Em. amm. 8m. E823
25. 08. 8o. 8o. 25. 8o. ax A a
saw as mom EN.- :3 8w 2

:3 o..m- 8.“ ...-Bow. 8.: ...,-was- 8.: oas- mw: ...-one- 2: 1&3- 25250

210

Voting and Electoral Participation Among the Immigrant Generation

In order to assess the effect of exposure to US culture within generational status, the
same model was re—run with the addition of years lived in the United States and ties
to one’s country of origin? Table 6.4 shows the results for foreign born Mexican and
Cuban immigrants. For example, among l\-"Iexicans, in addition to education, income,
economic attachment and American self-identiﬁcation, years lived in the United States
and decreasing ties to one’s home country signiﬁcantly predict registration status. For
each additional year lived in the US, the odds of voting and registering are increased by
a factor of 1.13 or 13% for Mexicans. Each standard deviation increase in years lived
in the US increases the odds of voting fourfold. Additionally, among ﬁrst generation
Cubans, decreasing ethnic ties and years lived in the United States are positively and

signiﬁcantly related to the probability of registering to vote.

 

2The questions pertaining to length of residency, ties to country of origin and
relative perceptions of US versus country of origin were only asked to respondents
who indicated they were foreign born. Therefore, it is impossible to determine how
years lived in the US effects voting and registration for second and third generation
immigrants. Nor is it possible to examine this variable for Puerto Ricans, who are

residents of the US by birth.

211

 

830:3» 53:03 «Ea :035:ﬁ:00<

6036.2?» 03mgw080c080m £0305 835:8, 35:00 $50 @302

 

 

 

 

9:. :5. was. a:- mm:. as. 2:. Ea- acosasé 35332
Ema-HO
as. time.- SN. :35,- mmm. Es.- me. town.- 3 $550 8 may
go. :22. EN. :33. 9.5. its? as. :32. ...-saw was: a: s
@034 23>
...; a .0.m a .0.m a .0.m a
53:0 :0252 $330 53082
:0302m m: E 88> 80> 0a cosumwmom

 

BEEHE/ 03mgm0§0c060m 98 55:03

:053:E$< :0 :03mbmﬁmmm EB m:$0> a0 :ocmmmoEom osmﬁoq ”ac 0308

 

212

In this section, I have demonstrated that exposure and increasing ties to the
host country unequivocally predict Latino political participation in American politics.
Most scholars, however, distinguish between increasing ties to the US and decreasing
ties to one’s home country, noting that the two are not necessarily mutually exclusive.
Above, I showed that an examination of the hypothesis that strong ties to the home
country has no effect on electoral participation reveals that it must be rejected among
the respondents in this sample. In other words, Mexican and Cuban immigrants with
strong ties to their country of origin are signiﬁcantly less likely to participate in
electoral politics. For example, the change in predicted probability of voting in US
elections over the range of ethnic ties is .176 for Mexicans and .148 for Cubans, holding
other variables constant at their means. In comparison, the predicted probability
of voting for Mexicans with strong ties is .042 compared to .011 for Cubans. To
investigate the interaction between identity and ethnic ties, I plotted the probability
of voting while allowing ethnic ties to vary over its range. The result for Mexicans is
shown in Figure 6.1. The figure provides visual evidence that as ties to one’s homeland
decrease, the probability of registering and voting in US elections increases. Among
Mexicans, those with weak ties to Mexico and who identify primarily as American

have the highest overal predicted probability of voting.

213

: 38:83: :0500c583 3 00:82 on. moat USN m:0303m m3 5 w:$0> “0 3230395 08200.5 :6 Esz

0885. 3 8: m8. .0 935$ 0.5 8sm> 938:. ”202
8g 03...: 05 mug... 83.05
a m m e
, I r 8...

 

 

2.6
, cud
. and
£36
.86

8: n as: - 8...

033:3... Bodega .
33:85.3... Seams: a:
8.0

 

 

§§_§829§=<88>ba_382a 863: gees.

214

Yet another important question to be addressed here is how sustained political
contact with one’s home country impact participation in US politics, a question that
has not been rigorously addressed in the literature. Tables 6.5 and 6.6 provide output
from a regression of citizenship and voting behavior in one’s home country on level

of acculturation, assimilation to the US and sociodemographic characteristics.

215

 

0030?: 0:0 muotm 0000:00m .0030m 0000 020005000 :06000w0m 0300004

 

 

 

no». ommo. mooo. moo. AEmo. Svor omm. vmoo. Hooor 09$
03008:? 33:00

ono. CE. Emo. mod owov. wvmo. to. $3. mmoo. 0050003003 0:300:0m oz
:omuwmammmdw _0:0$00E0:00H

mom. mmmw. owmm. woo. moww. Sow. mvo. wmom. ovoor 00000 00:03 0000:30>

m5. omE. mono. ono. oomv. omoﬁ- ono. nmom. ovomr #02955

oov. ooEu ooHo. omm. Sow. moon: to. «Sm. «own. 0:500 E00000

0mm. moo: SQ: o:. 3:. 23.- Hoo. owoo. mommr :MwEO mo 00.05000 00 00?

I I I I I I I I Hoo. wooo. ommor m3 5 00000»
:omumzammﬁw 0:0 :omuwusﬁsﬁﬁx

woo. homo. mvoo. H2. wwmo. vowor 5mm. omﬂo. Hoﬁo. 0:805

vow. no: moo: Hoo. wood. uEmo. oom. oovo. ammo. :050000m
:030_m:bmm< 12:00:03

woo. wﬁoo. oom.m m9. oomm. mmmw. omo. onH. ovum. :0000

moo. Moo: 23. wow. Noun. 3%.. omw. momm. Hovo. 0005 8:0an

moo. 0mm». mﬁoa- So. Bow. wow? ovo. out. Hmomr 002082
95.10 5300 10:05.02

0 .00 8.0 m a .00 M00 0 a .00 .md n
:00 0.5 :00 0:m :00 00H

 

33:00 080m 000300 :0500:0EO Eomﬁoom 0EOE0> 0:00:0Q0Q ”no 030%

 

216

 

.006 20800. 2080 00 0:0 000E030 000 00008500 00000 0030350003 0000005

 

00 00504 008000 5:002:00 smmc00m ”00203000 003200 00000 00000000 000000 603??» 30502030 ”00002

 

 

I I oohwwo- owoz- So-
I I ooo moo z
000. ooo. ooo. :o.- ooo. ooo. ooo. 03o: 00.0
Q—ﬁwmhwxr ~0huﬁoo
woo. ooo.- woo. woo- 0:. ooo.- ooo. ooo.“. 8080052 290200 02
030008003. 38:80:82
owo. ooo.- ooo. owo.- oow. woo. ooo. oooo. 00 8 5300
ooo. ooo.- ooo. ooo.- ooo. ooo.- ooo. Howo. mo s 85 28>
go. 0: Bo. o3 ooo. ooo. ooo. oooﬂ. 2330 80
0:. 0o; ooo. of- ooo. owg- ooo. wooo- 050050
ooo. ooo.- ooo. 000. I I ooo. 33. 005800 020:0
:Omuwﬁmgmmmdw Una—w ﬁomumhnﬁzﬂoo<
ooo. 02.- ooo. ooo.- ooo. woo. ooo. ooo.- 0583
mow. to; woo. oow.- oww. ooo. owo. 2on 808000
:Omuwmammméw umhﬁuoﬁhum
0 0.0 0 o 0.0 0 o 0.0 0 0 0.0 0
00 88> 00 8000 00 88> 00 52:0
muammwwEEH c.0350 wuswuwmeam £60382

 

005. Rooﬁom 00030000008 Moo 0308

 

217

The results clearly show that as years in the US increase, political ties to one’s
home country decrease among ﬁrst generation immigrants (b = —.O23,p = .001).
Similarly, first generation immigrants who prefer to retain their culture rather than
blend into American society are more likely to be politically active in their country
of origin, other factors equal. Mexicans are less likely to claim Mexican citizenship
with age, income, as proﬁciency in English increases and as they come to identify as
a member of a panethnic group. The only signiﬁcant variable pertaining to voting
behavior was years lived in the US, which decreases the probability of turnout in
Mexican elections. According to Figure 6.2, both English and Spanish dominant
Mexicans are less likely to vote in Mexican elections with increasing exposure to US
culture. However, the steeper slope for the Spanish dominant in Figure 6.2 suggests
that exposure has a greater effect for Spanish dominant compared to English dominant
Mexicans even though the English dominant are less likely to vote in Mexico at all

levels of exposure.

218

 

0030052000< m9 900 mnoﬁoﬂm 0.02082 5 w500> o0 bawﬁdaem ﬁwuomvmum .m.w 003mg

568.8... .329 §<Pa§o§z
gas 02:0 830.539.0008. $8.00 8&6: =_ $5080 032
gugognﬁoﬁagsgtgﬁ

 

 

 

om . ow . 8 . on o. o
l I ﬁll _ L 8.:
.02
as
ooo. n 050
moo. .I. 0&0 - a...
05.1600
.80
8.. A
5283+ EranﬁiﬁI? £389:
88?:
038mg. 5000.05 009:4

80:80 p.808: 5 93> .0 5:332“. 3802.".

219

6.5 Group-Based Resources and Latino Political

Participation

Wong et. al (2005) discuss the distinction between group identity and group con-
sciousness. In particular, whereas primary identification merely means that an indi-
vidual identiﬁes with one group more than another, group consciousness implies the
awareness of shared status with other members of the panethic community as well as
a strong sense of solidarity (Wong et. al, 2005; Bobo and Gilliam, 1990). Although
the literature portrays these concepts as independent, the indicators of group-based
resources may indeed prove to be related (Wong et. a1, 2005). To examine these
relationship, if any, I ﬁrst explored the bivariate correlations between the variables
measuring group consciousness and identity. Table 6.7 shows that those with a strong
sense of panethnic group consciousness are signiﬁcantly more likely to identify primar-
ily as Latino than are those without such a sense (p = .0075). Similarly, individuals
who report having a strong sense of panethnic group consciousness are significantly
less likely to identify as American (p = .0254). Nevertheless, these measures are not
highly correlated indicating that although the two variables are related, they can also

be distinguished from each other (Wong et. a1, 2005).

220

.S. W a is." was mo. IV. a "In .3. W a ... em 83$, ax Emowawmm gamma?” mxﬂoumxx

mmwgcoocma 38 m5 $3.28 :90 ”Hm—BO Z

 

 

 

 

mama Em nmo Z
0503 mm om 05mm 02
0503 mm Hm 05mm 3%
3:83.22“ umﬁmwoa @5203 m3
mocﬁﬁoo Engage Soc messed
@002 on S Ram oz
oboe vm ”Inca 13.3..ng m®>
3315 «:8 BER
85:28 economﬁ 89a messed
130% EwEO Mo 33:00 magnesia owcmamﬁoﬁﬁmq

 

55:92 macaw waEm wwwpcwocom

mmmcmsoﬁmcoO @580

 

nmocmzomomcoo Q5920 was 33.53 9.20 :83th QEmsoEEmm ”Nam mEaB

221

6.5.1 Bivariate Relationships Among Group-Based Re-
sources and Political Participation

Is there an association between the measures of group identity, group-based resources
and political behavior among Latino subgroups? Wong et al. (2004) asked a similar
question with respect to Asians. In contrast, with respect to Latinos, the results
suggest that Cubans who exhibit a strong sense of panethnic group consciousness are
signiﬁcantly less likely to register and vote than those without a strong sense (see
Table 6.8). Mexicans and Cubans who prefer an American mode of identiﬁcation are
signiﬁcantly more likely to vote and register than are those whose primary identity is
not American. Puerto Ricans who prefer to identify as an American are signiﬁcanly
more likely to vote than those who do not prefer such an identity. Mexicans and
Cubans who prefer to identify with their panethnic group are less likely to register
than those who do not. Puerto Ricans, Cubans and Mexicans who identify primarily
with their own ethnic group are signiﬁcantly less likely to vote than those who do not
and the latter two groups are also signiﬁcantly less likely to register to vote. Among
Mexicans, volunteering to an organization speciﬁcally representing Mexican concerns
is associated with higher levels of registration and voting. Similarly, belonging to
an ethnic-based organization tends to increase levels of voting among Puerto Ricans.
Mexicans who report experiencing discrimination are more likely to both register to
vote and vote.

Clearly, group-based resources are associated with more participation at the bi-

variate level but the effect of these resources for each Latino subgroup is quite dif-

222

ferent. Are these relationships maintained once potentially intervening variables are

considered, such as English language, socioeconomic status and generational status?

223

 

Vb ow wv MK. V5 gov
33.3 .1330 ***mm “Item .2. "......BARN
ow Np hm no on. Room
mm. an ion mm mm. *okcmm
mm. on on vm mm. gmm
***m.w ....Lw .1335 fiww E. 1.35:.
Ho Nb mm we mm. “Room
ow MK mm mo .K. Ream.
ac MK. on no wk “Nowm
"ISM mm mm ﬁnale C. Room
55:0 udowm 3.8an :mowxga 5330 “=me 8.8an SUCSSN

 

sonata .m: a 88>

30> 3 vocmummmmm

@580 253mm 3 wquoEn
:oswﬁcwwho 8m Umummucﬂo>
oz
8%
00 3353 395.5
02
8%
OEEA mm .3983 baatm
oz
3%
caoiwﬁxx mm .3353 ids—Em
53:3: Q5910
oz
mm;
3:83:09 cmnumwou magic? 0.8
85:58 unaware 80b managed
oZ
mw>
23:6 28 2.2? 823:8
usmuogw 80¢ 85qu

mmwcmzowomnoO 9.qu

 

 

mohsmwmz nowuwowSQmE QSEU ES 5in mo mun—580 .3 msﬁo> was :oﬂmbﬂwmm .w.© 033.

224

.5. W A u5.3.n was me. W a in .2. W a * em 83$

«X Ea£8mﬁ 39358 mﬁpmuma»

 

mm
mm

mm
on

E.
E.

nw
*3.

mm
***mv

E.
**mw

so
mm

mm
vw

E.
E.

R
vw

mwmm
35253”

ORE.
“.31,me

OZ
8%
95.10 umﬁmwdx 2039555me
OZ
3%

225

Multivariate results of the effect of group consciousness and identity on voter
turnout and registration is presented separately for Mexicans, Cubans and Puerto
Ricans (see Tables 6.9, 6.10 and 6.11). For Mexicans, self-identification as an evalen-
gelical Christian is positively related to registration and voting behavior. The odds
of voting among lVIexicans who are of this type of faith is 1.92 times higher than it
is among Mexicans who do not hold such beliefs (Table 6.9). In addition, the per-
ception of group-based discrimination is shown to increase the probability of voting
in US elections while personal discrimination increases the probability of registering
to vote (Table 6.9). An examination of the contrast between adopting a Mexican
versus an American identity reveals that the former decreases the probability of both
voting and registering to vote (Table 6.9). Likewise, adopting a panethnic identity
over an American identity also decreases the probability of voting (the table shows
that American — CO is positive and therefore the contrast CO - American is nega-
tive). Turning to the effect of group-based resources on Cuban participation, it is
apparent that the perception of a common linked fate decreases the probability of
Cuban participation with respect to both voting and registering to vote (Table 6.10).
In addition, primary identification as Cuban or as a member of a panethnic group sig-
niﬁcantly and consistently decreases political activity. This is shown by the positive
coefﬁcient on the contrast between American versus country of origin and American
versus panethnic identiﬁcation in Table 6.10. The results with respect to Puerto Ri-
cans were not as informative. Nevertheless, we again see the role of discrimination
in motivating political behavior (Table 6.11). The perception of individual level dis-

crimination significantly increases the likelihood that Puerto Ricans will register to

226

vote (Table 6.11). In addition, adopting an American identity over a Puerto Rican

identity signiﬁcantly increases the probability of registering to vote (Table 6.11).

227

 

amm. H am 8%& mean. H mm 88mm
8.0 u a .88 u ax 9%, 8s n a £2 n «X e33

 

 

 

man u 2. man u 2
3a. is: amm. a: 3:52 25:qu was; :85st
mam. wmm. com. 2:. 3:52 00 e, saﬁgﬂ
amm. :3: RN. £ng 3:82 00 9 $253.
53:3: 95.10
So. «so. Go. 3.5. 5358829 382$
So. tome. one. one. assassin Bamdzoé
as. as. Sm. tame. seamed aozmmcgm
m2. m2 .- E. amm. $282880 scion
Sm amm. mom. gm. 35:8 25v 8&3 5:8 as, as e33
mmoqmaomomaoO 95.10
mm annex n mm annex a.
mﬁmNmumU wQOEda mimuo> mammsmU wcoa< Gowuduamﬂwwm

 

£82on wsoE< Horagom wsso> was $0.30QO 38m @580 do 2an

228

 

22. n at 82::
28: n a .5? n ax 2:5

33. H mm 02st
:23 u a seat n «X 235

 

 

as u 2 as u 2

2:. in: 2:. :35 3:52 22:25 See, cﬁ:2:<

m3. ....sa. new. mom. 3:82 00 e, oases:

com. 3.an 2m. ...:wma 3:82 00 e, sess<
33:3: 95.10

35. so. awe. So. 8:22:03 sass:

2o. 18.- go. as. aesséaao Esmaseo

S: was may, :2. 33:20 32:35.3:

mom. e8. am. 2%.. 28:28 5:232. :83

2:. ram: 2:. 1:2- 85:3 .550 :3, as :8an 2:32
33:63 98 mmoamsomomzoo @380

mm 2502 n m: 32532 :

 

QONED wnoe< wan—9w

mamnmﬁo magi/w nomuwbmmwmﬂ

 

283:0 mcoﬁ< Hot/anom waﬂo> was moBDOmom wommm 95.20 .3: oﬁeﬁ

229

 

88 H mm 0888
28.0 n 82.8 u ax 25$

38. H mm 088:

38.0 n a .88 n ax 22>

 

 

2m n 2. 8n n z
wmm. 02. wmm. amm. 33:03 2838an may? cwotmﬁxx
s8. 88. 98. 28.- 3:82 00 e, 0282an
8m. 8:. am. :28. 3:82 00 e, 8283.
35:3; Q3050
:6. moo. 5o. .32. 8:88:85 8888
So. 20.- 3o. :82. 8:85:85 889220
8m. 8m. 28. 8m. 8:85 88885
2m. 8m. 8m. 32. 28:8: 8:889 :83
28. 88. 5m. 8m..- 8::3 880 88 as 28::8 822
mmmnmsomomzoo 95.10
.m.m :88: : .m.m 8282 n

 

 

muoumﬁo muoE< mamas?

mcosmﬁo waoﬁ< nomumﬁmmwom

 

382m 8.85m waoE< Hogmnwm maso> was meSOmmm vommm 9820 .36 @3er

230

The next set of results explores the relative outcomes of discrimination dependant
on generational status and ethnicity. This hypothesis requires the addition of inter-
acted variables to the models speciﬁed above3. The results suggest that the odds of
voting among third generation Mexicans who perceive discrimination to be a prob-
lem are 1.62 times that of first generation immigrants who perceive discrimination
to be problematic (results not shown). An alternative interpretation asks about the
generational effect on participation among those who perceive discrimination to be
a problem and those who do not. The results show that the odds of voting among
third generation immigrants are 3.69 times higher if they perceive discrimination to
be a problem than if they do not.

To assess the relationship between voting, exposure, identity and language reten-
tion, predicted probabilities were generated for a range of values corresponding to
years lived in the United States while the other covariates remained ﬁxed at their

means. Symbolically, the predicted probability is given by
Pr = (Vote =1|x1,11=j,12 = k,I3 :1) = @(xoﬁ),j = 0,1;k = 0,1,1: 1,2

where 1:1 is years lived in the US, 11 is an indicator variable corresponding to re-
spondent’s ethnic origin, 12 indicates respondent’s primary mode of identification and
13 indicates Iingusitc capability. Table 6.12 shows the average of these probabilities
for individuals at various lengths of stay in the US. Disregarding exposure for the

moment, Table 6.12 reveals that English dominant Mexicans who identify primarily

 

3These results are included in the text only and are not presented in any table or
ﬁgure.

231

as an American register and vote at higher rates than Spanish dominant l\/Iexicans
who identify with their country of origin. The situation is similar for Cubans. Turn-
ing now to the interaction between exposure, identity and language, it is apparent
that exposure has a slightly greater effect on voting and registration for Cubans than
Mexicans. Initially, the probability of voting and registering among Cubans is lower
than among Mexicans but at some point their socialization into American culture out-
weighs that of Mexicans and the probability of voting and registering among Cubans
becomes greater than it is among Mexicans. Clearly, exposure mediates the effect of
language retention and identification as Mexican or Cuban for Mexicans and Cubans,
respectively, but the mediating effect is stronger for Cubans. This is shown by com-
paring the predicted probabilities for Mexicans and Cubans in the last row of Table
6.12. The only exception to this pattern occurs between English dominant Mexicans
and Cubans who self identify as Americans (.996 versus .976 in Table 6.12). The
effect of exposure among highly acculturated individuals matters less than it does
among the Spanish dominant who identify with their country of origin4. Overall,
Mexicans who speak English and identify themselves as Americans have the highest

overall predicted probability of voting among all the categories.

 

4There are competing explanations for these results. For example, it may be that
the model was not capable of providing a good estimate for individuals who have
lived in the US longer periods of time due to their small number.

232

 

:mmEO o0 x5550 H CO SENEEOD smmﬁmqm H Qm 33353 :motmE< H ~< SENEEOQ :wszm H mm
.83? :88 :65 um gauge“: 3o: 98? 855:3 Eovgaovﬁ $50
2: £33 3:238: osmmswﬁﬁ «Ed :osggémE op wﬂﬁaﬁom 8:3, 093 2: wﬁig .3 30> 8.

wEEEmE o:w m:$o> mo £358 :ommmogo: osmmwoﬁ 23 Bed @8062: 0.83 83234303 $23. @302

 

 

 

 

woo. owo. moo. o3. owo. moo. owo. ooo. on
nos. mm». o3. wmw. mom. 5mm. moo. So. on
32. mt. pom. own. oov. mom. new. wmo. om
mmo. 3o. m8. mmo. go. So. moo. m3. 3
:Smwwom wEuo> :Bmwwom w:30> :Bmmwmm w:$o> :Bﬂwmm m:so>
85m 25m 005m 2.5:
£5220 38:82 3.38me m: :m m3“;

 

>30 £8320 “Em 3:8me Sogﬁcem 263mg:

95 :ozwowsgg E :osgxosemm ~30...an 98 mooSOmom vommm 95:0 ”So 2an

233

 

at. 92.- I I I I I I 535682:

126365 x 00

 

 

ooH. ooo.- I I I I I I 55:58:85
95:0 x 00
m3. ooo. I I I I I I :osdﬁewuoma

@520 x 88:95:

 

I I I I I 3:. gm. 8:28:35

36363 x 8:3

 

&

mmv. mm: 2m. 2:. $4.- mow. :oSaEECQmE
assoc/65 x :50Em8<

.m.m o & .m.m o & .m.m o

 

:mnso Smog 8.85m :8:me

 

@0385 mb 5 w:Eo>
o:< 5:358:85 msgw mo Eoomm 830385 2o 2an

234

 

m5. vmmv nwﬁ. mom.. I #8323878me

3:329: x 00

 

 

RH. ccm.- SH. 30.. I I I cosaﬁﬁiuma
9650 x 00
m3. mwo. 2:. «me. I I I :oswEECOma

9:50 x smoEoE<

 

03. mam. I I I I I :oSmEEEoma

@320 x 0534

 

l I. HON. 3;. I I GOSGSMECUBQ

$53365 x smutw8<

A dd. a & dun a & .m.m c

 

mango 58$ 35an H8252

 

30> 3 wqcﬁmwmmwm
93. :oSmEEEomMQ @280 go @635 2050385 E.© 2nt

235

V BE 25H iaEEDm 8980 m3 ”8.50m
I». 8 8358 cocabamog :m wE8> km m a: :2 S macaw E I. 8v H a Se 3 A w n a 8 a mac 89m 0 0mm ”88
H . . . c . n H v. v . R R tﬁ a o m w m a a s m E Z

 

 

 

to. who. .80. mmo. mwo. mwo. mnmo. immo. moo. one. wmo. 3:3. 884 @885 3
83825 8 28m
3m. oovm. 33. E3. new. mom. *wnm. Em. “53mm. ***©m©. cam. “Inﬂow. 2856
888%
mg. mg. 33. 3:3. mam. immm. “LS. :32. ".3.“an *mmm. ..cLZ. :32. Eom swatch
888m
38. ion. 0:. :33. mvm. mam. we. ing. :03“. E3. 1me. ***©3. 25252
88.5%

mom 88> mom 88> mom 88> mom 88> mom 88> mwm 88>
Eom :mESL Eom m3 50m :wwﬁom Som— mD Eom :mmmpom Eom m3 832:3

58:0 SSE 88an ~83on

 

 

EmEO 8:282 cam 825$“on .wE8> ma :EEwomm 8 $8800 bﬁc Saab

236

To test segmented assimilation theory, I examined the effect of context of recep-
tion on voting and registration. Table 6.15, which is a crosstabulation of context of
reception and neighborhood variables by nativity status and national origin, provides
some insight into how socialization and neighborhood context impacts voting and reg-
istration. The table provides the results of a test that there is no statistical difference
in neighborhood context between voters/non-voters and registrants/non-registrants.
In other words, it is a test of whether the difference in these contextual variables is
different from zero. Mexicans who voted and registered are signiﬁcantly less likely to
live in areas characterized by high Latino concentrations. Similarly, US born Puerto
Rican and Cubans are who live in areas with high concentrations of Latinos are less
likely to vote. With few exceptions, Latinos who live in areas characterized by large
foreign born populations are less likely to vote and register. Foreign born Mexicans
are significantly more likely to vote and register if they live in areas with a large
proportion of citizens. Finally, US born Puerto Ricans and Mexicans are signiﬁcantly
less likely to vote if they live in areas characterized by high ratios of poverty to income
levels.

Do the univariate results change when other, possibly confounding variables are
added to a multivariate logistic regression? To ascertain the effect of these contextual
variables while holding other variables constant, I next re—ran the regressions includ—
ing the contextual variables and the interaction of these variables with generational
status, language and discrimination variables. Only two significant results emerged
from the analysis. The results remain substantially similar to those presented above.

For example, the cultural and structural assimilation measures and the context of

237

reception variables remained statistically signiﬁcant. Additionally, I found that Mex-
icans who live in highly concentrated Latino areas are signiﬁcantly less likely to vote
(b = —.977,se = .536,p = 0.06). For each standard deviation change in percent
Hispanic (ad. = .201) the odds of voting decrease by 17.7%. Each percentage point
decrease in percent Hispanic increases the odds of voting by 1.22, holding the ac-
culturation, group-based resources and sociodemographic variables constant at their
means. An additional ﬁnding was that of a signiﬁcant association between the per-
centage of citizens in the area and Cuban political behavior. Cubans who live in
areas characterized by high percentages of citizens are signiﬁcantly more likely to
vote (b = 1.49, 56 = 683,19 2 0.01). For each standard deviation change in percent
citizens (ad. = .119), the odds of voting increase by 19%.

According to segmented assimilation theory, immigrants living in low—SES com-
munities are worse off if they assimilate fully than if they do not. The estimated
coefficients from the various interaction terms are presented in Tables 6.16 and 6.17.
Table 6.16 presents the results using income to poverty rate in the neighborhood.
Among a total of 54 coefﬁcients, only two attained statistical signiﬁcance. Both,
however, are in an opposite direction than predicted by segmented assimilation the-
ory. For example, the results suggest that Mexicans living in a poor area are more
likely to vote and register as years lived in the United States increases. In Table 6.17,
I change the context measure to perceived discrimination against one’s ethnic group.
Again, of the 54 coefficients estimated, only 5 reach statistical signiﬁcance. Of these,
none were in the direction of segmented assimilation theory.

On the whole, it seems there is very little evidence in favor of segmented assimila-

238

tion theory. Since well over 100 tests were conducted, some of the coefﬁcients should
be signiﬁcant purely by chance. Furthermore, a large number were in a direction

Opposite to segmented assimilation.

239

+++++++

+++++++

+

+
+

**+

1+

 

 

vmgopmmwom «58>

#888“on Uo8>

888$me “88>

 

 

 

$8930

mﬂwomm Owhmﬂnm

£82me

 

22.855594. was 9.8800 8886635 cook/8m cosogopﬁ ”2.0 63.3.

vooﬁopnwmo Z 80m

owezmswq

E. A Eom ammopom 05

50m swatch AN.

mm. A momcﬁmoO 0A3
vooﬁongwmmz E $858.00 ea
mm. A 8836 05
eooﬁoﬁwsz s 83:6 ca

m A mwum 8 summed

Rheum .8 4%:me

Gomumﬁawmm< Uﬁd Gomuduﬂﬁduoaﬂ

240

* * *- - + - + + wwwzwzwq
- + - - - + Hm. A 38m cmwogom 05
+ + + - + + 50m ammohom RV
- *+ + + + + mm. A maﬁa—MAG “NU
- + + + + + 8003835382 5 8:38.00 05
- - - + - + mm. A msmNEO ob
- - - + + + wooiopgwﬁz E 28$qu ob
- - + + - 3*- m A Em Mo £93
+ - - - - - 8am 8 £984

 

wmuwuﬂmom 88> vmumumwmmm 88> vmhmpmmwom 88> :oSmEEmm/x cam coﬁwusﬁsooaq.

 

9:330 983m 355m 983me

 

2285820me wmmmmdsoHU 8 :oSQwoBm

Coswzgﬁwwaﬁ USS HXQQQOU UvWGQC‘E/U‘mmwa ﬁmvauwm ﬁOEUﬂHQpCH .20 wﬁﬁdrH‘

241

6.6 A typology of Latino voting behavior

In this section, non-voting is regarded as a possible outcome of the voting decision, and
it is assumed that individuals manifest an underlying propensity to vote or not based
on their disposition towards political objects at various different structural levels.
It is possible to objectively identify different “typologies” of voters according to this
underlying propensity and this allows an explanation for the causes and consequences
of non-voting. Accordingly, I provide a latent class analysis of Latino voting behavior
using data from the 2002 National Suvey of Latinos. The items selected to represent
voting behavior are strictly related to voting and not other forms of participation.

Figure 6.3 shows the elements of the latent class model (Lubke and Muthen, 2005).
The Yg’s are dichotomous observed variables that act as indicators for the latent trait,
7), such as the propensity to vote or not vote in a US election. Each 6 represents the
error variance unique to each Y variable. The factor is therefore composed of a pure
measure of the latent construct separated from the error variance (Shrayne et al.,
2006)

Not all variables must be endogenous to the latent class variable represented by
C in Figure 6.3. For example, the measures of acculturation, assimilation, group-
based resources and context of reception can be included in the model to predict the
distribution of C, the latent class. These variables are represented by X in Figure 6.3.
Because C is considered a discrete variable by nature, non-linear methods, such as
multinomial logistic regression, are required to estimate the effect of the covariates on

latent class membership in cases when more than 2 latent classes emerge to provide

242

Class Data/Voting/LCA modelpdf

 

 

 

 

 

U1

 

 

 

 

 

 

 

 

 

 

81 87 83

Figure 6.3. Latent Class Model of Latino Voting, Source: Shryane et al., 2006

243

the best fit to the data. When there are only 2 latent classes, multinomial logistic
regression reduces to binary logistic regression, similar to methods used to estimate
the effect of several variables on a discrete set of dependent variables.

One goal of this chapter is to test whether or not there is significant heterogene-
ity among Latinos with respect to voting in particular at different structural levels.
The hypothesis is that l\-=Iexicans, Puerto Ricans and Cubans have different latent
dispositions towards electoral politics and that these latent dispositions depend on
sociodemographic characteristics, level of acculturation, reception into US society and
other structural or attitudinal mechanisms traditionally believe to be barriers to par-
ticipation. The following three items from the 2004 Latino National Survey: Politics

and Civic Participation were chosen:

QN48 Have you ever voted in a United States Election? (Yes Voted = 1, No, did not

voted 2 2)

QN49 We often ﬁnd out that a lot of people aren’t able to vote because they were not
registered, or they were sick, or they just didn’t have time. Do you remember
for sure whether you voted in the November 2002 congressional election in your

district? (Yes Voted = 1, No, did not voted = 2)

QN5O Do you remember for sure whether you voted in the November 2000 presidential
election when George W. Bush ran against Al Gore and Ralph Nader? (Yes

Voted 2 1, No, did not voted 2 2)

The crosstabulation for each ethnic group is shown in Table 6.18. It is apparent

244

that the largest probability corresponds to Y1 = 1, Y2 2: 1, Y3 = 1 for each subgroup

of Mexicans (G = l)(.68), Puerto Ricans (G = 2)(.67) and Cubans (G = 3)(.87).

In latent class analysis the number of latent factors is specified by the data. As

shown in Table 6.19, the two class model provides the most acceptable fit and also

the most meaningful theoretical intereptation with respect to voting behavior. This

model can be interpreted as a highly participatory class of voters and a highly non-

participatory class. Therefore, the two latent class model offers a meaningful typology

for analyzing voting behavior.

Table 6.18: Crosstabulation of Voting Patterns

 

 

0 Y1 Y2 f P(Y1,Y2, Y3 = J} k, l)
l l 323 .68
l l 2 40 .08
l 2 1 16 .03
l 2 2 99 .21
2 l l 122 .67
2 1 2 13 .07
2 2 l 12 .07
2 2 2 34 .19
3 l l 187 .87
3 l 2 ll .05
3 2 1 4 .02
3 2 2 14 .06

 

Notes: ngroup; 1=Mexican; 2=Puerto Rican; 3=Cuban

 

Table 6.19: Type III Sums of Squares for Goodness of ﬁt

LL

AIC

BIC

x2

P

 

Independence

2-class with complete heterogeneity

2-class with complete homogeneity

Sample Size

—1773.910 3559.819 3588.465 396.47 0.00

-1598.194 3226.389 3298.002
-1599.956 3213.914 3247.333 3.785

245

875

0.00

0.00
.151

The results of fitting a two—class latent structure to the data are shown below.
Figure 6.4 presents the probabilities, 7rtX of falling into each of the two classes for
Mexican, Puerto Rican and Cuban immigrants, respectively. Notably, the respon-
dents are not equally distributed across classes. Roughly 85% of Cubans fall into
class 1, whereas only 67% of Mexicans and Puerto Ricans were assigned to Class

1 respectively. These estimates are very close to the marginal proportions reported

above (Table 6.18).

The cell entries of Table 6.20 refer to the conditional probability (fr—AX ) that a re-
spondent obtains a particular score on a manifest variable given that he or she belongs
to a particular class. For instance, a Mexican immigrant who belongs to latent class 1
voted in the 2000 presidential election with probability 1.0. Similarly, the conditional
probability of having voted in the 2002 congressional election given membership in
class 1 is .99. The table reveals that a similar latent class structure exists for Cuban
immigrants. The conditional probabilities of voting given membership in class 1 are
1.00 for both the 2000 and 2002 elections, respectively. We again observe a similar
latent class structure for Puerto Ricans. For example, the conditional probability
of voting in the 2000 election given membership in class 1 is .99 for Puerto Ricans.
Similarly, the conditional probability for 2002 is .99. From these probabilities, it is
evident that there is a highly participatory class to which 67% of Mexicans, 68%
of Puerto Ricans and 85% of Cuban immigrants belong. On the other hand, there
is a highly nonparticipatory class to which 33%, 33% and 15% of Mexicans, Puerto
Ricans and Cubans belong, respectively. In this class of non-voters, the conditional

probability of not voting in the 2000 presidential election given membership in class 2

246

is .75 for Puerto Ricans (cf. .73 for Mexicans and .56 for Cubans). What, then, is the

overall pattern that emerges from these data? There is a highly participatory class of

Mexican, Puerto Rican and Cuban immigrants. It therefore seems reasonable to label

this class of individuals “electorally responsive.”

Conversely, there is a smaller but

highly normarticipatory class which is deemed the “electorally nonresponsive” class

(Figure 6.4). The next section delves further to explore the characteristics that the

predict membership in each class.

Table 6.20: Estimated Parameters for the Unrestricted

Heterogeneous and Partially Restricted Homogeneous LCM

 

Group Latent

 

Model 5 Class t g? 7rv00 G X 7rv02 G X
111 1 .67 1.0 .99
2 .33 .27 .14
2 l .33 .99 .99
2 .68 .25 .23
3 1 .85 ls0 ll)
2 .15 .44 .22
H; 1 1 .70 .99a .96a
1 .30 .21 .16
2 l .69 .99“ .96“
2 .31 .21 .16
3 1 .90 .99“ .96a
2 .10 .21 .16

 

a The parameter value is subject to an across-group

homogeneity constraint.

«VOOGXz Voted 1n 2000; as Vt=OQGX Voted in 2002

Table 6.21 shows the two-class model that estimated regression coefficients for

Mexican, Puerto Rican and Cuban immigrants and Table 6.22 shows the signiﬁcance

levels for the reduction in log likelihood when each variable is added to the model.

247

0.7 :
0.6 ‘

05

 

04- ,

0.3: A Nbo'cm ‘
I

02? 1

0.1; - ,___--,___ . . ‘ . _, ‘ i

0 02 0A 0.6 0.8 1 1

i

Pow)

 

Figure 6.4. Gamma Parameter Estimates for Latent Class Membership (Voting)

248

Latent Class Men'bership Parameters
H10 Estimates for Electoral Participalion

‘P(X=2|Y=1)
1.4 '-
12 ~. :
Imi
1 ".
‘ OPuerlao I
1 0,3 ,, Rican
ACme 5
0.6 *‘ .
A
} 0.41 0
02 + _ 1 i a I
0 0.5 1 15
Pfx=1lY=1l

Figure 6.5. Rho Parameter Estimates of Electoral Participation

249

The effects of assimilation on class membership were quite s1_1l_)stantial. The multi—
plicative effect of assimilation on class membership reveals that each additional unit
increases the odds of being in class 1 (electorally responsive) relative to class 2 (elec-
torally nonresponsive) by 21%, 7% and 80% for Mexicans, Puerto Ricans and Cubans,
respectively. Linguistic assimilation tended to be related to class membership in di-
rections one would expect. As proﬁciency in English increases, the odds of being in
the class characterized as electorally responsive increases for both Puerto Ricans and
Cubans. The change in the predicted probability of being in class I relative to class II
over the range of linguistic proﬁciency is .087 for Puerto Ricans and .272 for Cubans.
The effect of group consciousness is largest among Cubans and opposite to that of
Mexicans and Puerto Ricans. Believing in a common linked fate increases the odd of
membership in class I by 1.68 for Mexicans and 1.36 for Puerto Ricans. On the other
hand, the pattern is reversed among Cubans. Cubans who believe in a common linked
fate among all Latinos are signiﬁcantly less likely to be in the electorally responsive
class (OR. = .568). Mexicans and Cubans who believe it is important for Latinos to
maintain aspects of their culture have higher odds of being in class I relative to class
II, a ﬁnding that is not inconsistent with the ﬁnding for linked fate. The latter refers
to unity among all Latinos whereas the former refers to unity among one’s ethnic
group. Volunteering in an ethnic based organization signiﬁcantly increases the odds
of being in a highly participatory class for both Mexicans (1.31) and Puerto Ricans
(1.35). Finally, increasing levels of education and age increase the odds of being in a

highly participatory class.

250

 

 

 

82 Sec 8.9 2 N NE
e
2: 98.0 88 38. 2: 88 @885
88 £8- 88 was- $8 $3- 8:858
mdudum OHSOdOowomoom
ﬁomuwmmammm< :whﬂuUH—uum
82 22 m3” m2 of $8 885888
83 amm.- mm: 38 2: 88.0 885.888 8388
as was $8 8.2. 58 SS- Seesaw:
83 2:. a: 8.0 3.8 82- 23:6 seem
88 3.0.0- a: 88 a: 88 338:5
a: $3 an; 88. 33 28- was 8%:
mega—Owed pemmmdzemv
a: was Q: 28 mi 8.0 m: sewer
mm; sad- as 58 82 ammo 28288
m: 83.0 2: 8.0 as 83 85288.2.
28 38. was $8- 2: ES 8883
GOmHM—mammmd‘ «vim Comuwhﬂu—H—UU<

e n e e. e n
:mQSO Smomm Ouuwﬂnm Gdoﬂﬂwaz Bﬂdiwx/

 

8883808880 Z 8882mm 889,. mmecmimsoamom 8882mm

AC 881$,on 8.30 95 89 885mm Hopoagem $882880 ammo 83.8%

 

251

85pr :< :63?» am? 3:02me 23 E > ,mmﬂo ESE 2: mm X

.2 n 2 u N2 u hm + 2 n x: n hm .3035 8552 2: m2 2 was H$82

 

$3 a u N: n bu
$3 2 u x: u hm
88 88>

$3 2 n x: n bu
mad 2 n x: M \cm
88 88>

q

553 $3 32 a N NE

252

 

Table 6.22: Beta Parameter Test Type III Sums of Squares Based on 2 x LL

 

 

Covariate Exclusion LL Change in 2LL (if p

Panethnic -620.20 6.47 3 0.09
Assimilate 659.16 12.74 3 0.01
Retain Culture -682.32 6.76 3 0.07
Melting Pot -672.98 3.61 3 0.31
Generation 1 -683.98 14.00 3 0.00
Generation 2 -685.91 .787 2 0.68
Language -621.80 9.67 3 0.02
\Nork together politically -690.55 3.13 3 0.37
Volunteer -680.52 7.64 3 0.05
Home Country Politics —675.86 2.56 3 0.46
Education -622.56 11.21 3 0.01
Income -617.95 1.98 3 0.57
Individual Discrimination —679.96 1.32 3 0.72
Group-Based Discrimination -675.52 0.65 3 0.88
Political Efﬁcacy ~677.27 9.99 3 0.01
Political Apathy -683.98 6.45 3 0.09
Age -637.28 40.64 3 0.00
Married -684.32 12.91 3 0.00
Female -683.98 5. 28 3 0.15

 

6.7 A negative binomial model of Latino citizen-
ship acquisition

Among ﬁrst generation immigrants, citizenship acquisition is an important indicator
of political assimilation. Researchers exploring patterns of Latino naturalization tend
to model it as a dichotomous response using logistic regression analysis. Consequently,
we know the factors that predict citizenship versus noncitizenship but we do not know

the factors responsible for decreasing the rate at which immigrants naturalize. This

253

knowledge would greatly help us understand how quickly immigrants are assimilating
to the polity once they are eligible to do 805. The rate at which immigrants choose
to become citizens once eligible has further implications for other types of political
activity since citizenship is a legal prerequisite for voting (although in practice, the
lack of citizenship may not preclude individuals from registering to vote and even
voting). In any event, the most often cited reason given by Latinos for attaining
citizenship status is that it provides them with the ability to vote.

Among Latinos, there is substantial inter-group variation that characterizes
whether or not to become a United States citizen or to participate in politics. Natu-
ralization is less of an issue among Puerto Ricans because they are granted citizenship
status due to their commonwealth origins (McClain and Garcia 1993); it is, however,
an important indicator of attachment to the US polity for foreign born Mexicans and
Cubans. Overall, there is broad agreement in the literature that naturalized Hispan-
ics vote at lower rates than the native-born. The fact that the majority of Mexicans
and Cubans living in this country were not born in the United States highlights what
has been considered to be a major barrier to political participation: the large pro-
portion of Latinos that are ineligible to vote. In an attempt to explain naturalization
rates among the foreign born, several studies have found that a process of reinforce-
ment through exposure to the political system underlies the development of political
attitudes among immigrant groups (Wong, 2000). Highton and Burris (2002), for

example, studied the interaction between nativity status and exposure and report

 

5There are laws dictating how long an individual must wait before he or she is
eligible to become a citizen. On average, these laws mandate a 5 year waiting period.

254

that nativity status has a powerful effect on turnout, but only when considered in
conjunction with the number of years lived in the United States.

Many researchers have noted that Hispanic immigrants naturalize at a very slow
pace. In Jones—Correa’s (1998) well-cited study of naturalization rates among 112
Latinos from New York city, he found that naturalization is often delayed due to
the conflict that immigrants feel from on the one hand desiring to join the polit-
ical community and on the other feeling as if they are betraying their homeland.
Cubans are unique in that strong anti-Castro sentiment coupled with a unique po-
litical status have produced high naturalization rates (McClain 1993). In addition,
discrimination may be causally related to one’s decision not to acquire citizenship
status. For example, in my research on Cuban refugees, I found that in addition
to income, age and satisfaction with the United States, perceived discrimination is
signiﬁcant to a Cuban’s decision to become a United States citizen. Additionally,
lDeSipio (2001) found that individual differences such as education and income are
more significant than nationality differences in explaining immigrant naturalization,
even though national origins continue to have predictive value. Another factor that
should be considered is that individuals in the second generation have long had a
pattern of distancing themselves from immigrants rather than linking them to the
polity (de la Garza, 2004; Browning and de la Garza 1986, Mollenkopf et al., 2001).
These researchers believe that the extent that different nationalities adhere to this
pattern, they could reduce the pace at which conationals naturalize.

This section explores the factors related to the waiting time to citizenship among

first generation immigrants. Two continuous variables derived from the 2002 National

255

Latino Survey were used to model the dependent variable. First generation immi-
grants were ﬁrst asked the year they came to the United States. They were further
asked to provide the year in which they became a citizen. Subtracting these two
values yielded a positive count variable that indicated length of time it took them to

acquire citizenship status.

6.7. 1 Results

The average waiting time to citizenship over the sample was 13.3 with a variance of
approximately 64, which indicates signiﬁcant overdispersion of the data. The form of
the overdispersion is shown in Figures 6.6 and 6.7. The diamonds show the predicted
probabilities from a Poisson distribution with )u = 13.3. Compared to the Poisson
distribution, the observed distribution has substantially fewer counts less than ﬁve,
which makes sense given the legal restrictions, and more observations greater than
ﬁve. Overall, the sample variance is larger than would be expected if naturalization
was governed by a Poisson process in which both Mexicans and Cubans naturalized
at the same rate. This leads quite naturally to other estimation techniques, such as
the negative binomial response model, which accounts for such heterogeneity in the
data. Both ﬁgures indicate that this model represents a signiﬁcant improvement over
the Poisson distribution.

The simplest way to interpret the results of the negative binomial response model
is by using the factor changes in the expected count (See Appendix C for variable

codings and regression output). The separate regressions for Mexicans and Cubans

256

-o—observed ' + Poisson lEBirrirurid

". .
o

 

 

. 012345678910
. NurberofComts l

Figure 6.6. Negative Binomial and Poisson Predicted Probabilities vs. Observered
Proportion for Model Fit

257

” ”"H I'_B'_'IW_PI'I IP lul—‘IitiesB lg" ’
Esﬁnaedlvhms

—o——C1:sen.ed -+—Poisson +NegBimm‘d

 

 

 

 

 

 

 

Figure 6.7. Negative Binomial Predicted Probabilities Based on Estimated Means

258

are given in Tables 6.23 and 6.25, respectively. For Mexicans, the expected waiting
time decreases when individuals identify with Mexico versus as a Latino/Hispanic
(,6 = —.19,p = .07). Mexicans who do not believe that group-based discrimination

is a problem naturalize at faster rates than Mexicans who do believe that group-

based discrimination is a problem (,6 = —.06, p = .01). l\-‘Iexicans who believe that
life is better in Mexico (6 = —.085,p = .07) and who have weak ties to Mexico
([3 = —.06, p = .01) naturalize at faster rates than Mexicans who believe that life is

better in the US6 and who have strong ties to Mexico. Finally, the results show that
educational attainment (6 = -.187, p = .00) decreases waiting time while years lived

in the United States (6 = .041,p = .07) tends to increase it.

 

(This result is counterintuitive. See Appendix C for ﬁnal regression model.

259

end-

NON-

Swd-
mm.m-

mm;

mwdﬁ
$.m

ed

va-

SA-

22d-
mvN-

mwd

omA
wma

Nd

no.3-

ow:-

III»

owN-

Bd-

had

wmdm
3.2

838% 4

Swam-

howd-
mm:-

$5-
www-

www.m-
dmd-

SA-

www.c-

$4-

0.36-
ﬂam-

mwd

mm:
mm:

ooh.-
www.m-

Bmﬁm 323-83.

0333-? 3.5:00

wwwd Sud 02me 3 xodn 968 3 ESE

ovwd mmwd 02er 8 meat
:mwmuo mo FEE—CO 3 $5.

mowd Rad aossﬁﬁtommw Enememm

Rad vmmd mmmcmsowomqoo Eossom

mead Ewd comueocscog 282382
mmmamsomomcoU 9.qu

vvmd wood 2:83

wwwd mvwd cos-832m
nomad—mammmd- Eusuozﬁm

mo; 2: m2 E Sﬁmm £5

Raomuamomm Euogwsom \3263334

m3 3 Egg
Baum owwswcwq

2.2 8.2
8.2 o:
somew~m8mmm< Rena—:0
owes ES 8652 as... s 8%: Basses
$2 $3 assesses e885.
eczemasma. ﬁnesseSEeE

3% 35 33.223

 

83:50 Segue-«Q

3:35 .28er

353356 on. was magma mo :oﬂmwoewom #380:me wig-ammz ”mmb @558

260

3:0 £23,282: :32me 8302

 

mm.m

mhmd

5.:

msmd

RA

No;

$3

261

 

3:0 $08352: 202802 .2 H 3. ”0092

 

 

 

 

as see 225 sass: a

one 2.: o

:3 8.2 As:- meesmv a. 8F 858

was as ceased s SE a

was 8s 2858a 8:06 e 858598

as $2 Amazed 8.286 m seem-e85-

2.0 5.2 28225833 messages"; so.

as 2.2 580 so 8550 252 based

was 3.2 sssm ease:

mmd mmhﬁ 03082 9220803 02H-
3 H «s02 A 3K 3. :85 08060.5 Spats?

 

0005 :02: ad moﬁmtg .2050 9:30: 005. 03sz 020 00505820me

2030.53.03 0.82025 “0 mos; cmxc smﬁw 9:22.336 *0 523030.22» 08060an gammy 053%.

 

262

Turning next to the output for Cuban respondents, we observe a different pattern.
For example, income (,{3 = —.023,p = .04), identiﬁcation with the US versus Cuba as
one’s true homeland ([3 = 35,1) 2 .02) and years lived in the US (5 = .030,p = .00)
are the only variables that signiﬁcantly affect citizenship acquisition. More speciﬁ-
cally, Cubans with high income levels have shorter waiting times than Cubans with
low levels of income. Identification with Cuba as one’s true homeland and years lived
in the US increase waiting time. The positive coefﬁcient on the exposure variable for
both Mexicans and Cubans suggests that there may be a window of opportunity and
once that opportunity is missed it is gone forever. Tables 6.25 and 6.26 are alterna-
tive ways to analyze and think about the data. The tables Show the probability that
waiting time will be greater than expected for the signiﬁcant variables ﬁxed at various
levels while the other variables in the model are held constant at their means for each
regression separately. For example, Mexicans who identify with Mexico are predicted
to wait 15.33 years before acquiring citizenship versus 10.76 years for Mexicans who
identify the US as their true homeland. This suggests that with respect to political
assimilation, ties to one’s home country matters less for Mexicans than it does for
Cubans. The predicted probability that waiting time will exceed its expected value
among individuals who identify with Mexico is .49 compared to .53 among Mexicans
who identify as Latino/ Hispanic. The probability of exceeding the expected waiting
time for individuals with strong ties to Mexico is more than twice that of individuals
with only weak ties. Among Cubans, on the other hand, individuals who identify
Cuba as their true homeland are expected to wait approximately 2 years longer than

Cubans who identify the US as their true homeland (7.54 - 5.29). The predicted prob—

263

ability that waiting time will exceed its expected value is twice as large among Cubans
who identify as Cuba versus as Latino/ Hispanic. Finally, lower income Cubans have

signiﬁcantly longer waiting times than Cubans with high incomes (Table 6.26).

264

 

l l. l mum. Cm. mo; 004 0000 00 0000. 0>0E 00 0005 0000a
00in 00 .300000 00 00TH

amm.- NE: SN- l m3; mg. 50. 005 #0000090

I l l 000. com. 00A 004 0000000000000 Bossom

l l l $0. go. 004 SA 0030005003 00000
000000300000 00000

no.7 mm? 3.0- l mmﬁ- amm. mwm. 0083

Sam- mﬁm- wmé- l 20- man. omw. 000.0000m
0050:00mmm< 100000003

30.. 0mm.. 00m- l 0mm..- mg. $0. mD 0_ 00000m
E0050000ﬂ _000m>0£0m\_00m00030<

0:- 00m..- 03- l 000.- 80. 000. 0000003

3.: SA 3.3 l 2: m0; moa mD E 200%
00m00~m8mmm< 70.50100

l l l no.0 no.0 2A mm; 00.200000 00:. 0.0 0000 0050002

| l i 00.0. 00.7 mg. mum. 003023003 000C0E<
0030:5604 10003005000":
0 4 N 0 00:00 4 N T 0 00.00 30 000 000.200

00000.0 000.8qu 000.0803 000009 380%

 

000.0000 000.8qu

00000.0 083$

05000036 0.0 0000 w0§0>> 00 008000w0m #000005 0>E0w0z ”mmb 03.0%

 

265

E00 00000w0000: 00000 ”0002

 

mm;

$3-

93.

mwmr

3w

mag.

02.

.0ri

cm;

Em.

SA

03.

0w<
0030000.? 300000

3:0 8 0F

266

 

E00 0000000000: 00000 .mH H i ”0.002

 

 

 

3.0 00.0 @5800 0000 00

mod $40. 300005 00020006 3 00503000005
wad mm.» 300005 305 H 000005
00.0 80 20805803 80800002 .6
3.0 990 00020 m0 @00000 0002 waE&
nod and 000.0% 00000D

mmd was 0000 00000000E 0009

A: H «.00: A 3%

1 00000 00000000m

0E0E0>

 

:5me :QSQ paw wriﬂdmhds “0:00 WQQJOS OEOUGH Uﬁd

00300030003 000E0E< .00 0.00—00> 00000 00.0% 90000055 00 5:50.000m 00560.5 ”away 0308

 

267

6.8 Summary and Conclusions

This chapter explored three competing models of assimilation theory with respect to
Latino electoral behavior. The various measures of acculturation and assimilation did
a relatively better job than either the ethnic disadvantage or segmented assimilation
models at predicting electoral participation across all Latino subgroups. More speciﬁ-
cally, generational status and years lived in the United States increased the likelihood
of registering to vote and voting among Mexicans, Cubans and Puerto Ricans. High—
ton and Burris (2002) suggested that socialization and exposure to American culture
explain why turnout is higher among Latinos who have resided in the US for greater
periods of time. According to these authors, citizens who have lived in the country
for longer periods of time tend to be more linguistically assimilated, which may assist
the acquisition of political information and further increase turnout. In fact, English
dominance was highly related to political participation across all subgroups further
supporting the predictions of straightline assimilation theory. In addition, those who
have resided in the US longer tend to be more structurally assimilated. Structural
assimilation was also significantly related to electoral participation but the precise
relationships differed by subgroup. Interestingly, identificational assimilation, that is
prefering an American identity to a national origin or panethnic identity, was consis-
tently and strongly related to voting and registering to vote, with its strongest effect
being exerted on Cuban participation. I suggest that in addition to socializing and
exposure factors, the meaning given to “being American” explains, in part, why elec-

toral participation is higher among Cubans who identify as American. Researchers

268

have suggested that American national identity is meaningful because it leads to po-
litical outcomes (Citrin, 2003) and the present findings tend to support this assertion.
Latinos do not merely make empty claims that they are American, claiming mem-
bership in the American community has political consequences as well. The variables
measuring attitudinal and behavioral assimilation were not in the direction predicted
by straightline theory, and therefore do not support the model in that respect. For
example, the variable pertaining to the relative conditions in the US being better than
they are than one’s country of origin did not consistently predict electoral participa-
tion among Mexicans, Puerto Ricans or Cubans. In fact, a counterintuitive ﬁnding
was that Mexicans who believe life is better in the US had longer waiting times to
citizenship acquisition. Therefore, the “better life” hypothesis was not supported by
these data.

An important finding from this chapter is that not only do increasing ties to
the US increase political participation, but decreasing ties to one’s country of origin
increase it as well. In addition, I showed that increasing ties to the US simultaneously
decrease political participation in one’s country of origin. Therefore, this provides
evidence that ﬁrst generation immigrants display high levels of committment to the
American polity and as they assimilate both culturally and structurally, political
ties to their homeland decrease. Because decreasing ties to one’s country of origin
implies increasing ties to the US, this variable is taken as an important measure of
assimilation.

Turning to the effect of group-based resources, I found very little support for the

hypothesis for the ethnic disadvantage or segmented models. Overall, the results sug-

269

gest that the effect of group consciousness, group identity, sense of common culture
and experiences with discrimination on registration and voting all appear to differ by
ethnic origin and vary in importance both within and between national origin groups.
For example, individuals opting for either a national origin or panethnic identity con-
sistently participated at lower levels in comparison to those who prefered an American
identity. Previous research has viewed attachment to one’s culture and identification
with that culture as an important group based resource, but here I found that these
variables had a negative effect or no effect at all. The analysis further suggests that
identiﬁcation as American, in addition to being a measure of assimilation, should also
be considered a group based resource in future studies. Not only is claiming member-
ship in the American community directly related to political participation, I have also
shown that American identiﬁcation has a protective effect in the face of perceptions of
discrimination. This was demonstrated in the interaction effects between discrimina-
tion and identiﬁcation as an American which suggested that individuals who prefer an
American identity and who perceive discrimination to be problematic have a higher
probability of voting and registering to vote than those who do not perceive dis-
crimination to be problematic (Tables 6.13 and 6.14). Importantly, the consistent
ﬁnding of the impact of discrimination on political participation supports the ethnic
disadvantage model. Nevertheless, in contrast to previous research, I showed that
it is not the adoption of an ethnic identity that mitigates the effect of discrimina-
tion, it is actually the preference for an American identity. This result held across
several regressions and among all subgroups. Among Cubans, group-based resources

tend to suppress political activity whereas the traditional indicators of assimilation

270

consistently increased political behavior. This is a very important ﬁnding that has
not been discussed in the literature, despite the acknowledgement that Cubans are
distinct from other Latino groups. The perception of a common linked fate, as well as
the preference for a panethnic identity, both signiﬁcantly decreased Cuban political
participation. As discussed in the previous chapter, Cubans have a strong prefer-
ence for participating into American culture almost to the exclusion of their own.
Individuals clearly take pride in being American and identifying with other Latinos
has a strong negative effect on political assimilation, at least with respect to voting
and registration. Finally, there is some evidence that participation in the evangelical
church motivates Mexican political behavior.

The latent class analysis supported the results of the logistic regression, and ex-
tended those results to characterize Latino voting behavior into clusters of highly
responsive and highly nonresponsive individuals. In addition, the latent class analy-
sis revealed signiﬁcant homogeneity in voting behavior across subgroups. However,
the negative binomial response model revealed signiﬁcant differences in the rate at
which Latino subgroups naturalize, with Cubans tending to acquire citizenship at
faster rates than Mexicans. This suggests that the participation proﬁle of Mexicans,
Cubans and Puerto Ricans may be similar but that differences in naturalization rates
create the appearence that Mexicans are not as politically active as Cubans.

While the unadjusted estimates show that context of reception and neighbor-
hood quality effect. participation, the importance of those variables dissappeared when
controlling for other covariates pertaining to assimilation and acculturation. Never—

theless, there is some evidence that as the percent of Hispanics increases, Mexican

271

participation decreases. In addition, I found that Cubans living in areas characterized
by high percentages of citizens tend to vote at higher levels. A more direct test of

segmented assimilation theory failed to reveal any signiﬁcant results.

272

Chapter 7

Group Solidarity, Pan-Ethnic
Identity and Acculturation: A
Resource Model of Latino

Nonelectoral Political Participation

7. 1 Introduction

Due to the large proportion of noncitizens, nonelectoral political participation assumes
a peculiar salience among Latinos. Engagement in nonelectoral types of activities
represents one of the few ways in which noncitizens can effect their personal policy
preferences (de la Garza, 2004). It also functions “to stimulate the development of

local civic institutions that produces social capital, which strengthens neighborhoods

273

and empowers ethnic groups (de la Garza, 2004, p. 95).”

Nonelectoral behavior among Latinos is highly correlated with electoral involve-
ment. Diaz (1996), for example, found that Puerto Ricans and Mexicans involved
with political organizations vote at higher rates than those who are not involved in
political organizations. Nonelectoral participation has been further found to be a
function of social capital rather than of ethnicity. Nevertheless, nonelectoral involve-
ment among Latinos is lower than it is among other ethnic groups, such as African
Americans and Anglos. Signiﬁcant differences exist, too, between Latino subgroups.
Cubans have the lowest rates of nonelectoral participation and, despite having much
higher incomes, are the least likely to make contributions to political or social causes.
Moreover, Mexican Americans and Puerto Ricans are more likely than Anglos to at-
tend political rallies or wear campaign buttons but generally speaking they are more
active than Cubans.

With limited exception, scholars have not found any signiﬁcant differences in
nonelectoral participation between foreign— born citizens and natives: both groups
have been shown to have very low levels of civic engagement. One exception is a
qualitative study undertaken by Segura et al. (1999) of four heavily populated Latino
communities in California and New Mexico. They found that foreign-born Latinos
had lower levels of civic and political engagement than the native-born Latinos, but
the results are not generalizeable (Segura et al., 1999). More recent studies have
found that, irrespective of national origin, both foreign born and native-born Latinos
have similarly low rates of participation in activities such as political demonstrations

or contributing money to an election campaign (de la Garza, 2004).

274

7.2 Data and Method

The data for this study are derived from the 2004 National Survey of Latinos: Pol-
itics and Civic Participation. To explore the effects of group identification, group
consciousness, context of reception and acculturation on nonelectoral engagement
among Latinos, the following indicators of nonelectoral political engagement were

utilized as dependent variables in several regression models:

0 Have you attended a public meeting or demonstration in the community in

which you live;

0 Have you contacted an elected ofﬁcial;

0 Have you contributed money to a candidate running for public ofﬁce;

Have you attended a political party meeting or function; or

Have you worked as a volunteer or for pay for a political candidate.

Possible responses included “1 = Yes” and “2 = No.” An ordered set of responses
was created to quantify the level of each respondent’s nonelectoral political partici-
pation. Responses were recoded to indicate a range of possible participatory activity
from no participation to frequent participation. The ﬁve participation variables were
added together to create an index that ranged from 5 to 10, such that individuals
with a score of 5 participated in all ﬁve activities and individuals with a score of 10
did not participate in any activity. A score of 10 was recoded as “1 = None of the

H

time.” A score of 9 was recoded as “Rarely. A score of 8 was recoded as “Often.”

275

Scores 5, 6, and 7 were recoded as “Very Often.” The distribution of responses is
shown in Table 7.1.

Table 7.1. Distribution of Responses of Nonelectoral
participation among a sample of N = 2154 Latinos

Source: 2004 National Political Survey of Latinos

 

Frequency Percent of Sample
None of the time 70%

Rarely 16

Often

Very Often

 

Due to the nature of the dependent variable, ordinal logistic regression techniques
were used to estimate the model. This has advantages over other techniques because
treating these measures as if they were an interval level involves the implicit assump-
tion that the intervals between adjacent categories are equal (Long, 1997). Finally,
estimating simultaneous equations using proportional odds models takes advantage
of what seems to be the natural ordering of the outcome variable. A test of the

proportional odds assumption, while stringent, was met in all models reported.

276

00000 30505000 00 00000000302 00 000000030>HUO\Z ”003003 00 “0030000 .3000 #005200 000003<Hm\2m
0030000000000 00 w030000 0:000 0 0000000400 HOD)?» ”000000000 0 .000 0000039 0 00 000003 HD>

00006000 0 00 00002 00000000000HEU £00500 00000000>0w 0 000000000HOUM0000>H> ”000000mw0mum W000§OHO
00020-000HOZ M0000. 0m000omnmmw0000m 00000mﬂmmM00000HOM000§02H2 “00002

 

 

 

 

 

0t. 2 000.0 2 000 8 a... S 000 R 003 00. 000.0 mm 002
0: m $0.0 0 000 2 0% z 000 s 02.0 2 0.0.0 0 050
at 2 $0.0 E 0.00 00 80 S 000 00 02.0 00 000.0 2 0:00
0: m 30.0 m 00.0 0 a0. 0. 000 0 003 0 02.0 ... 0>
0t a 000.0 0 000 2 00. 00 000 0 02.0 0 000.0 0 20
0t. 0 000.0 0 000 2 09. 2 000 2 02.0 2 0.0.0 2 00
0.000 0 0.00 E 000 m... E 00 000 00 000 E 000.0 E >
0.02 0 0.0. mm 000 00 000 00 03 00 .30 t 03.0 E 0
0.02 0 000.0 0.. 000 0.: $0 00 000 000 :00 mm 20.0 a. 0
0 .0 0 .0 z .0 0 .0 2 as 0 .0 ... .0
0200 0.0 .00 0 00 2 :0

 

000304 w0000< 000E0w0m0m 000000000002 00 0030303 0>$E0000Q .m.» 0308

277

7 .3 Results

Compared to registering and voting, fewer Latinos participate in activities such as
contacting government officials (12%), contributing money to a candidate (9%), work-
ing as a volunteer or for pay for a candidate (5%), attending a public meeting (19%)
or a political function (9%) (Table 7.2). In terms of nativity status, 31% of the US.
born claimed to have volunteered their time to a neighborhood, business or com-
munity group versus only 15% of the immigrant sample. Across every measure of
political participation beyond voting, the US. born report higher levels of involve-
ment than do immigrants. For example, with respect to contacting a government
ofﬁcial, 19% of US. born Latinos answered afﬁrmatively compared to only 8% of
foreign-born Latinos and 4% of foreign born non-citizens. Curiously, and in line with
previous research on minority political participation (Wong, et al., 2005) immigrant
status appears to matter more for activity beyond voting compared to voting, at least
at the bivariate level. Wong et al. (2005) provided, as an explanation, the obstacle

of citizenship as opposed to immigrant status itself (Wong et. a1, 2005).

278

 

Table 7.3 Ordinal Logistic Regression Models of Nonelectoral Political

Participation Among Mexicans, Puerto Ricans and Cubans

 

 

 

Mexican Puerto Rican Cubans

Variable b p—value b p—value b p—value
Acculturation and Cultural Assimilation

Years in US .014 .406 .315 .608 .027 046**

Gender Assimilation -.116 .258 .047 .692 .120 .134

Retain Culture -.235 .392 -.335 .263 .476 .202

Language .269 .000*** .074 .406 .202 056*
Attitudinal/ Behavioral Receptional

Better in US .090 .191 — — -.255 019**

Personal Disc .672 009*“ .618 .039** .709 094*

Discrimination Problem -065 .806 .114 .696 -.549 .138
Structural Assimilation

Education .109 087* .248 .000*** -007 .920

Income -.001 .808 .001 .736 .005 .300

US Citizen -.191 .574 —— — .540 .231
Group-Based Resources

Panethnic Identity .104 .665 .290 .310 .233 .451

Political Group Consciousness .218 .373 .428 .141 -.398 .225
Controls

Age .005 .694 .006 .514 .010 .386

Female —.495 .05** -.422 .149 -.113 .732

Married .122 .622 -.246 .384 -.267 .401
Model Log Likelihood ~307.92545 -229.86469 -208.9628
n 436 230 244
Prob > ch22 .0000 .0267 .0000

 

Notes: ***p S .01, ** p S .05, * p g .10

Turing now to an examination of several hypotheses regarding the impact of group

identity, group consciousness, context of reception and acculturation on nonelectoral

political participation, Table 7.3 presents the results of the ordinal logistic regression

279

for Mexicans, Puerto Ricans and Cubans separately. The results suggest that several
measures pertaining to context of reception, acculturation and assimilation matter
for political engagement but their effect differs slightly across subgroups. Among
Cubans, years lived in the US and proﬁciency with the English language are pos-
itively and signiﬁcantly associated with nonelectoral political activity. In addition,
context of reception, as measured by respondent’s belief that the relative conditions in
the US versus one’s country of origin are better, is signiﬁcantly and positively related
to Cuban political behavior. That is, Cubans who perceive relative conditions in the
US to be better than they are in Cuba are signiﬁcantly more likely to engage in non-
electoral types of political activities. Among Cubans, too, the odds of participating
increase signiﬁcantly among individuals with personal experiences with discrimination
(p = .094). Experience with personal discrimination also seems to motivate political
activity among Puerto Ricans and Mexican Americans. Moreover, Puerto Ricans and
Mexicans who are structurally assimilated are more likely to participate in politics.
For example, the odds of participating increase by 28% and 11% for each unit increase
in educational attainment among Puerto Ricans and Mexicans, respectively. Accul-
turation, as measured by linguistic proﬁciency, is a signiﬁcant predictor of Mexican
American involvement in nonelectoral types of political activity. In addition, among
Mexicans, females are signiﬁcantly less likely to engage in nonelectoral participation
than are males. Importantly, one’s preference for a panethnic identity and status as a
US citizen did not signiﬁcantly predict participation in any of the models. Including
the discrimination variables decreased the p—value of the variable measuring years

lived in the United States, which became statistically insigniﬁcant. This implies a

280

positive differential between Mexican American individuals with similar perceptions
of and experiences with discrimination and increasing exposure to American cultural
norms.

The presence of multicollinarity created less precise estimators, therefore it became
necessary to test for the joint effect of the various measures of discriminatory context.
Table 7.4 examines the joint effect of the various contextual measures. As can be seen
from the table, there are several statistically signiﬁcant results. Among Mexicans, for
example, the joint effect of having been discriminated against, feeling unwelcome in
the US and perceiving conditions to be better in one’s country of origin signiﬁcantly
and positively impact nonelectoral engagement. A similar analysis was performed
with respect to Cubans. In contrast to Mexicans, the joint effect of discrimination
and perceiving conditions to be better in Cuba had a negative impact on Cuban

nonelectoral participation.

281

 

mHh. H 05000030000300 ”000070

 

 

1.0.8. 00.0.0 000. 0.0.0 0000. 00.0 0..
00.0. 0.0.... I I ......0000. 00.0 0
......0500. 00.00 000. 0.0.0 000.. 0.0.0 0
008. 000.0. 0.00. 00.0 000. 00.0 0
000. 00.0.. I I .300. 00.... 0
......08. 00.00 000. 000. 00.0. 0.0.0 0
......08. 04.00 I I .08. 8.0 0.
.0 .00 .0 .00 .0 \s .00.
0003000 00000m Opumﬁnm 000000032

0000.005 I|.. 000000.006 H 00.000.000m0
$00003 H 0.500003%. H owwﬁwgwah
$00.50 N 0083000

.0000on .II. owwwwamab.

0.500003%. H 00.000.000Q

0200.0000Q H 000.500.006.

$00006 H 08.000006. H 0200.0000Q H aoémwnﬂ

00000000003

 

00000000000000 00000050 00 0000000000: 00000000000 00 .0me 000005 000.0. 030%

282

7 .3.1 Sustaining Participation Beyond A Single Act

A question that has previously not been addressed in the literature pertains to the
factors that sustain political engagement beyond a single act. To address this ques-
tion, I used continuation logistic regression models of nonelectoral engagement. This
approach takes advantage of the sequential nature of decision-making. In what follows

below, nonelectoral participation was recoded into the following speciﬁcation:
0 y,- = 1 if an individual participated in less than 1 participatory act (i.e. none)

0 y,- = 2 if an individual participated in at least 1 participatory act (i.e. one or

more)

0 y,- = 3 if an individual participated in more than 1 participatory act (i.e. more

than one)

Although the variable can be considered ordinal, proportional odds models are
restrictive in not allowing separate structural mechanisms across different categories.
If different mechanisms are at work for different levels of participatory transitions
(i.e. going from none to one) the process can be broken down into a sequence of
transitions. In this case, conditional probabilities model the probability that y,- 2 j
for j > 1, given that y,- = j — 1, denoted as PS. The process of participating in
nonelectoral politics can be viewed, then, as a series of binary choices.

The models presented below were estimated by conditioning on the appropriate
subsamples of the data. The entire sample was used to model the probability of 1 or

more participatory act (R15) using a continuation ratio logit model. In the model for

283

the ﬁrst transition, the response variable is coded as 1 if the respondent engaged in
at least one participatory act, 0 otherwise. Note that P“ = 1 — P5. The conditional
probability of having participated in more than one act Pig was modeled next using
the continuation logit model using only the subset of respondents with at least one
participatory act. In a model for the second transition, the response variable is coded
1 if the respondent has more than one participatory act and 0 otherwise such that

P22 = P350 -PE)

1 Z

and

Pi3=P$Piii

which shows that the sequence of binary logit models completely describe the

multinomial process.

7 .3.2 A Continuation Ratio Logistic Model of Nonelectoral
Participation Among Latinos

Among Mexicans, the results suggest that the odds of completing at least one par-
ticipatory act are signiﬁcantly associated with ethnocultural Americanism, panethnic
identity, English proﬁciency, personal discrimination, education, age and gender (Ta-

ble 7.5).

284

 

10.0.00. ww.m wmo. .1.me 30 N00.
......mmo. 2N mwm. Co. m0. moo.
.53. 0mg mg. New. mm. mmm.

m2. wmAI mgr to. $0. 000.-
amm. on. NS. ***woo. mwN mmv.
$0. mmA- 0.00.. mg. 00.0 moo.

.03..va SN wmﬁ 0.3.000. mud owﬁ

30. 0.0. ms. .030. 30 0am.

3&3. $0 vwm. one. 80. 000.-
......omo. 00.0 mvm. mwm. mod- 03..
0.35. mwa 03. 0.02.000. wad m3.

:0. $0 3m. 30. 000 Km.
mmo. 00. one. own. two- wmmr

00:9): QHSNKVIN a ®S_.®>IQ ®0JE>IN a

0mm:Am 0mm

0w<
0030000.? 00.50000
0000000000000 0000000000
80 0500...:
8030000 00000000800005
0000000800005 0000000090
30000500000000 000.00
00000000
0003000003.
0005000003040 10.0050500m
0000000500500 00 00000000000003
00000000000000 00000000000005
00030200000300 0000030050000“:
003000 000.330
0w00m004
000000000000 000000m

Gomudhoﬁmu umumnm

00000000000630 0:0 00030000500005ﬂ

2.02:3

 

0000000082 w008< 0000000000000

000000000 0000000000002 00 0000000003000 000m0w04 030m 00000000000000 mg. 05.0.0.

 

285

S. w a * do. w a i ,8. w at; ”$82

 

mam.
:20.

meg
ﬁn.”

mooo.
m3
moowww-

mom.
mmor

mmm.
***ooo.

mo.
wwﬁ-

So.
v36-

mx A 3i

2

4 m3 352
@2232

258mm

286

In stage one, Mexicans who believe that in order to claim membership in Ameri-
can society they must be citizens, vote in US elections, speak English and believe in
the US Constitution (ethnocultural Americanism) are less likely to engage in at least
one participatory act but individuals who identify as a member of a panethnic group
are more likely to engage in at least one participatory act. Once again, the measures
pertaining to acculturation and structural assimilation are positively related to polit-
ical activity among Mexican Americans. The odds of participating in at least one act
increase as proficiency with English increases and with educational attainment (Table
7.5). Additionally, individuals who have been personally discriminated against are
more likely to participate in the types of political activities considered here. Turn-
ing to the factors that sustain political participation, the results suggest that among
Mexicans, positive attitudes towards ethnocultural Americanism motivates partici-
pation beyond a single act. Therefore, while believing in ethnocultural American
values does not initially compel Mexicans to participate in politics, these beliefs are
significantly related to sustaining political activity over time. Taken together, these
ﬁndings suggest that Latino political assimilation has two components, attitudinal
and behavioral, that are self-reinforcing. It also suggests that political institutions
can be responsible for motivating political assimilation and actually sustaining it.
Further evidence for this fact is that individuals who believe that Latinos are work-
ing together politically are more likely to complete more than one act conditional on
having completed at least one. This ﬁnding suggests that Mexican Americans will be
more likely to participate in the American polity when they perceive group solidarity

exists with respect to achieving their political goals. In the second stage, individuals

287

who do not believe that it is important for Latinos to maintain their distinct cultures
are more likely to participate beyond a single act, as are older individuals and individ-
uals with higher educational levels. Importantly, the effect of panethnic identity and
personal discrimination are much smaller for later transitions. The analysis pertain-
ing to first generation Mexicans (foreign-born) is substantially similar as indicated
by the variables that are significant in the model (Table 7.6). The results suggest
that language, political consciousness and age are the most signiﬁcant predictors of

nonelectoral engagement over time (Table 7.6).

288

 

Table 7.6 Continuation Ratio Logistic Regression of N onelectoral Political

Participation Among First Generation Mexicans

 

 

P 2 1 P > 1|P 2 1
Variable b Z-value p—value b Z -value p-value
Acculturation and Cultural Assimilation
Years in US .246 .42 .674 -.181 -0.13 .897
Language .212 2.53 .011** .519 2.46 .014**
Assimilation .108 -1.08 .281 .026 .12 .906
Retain Culture -.214 -0.79 .427 .284 .51 .607
Identiﬁcational Assimilation
Panethnic Identity .016 .07 .946 .005 .01 .990
Attitudinal Receptional Assimilation
Better in US -.098 -1.44 .150 .117 .80 .427
Group-Based Resources
Political Consciousness .149 .62 .537 .990 1.97 .049**
Personal Disc. .696 2.72 .006*** -.445 —0.85 .396
Discrimination Problem —.045 —0.17 .862 .564 1.00 .320
Structural Assimilation
Education .127 2.09 .037** .121 .96 .336
Income .001 .18 .857 -.012 -1.04 .300
Control Variables
Age .008 .89 .373 .049 2.23 .026**
Female -.486 -1.96 0.05* -.421 -0.82 .415
Married .053 .053 .826 -.051 -0.10 .919
Model Log L -278.520
77. 474
Prob > x2 .1857

 

Notes: ***p g .01, ** p g .05, * p g .10

Turning next Cuban participation, the results show that language, age and per-

289

sonal discrimination are significantly and positively related to political activity in
the ﬁrst transition (Table 7.7). On the other hand, panethnic identity is negatively

related to participation, a finding that echoes the results of previous chapters. In the

second transition, none of the variables were significant, which is probably related to
the small sample size. Among ﬁrst generation Cubans, years in the US and age are
significantly and positively related to participating in at least one participatory act
(Table 7.8). On the other hand, context of reception is inversely related to partici-
pation. First generation Cubans who feel that relative conditions are better in Cuba
are signiﬁcantly less likely to have participated in at least one act. Factors that are
responsible for sustaining nonelectoral engagement beyond a single act among ﬁrst
generation Cubans are years lived in the US and marital status (Table 7.8).

Finally, among Puerto Ricans, experiences with discrimination, perceptions that
Latinos are working together politically and structural assimilation are the factors
most signiﬁcantly related to participation. The only factor found to sustain partici-

pation is income (Table 7.9).

290

 

Table 7.7 Continuation Ratio Logistic Regression of N onelectoral Political

Participation Among Cubans

 

 

P 2 1 P > 1|P _>_ 1

Variable b Z -value p—value b Z -value p—value
Acculturation and Cultural Assimilation

First Gen. .056 .10 .920 .863 .98 .329

Second Gen. .542 .91 .364 .762 .84 .401

Assimilation -.017 -0.15 .878 -.196 —1.00 .317

Retain Culture .175 .61 .544 .674 1.35 .178

Language .254 3.34 .001*** .083 .62 .535
Identificational Assimilation

Panethnic Identity -.479 1.85 065* -.399 .93 .354

Group Consciousness
Political Consciousness -.282 -1.10 .273 -.147 —0.34 .738
Personal Disc. .883 2.80 .005*** .069 .13 .894
Discrimination Problem -.400 -1.42 .156 -.665 -1.34 .180

Structural Assimilation

Education .068 1.23 .218 .089 .13 .894
Income .002 .49 .621 .003 .53 .594
Control Variables
Age .017 2.10 .036** .005 .36 .721
Female -.179 -0.70 .485 -.526 -1.21 .225
Married -.327 -1.28 .202 -.572 -1.24 .216
Model Log L -266.79858
n 354
Prob > x2 .6246

 

Notes: ***p S .01, ** p g .05, * p _<_ .10

291

 

Table 7.8 Continuation Ratio Logistic Regression of N onelectoral Political

Participation Among First Generation Cubans

 

P21

P>HP21

 

 

Variable b Z-value p—value b Z-value p—value
Acculturation and Cultural Assimilation
Years in US 1.41 2.17 .030** 2.79 1.81 .070*
Assimilation .124 .91 .365 -.300 -1.10 .270
Retain Culture .341 .91 .362 .749 1.10 .272
Language .200 1.92 055* -.167 -0.73 .464
Identiﬁcational Assimilation
Panethnic Identity .331 1.05 .293 -.138 -0.23 .821
Attitudinal Receptional Assimilation
Better in US -.247 -2.23 .026** -.081 -0.37 .709
Group-Based Resources
Political Consciousness -.505 -1.54 .124 -.664 1.02 .307
Personal Disc. .670 1.51 .132 1.43 1.51 .131
Discrimination Problem -.563 -1.52 .128 -1.38 -1.60 .110
Structural Assimilation
Education -.000 -0.00 .999 .063 .46 .647
Income .002 .61 .543 .004 .50 .620
Control Variables
Age .025 2.45 .014** .008 .44 .662
Female -.019 -0.06 .952 -.958 -1.52 .130
Married -.089 -0.28 .777 -1.55 -2.29 .022**
Model Log L -169.46567
n 258
Prob > x2 .5771

 

Notes: ***p g .01, ** p S .05, * p S .10

292

 

Table 7.9 Continuation Ratio Logistic Regression of Nonelectoral Political

Participation Among Puerto Ricans

 

P21 P>HP21

Variable b Z-value p—value b Z-value p—value

 

Acculturation and Cultural Assimilation

First Gen. -.320 —0.85 .397 .956 1.42 .156

Assimilation .048 .38 .705 .222 .83 .405

Language -.037 -0.42 1c.675 -.348 —.87 .461

Retain Culture -.324 -1.01 .310 .422 .61 .543
Identiﬁcational Assimilation

Panethnic Identity .420 1.34 .180 -.177 -0.27 .784

Group-Related Resources
Political Consciousness .518 1.66 .096 —.388 -0.58 .561

Personal Disc. .631 1.98 .047 .349 -0.49 .627
Discrimination Problem .125 .40 .686 -.118 -0.18 .856
Structural Assimilation
Education .262 3.61 .000 .113 .68 .499
Income .006 1.00 .320 -.057 -2.33 .020
Control Variables
Age .012 1.02 .308 -.044 -1.50 .134
Female -.318 -1.03 .304 -1.02 -1.59 .113
Married -.219 -0.71 .476 -.196 -0.35 .729
Model Log L -181.20919
n 230
Prob > x2 .4009

 

Notes: ***p S .01, ** p S .05, * p S .10

A Latent Class Model of Latino Participation

Is Latino nonelectoral participation heavily influenced by Latino issues, non-Latino
issues or both? An important question is whether the decision to participate is based

on a strong sense of ethnic awareness and common collective identity or whether it

293

is broader in scope. In 2004, respondents were asked whether their involvement in
nonelectoral politics was related to Latino issues, non-Latino issues, or both? The
ﬁve dichotomous indicators Y1,Y2,Y3, Y4 and Y5 are afﬁrmative statements regarding
degree of nonelectoral political participation. Latent class analysis enabled me to
identify subgroups by participatory and empowerment status.

Table 7.10: Type III Sums of Squares for Goodness of Fit

nonelectoral participation, Mexicans, Cubans and PR

 

 

Model LL x2 AIC BIC
Independence -4106.20 1311.18 1383.18 1581.28
2-class —3689.49 477.74 627.74 1040.45
3—class -3623.99 346.75 574.75 1202.06
3—class restricted -3655.79 410.35 494.35 725.47
4-class —3636. 10 370.98 484.98 798.63

 

Table 7.11: Fit Statistics for Political Participation

 

# Classes -2 11 likelihood G2 (If p E A

Independence -3081.33763 22828.7762 26 .000 .000 .000
2 -2684.78124 66.2812 20 .000 .0397 .7533
3 -2661.56207 12.2681 14 .5848 .1069 .5838

 

A critical part of the analysis is deciding how many latent classes are required.
The ﬁt statistics below suggest independence is strongly rejected. The three class
model provided the most acceptable ﬁt to the data (Tables 7.10 and 7.11). A hier-
archical analysis was performed to restrict conditional probabilities across all model
parameters. The resulting restricted 3—class model yielded a x2 value of 410.35 with
3413 degrees of freedom. Since the computed difference was greater than the critical

value of x2 at the .05 level, the restricted model was retained.

294

7 .3.3 A Latent Class Analysis of Latino Empowerment

The latent class analysis revealed an interpretable solution in terms of high partic-
ipation/ high empowerment, low participation/ low empowerment and low participa-

tion/ high empowerment (Table 7.12).

Latent Class I: Low participation/ high empowerment Class I is comprised
of individuals who are highly non-participatory but highly empoweredl. For exam-
ple, the conditional probability of not having contacted a government ofﬁcial given
membership in this class is .98 for Mexicans and Puerto Ricans and .97 for Cubans
(Table 7.12). The conditional probability of not having contributed money to a po-
litical candidate given membership in this class is .99 for Mexicans, 1.00 for Puerto
Ricans and .98 for Cubans. The conditional probability of not having participated
in other forms of political activity, such as (protesting or attending a public meeting
or demonstration is similarly high across all subgroups. The respondents who par—
ticipated in this class were more likely to report that their participation was related
exclusively to Latino issues as comparted to non-Latino issues or both. For example,
69% of Mexicans, 48% of Puerto Ricans and 43% of Cubans in this class reported
that their demonstration was exclusively related to Latino issues. On this basis, this
class of respondents is considered the group that is best characterized by low levels of
participation but is highly empowered by group—related issues. Puerto Ricans (.45)

are less likely than either Mexicans (.78) or Cubans (.71) to belong to this class (Table

 

1High participation and high empowerment are relative to the average values re-
ported in Table 7.2.

295

7.12 in row entitled “Class Prevalence”).

Latent Class II: Low participation/ low empowerment Class II is comprised
of respondents who are in a low participation, low empowerment class. For example,
the conditional probability of contacting a government ofﬁcial given membership in
this class is only .43 for Mexicans, .15 for Puerto Ricans and .27 for Cubans. The
conditional probability of having worked for a political candidate given membership
in this class is .09 for Mexicans, .03 for Puerto Ricans and .05 for Cubans. Only 31%
of Mexicans, 4% of Puerto Ricans and 23% of Cubans in this class reported attending
a public party meeting or function. The respondents who claim membership in this
class were likely to report that their participation was not exclusively related to Latino
issues. For example, 61% of Mexicans, 73% of Puerto Ricans and 64% of Cubans in
this class reported that their demonstration activity was exclusively related to non-
Latino issues. On this basis, this class of respondents is considered the group that
is best characterized as non-participatory and not motivated by group—related issues.
Puerto Ricans (.40) are more likely than either Mexicans (.14) or Cubans (.25) to

belong to this class.

Latent Class III: High participation/ high empowerment Class III is com-
prised of respondents who are in a high participation, high empowerment class. For
example, the conditional probability of contacting a government official given mem-
bership in this class is .61 for Mexicans, .53 for Puerto Ricans and 1.00 for Cubans.

The conditional probability of having worked for a political candidate given mem-

296

bership in this class is .43 for Mexicans, .28 for Puerto Ricans and .64 for Cubans.
More than 6 out of 10 Mexicans, 4 out of 10 Puerto Ricans and all of the Cuban
respondents in this class reported attending a public meeting or demonstration. The
respondents who claim membership in this class were more likely to report that their
participation was related to Latino issues as compared to non-Latino issues or both.
For example, 62% of Mexicans, 40% of Puerto Ricans and 60% of Cubans in this class
reported that their demonstration was exclusively related to Latino issues. Likewise,
59% of Mexicans, 43% of Puerto Ricans and 64% of Cubans in this class reported
that their volunteer work was exclusively related to Latino issues. On this basis, this
class of respondents is considered the group that is highly participatory and highly
empowered by group-related issues. Puerto Ricans (.15) are more likely than either

Mexicans (.08) or Cubans (.04) to belong to this class.

297

On— Uwuﬂ—Qh Gomuﬁhﬂmﬂoemn

 

a. a. an. t. am. an. on. 5. 2. £0:
2. 8. 9o 8. ma. 8. 8. a. 8. 853-82
a. E. E. 2. no. mo. 8. S. no. 83.3

Cu UQHH—thuﬁoo %®G02
8. R. o: 8. 8. o: an. 3. 3 oz
3 mm. 8. am. 3. 8. no. 8. 8. mar
:Omuoﬂa ho WGmaGQE huhdnm mwOMH—Zom a meuﬁmuu<
2. 2.. o: a. 3. am. 2. we. 3. oz
mm. a... 8. E. 8. :. a. S. we. 2%
EOMHNHHMGOEQQ ho manmummz Ugh—SAM m Ummvﬁmuu<
on. so: a: S. a. 8. E. a. 8. oz
8. so. 8. we 8. 8. Q. 8. S. 8>
mudﬁmﬁgo w you“ h®0933~0> mm ﬁwvmhog
S. E. ma. 2. a. o: S. mm. mm. oz
mm. mm. me. on. we. 8. mm. an. S. as
Gum—Umﬁgo Q Ca hwﬁoz mvmuﬁn—MHHGOO
8. ma. 3.. 2.. mm. am. an. E. ma. 52
o: a. no. 8. 2. no. a. a» S. as
—mmUEO aﬂ®8ﬂh®>ou a ﬁvuuwuﬁoo
aim: aim: aim: aim: aim: aim: aim: aim: aim: 6365>
E6350 mm gauchwz

 

9:550 Una £52m 3.8an

$82on macaw 2536.929an 303:0: 1230298: :0 magmasmm 5683.3: @8880 cam 05m ”was 323.

298

 

we.
co.
mm.
.3.

co.
ow.
ow.

mm.
00.
we.
vm.

oo.
ow.
ow.

3..
co.
av.
3.

mm.
mv.

3.
co.
hm.
mv.

co.
co.
0:.

0:.
co.

o4

co.

00.
0A

00.

mv.
co.
S.
cm.

co.
mm:
wv.

we.
co.
3.
mm.

co.
mm.
mm.

3.
co.
mm.
«.0.

co.

GA

co.

mu. 82:83am 8.20
ea. ﬁom
me. ocﬂaqéoz
um. 95.3

cu poem—o.— Managua—30>
co. 50m
8. 2584.82

me. 023:5

299

Finally, a multinomial model was estimated to predict membership in each class
(Table 7.13). The following generalizations are shown by the table. The results show
that Latinos are more likely to be in class III (high participation / high empowerment)
versus class I (low participation/high empowerment) with English proﬁciency, per-
sonal discrimination, income and education. Individuals are less likely to be in class
III if they believe that life is relatively better in their country of origin than in the US
and if they perceive discrimination against Latinos to be problematic (Mexicans and
Cubans). Mexicans and Cubans who believe that Latinos work together politically
(political consciousness), who prefer to identiﬁfy as Latino/ Hispanic and who believe
that it is important for Latinos to maintain their unique cultures rather than assimi-
late are more likely to be in an empowered class (e.g. Class 1) than in a disempowered

class (e.g. Class II).

300

S. as
s. 0:
8. 3m
2. as
8. $2
8. 32
S. 3.
8. 0.2.
E. Z:
S. no
8. 2...:
8. was
8. 0%
8. as:
a .5 4

3m.
3.

.3.

NA
mA

AA

8.
NA
3.

on.
N.N
©.mA
we.

£0

mod- om.
3.0- mm.
mod- CA
3.0- .3.
3. AA
Av. NA
no. AA
nmd- 2.
cm. 0N
oA- N.N
mod- Nm.
omd- NA
wn. ON
wN ON
©.N- om.
MHQHQ .m.o
50.230

Ed-
33-
No.

mod-

3.
NA.

mo.

©A-
5w.
Am.
3A.-

3.
am.
8.
NA-

NaAa

AA
mm.
3.

Am.

AA
NA

mA

mA
.vN
NA
cm.

.3.
mA
mN
mm.

4&0

mo.
vod-
mod-

AN.-

36
on.

mm.

Nvd
cw.
NA.
on.-

end-
ow.
0w.
mm.

ma~¢

mN.
mm.
oA

mm.

3.
3.

gm.

ow.
AN
mA
m.m

«13A
AA
3:.
N.N

AHO

000$ oﬁmsm

mA-
$0.0-
oo.

36-

0o.-
mod-

Ed-

35.0-
Nu.
mm.
NA

em.

3.
mud-

mu.

NLAQ

mA
cm.
CA

N».

NA
oA

GA

3.

NA

3.

3.

on.

u?
01

Pm
A0.

8.0

9.. S: «S. 8:52
Ed. 2: 2. 236a
8. o: no. 63.

0030205 33:00

was. 8. a..- «as: 826a
83. oasﬁm

mA. moA 30. 08005

3. mA 00. 253000va

somuwzammm< 12:00:05

no. a. 2.- m: s 388:
363265650302
3o- mm. 3:. 6638a .85

3. 3A No.
mod- mA wN.
Exo- mA mm.

.005 Ecomaam
.2 6832:6583

00000000300 E0320m

mmmcmaomomaonv 95:0

oA- wA oo. 05:00 530m

03. 3A New. owmsw:0A
mA NmN N3. .000 0:000m
mav- Nm. AA- .000 HES

GEE—1:004 3.5030 0:: somamasﬁaooé.
$0 «in

000082

maAo

005090Samm 1200020002 0503 00 0:030:54 00.30 Ewan-H 03mmw0A 32:05:52 UNA.» 050R.

301

7 .4 Summary and Conclusions

This chapter has explored the relative effects of group identity, consciousness, accul-
turation and assimilation on Latino nonelectoral political participation. The results
tend to support several hypotheses of both the straightline and ethnic disadvan-
tage models of immigrant assimilation as they related to political participation. For
example, linguistic proﬁciency was found to be signiﬁcantly related to nonelectoral
engagement across all models. Years lived in the US was a signiﬁcant predictor of
motivating and sustaining political activity among ﬁrst generation Cubans. There is
also substantial evidence that individuals who prefer cultural retention over assimila-
tion are less likely to engage in politics. Finally, many of the measures of structural
assimilation were positively related to nonelectoral engagement.

Several of the group-resmirce variables were statistically signiﬁcant. For example,
having been the object of discrimination was a consistent predictor of political activity.
Context of reception was also signiﬁcant, thereby lending support for the hypothesis
that Latinos with favorable contexts of reception are more likely to participate in

politics.

302

Chapter 8

Summary and Conclusion

Some scholars and commentators believe that the recent increase in the Latino popu-
lation will destroy the fabric of American society and result in linguistic fragmentation
and cultural hegemony. In contrast to immigrants of European descent, who accepted
their inevitable transformation into American political culture, post-1965 immigrants,
they claim, are maintaining political and social ties to their country of origin, seg-
regating themselves into ethnic enclaves and refusing to assimilate into American
society. Recently, scholars have challenged classical assimilation theories, suggesting
instead that the process of assimilation for America’s newcomers is best characterized
as “bumpy” and not a straight line. This dissertation has provided an empirical test
of the various assimilation theories as they pertain to Latino political incorporation.

More speciﬁcally, in this dissertation, I described the pattern of political incorpo-
ration among Mexican, Puerto Rican, and Cuban immigrants to determine whether
participation increases with exposure, identiﬁcational and structural assimilation,

context of reception and access to group—related resources. I examined the political

303

assimilation of Latinos by national origin group. Clearly, assimilation is a multidi-
mensional process, involving structural factors, language acquisition, and identiﬁca-
tion as American. There is also a political dimension to assimilation that involves
naturalization, the acquisition of American values, and participation in the political
process. Each chapter generally supports a conclusion that exposure to American
cultural norms, context of reception, discriminatory processes and access to group—
related resources alter the distribution of Latinos who identify with and participate
in the American polity.

I have shown that generally speaking, Mexican, Puerto Rican and Cuban immi-
grants do become politically incorporated over time. The longer they live in the
United States, the more likely they are to vote, naturalize and engage in nonelec-
toral types of political behavior. There is no evidence that recent Mexican, Puerto
Rican or Cuban immigrants are likely to be permanently disengaged from the elec-
toral process. For example, all three subgroups have a large class of individuals who
are highly electorally responsive. In comparing the citizenship rates and turnout of
Latinos by national origin group, I considered whether the “straightline” pattern of
assimilation applies to voting by comparing the political participation of different
immigrant generations within each ethnic group. The results strongly favored the
straightline assimilation model but there is more to the story. I demonstrated as well
that context of reception alters the landscape of participation by decreasing the like-
lihood that individuals adopt an American identity versus identiﬁcation as a member
of their country of origin or as a member of their panethnic group.

In Chapter Four, I demonstrated that the structure of beliefs about being Ameri-

304

 

can and American values is partially homogeneous across national origin groups with
some heterogeneous components. Conditional latent class probabilities revealed that
Mexicans are more likely than Cubans or Puerto Ricans to internalize ethnocultural
conceptions of Americanism. Nevertheless, this class of respondents comprised the
largest prevalence of Latinos across each ethnic group. In other words, the ethno-
cultural American class comprised the largest percentage of Mexicans, Cubans and
Puerto Ricans. The ﬁndings also indicated that membership in this class depends on
the context of immigrant reception, more speciﬁcally notions that Latinos are being
discriminated against as a group. Acculturation and exposure to American cultural
norms also predicted membership in the ethnocultural American class which suggests
that as Latinos assimilate they increasingly identify with American political culture.
This chapter also distinguished between Latinos who cluster around American liberal—
ism versus American liberalism coupled with what I called governmental paternalism.
Overall, the analysis showed that Latinos do accept the tenets of American liberal-
ism but differences exist regarding the role of government as limited versus expansive
and in the role of individual self—reliance versus governmental protection. Another
important ﬁnding from this chapter is the importance of the context of immigrant
reception to shape ethnic identity choices, particularly the preference for opting for
an American identity. This is important because American identity is shown to be
an important predictor of political behavior, especially among Cubans.

In Chapter Five, I turned to voting, registration and citizenship as important
indicators of political assimilation. Among all ethnic groups, age, education, and

income were positively associated with the decision to acquire citizenship, register

305

and vote in US elections. Structural assimilation as measured by degree of economic
attachment in addition to education and income clearly affects the level of political
assimilation, with the effect being more substantial among Mexicans and Puerto
Ricans. The various forms of structural assimilation differ, however, by subgroup.
For example, economic attachment is relatively more important in the decision to
register and vote among Mexicans than it is for Puerto Ricans and Cubans, who are
motivated more by educational attainment. Again, exposure to US cultural norms,
American identiﬁcation and increasing ties to the United States were shown to be
signiﬁcantly related to political behavior among ﬁrst generation immigrants. The
effect of group based resources varied signiﬁcantly across ethnic origin groups. For
example, in contrast to ethnic disadvantage models of assimilation, which suggest that
group consciousness facilitates political assimilation, I found that Cubans who believe
in a common linked fate between all Latinos are signiﬁcantly less likely to vote. On the
other hand, identiﬁcation as American was positively related to political participation.
Perceptions of linked fate had no effect on either Puerto Rican or Mexican political
behavior. These results suggest that Cubans internalize American values and as they
assimilate simultaneously shed their identities as Cubans. For example, Cubans are
more likely to identify as American versus Latino/ Hispanic or as a member of their
country of origin over time. They are also more likely to be oriented towards US
politics as ties to Cuba decrease. Cubans who identify with their country of origin
are signiﬁcantly less likely to participate in politics while Cubans who identify as
American are signiﬁcantly more likely to do so. Turning to the interaction between

context of reception and identity, the results of this chapter showed the protective

306

effect of identiﬁcation as American. Among all subgroups, individuals who either
perceived discrimination to be a problem or were themselves victims of discrimination
were more likely to vote and / or register if they adopted an American identity rather
than a panethnic identity or the identity of their ethnic origin group.

There also are differences in the citizenship rates among Cuban and Mexican im-
migrants. Acculturated Mexicans and Mexicans who abandon their ties to Mexico
naturalize more quickly than unacculturated Mexican immigrants. In addition, we
saw that perceptions of discrimination signiﬁcantly decreased waiting time. Inter-
estingly, psychological attachment to one’s homeland had differential effects among
Mexicans and Cubans. Cubans who believed that Cuba as opposed to the US was
their “true homeland” waited longer before seeking citizenship status but the oppo-
site held true among Mexicans. Again, this suggests that Cubans are less conﬂicted
about assimilating than are Mexicans, who prefer to retain some aspects of their
culture rather than fully assimilating.

Finally, with respect to nonelectoral participation, several important ﬁndings
emerged. First, the factors that sustain participation are not necessarily the same
factors that instigate it. For example, among Mexicans, ethnocultural Americansim
sustains political activity over time. Structural assimilation, acculturation and ex-
posure affect the political incorporation of immigrants. Interestingly, discriminatory
context is responsible for motivating participation, not necessarily sustaining it, at
least among Mexicans. Again, we saw how Cuban political behavior is suppressed by
the notion of a common linked fate as measured by their propensity to identify as a

member of a panethnic group. Context of reception was also a signiﬁcant predictor of

307

political participation among Cubans, as individuals who believe that conditions in
the US are relatively better than in their country of origin were found to be more likely
to participate. Again, among Puerto Ricans, political consciousness and structural
assimilation were the most important predictors of political assimilation.

The evidence is in favor of the assertion that the process of political incorporation
of the current wave of immigrants to the US is similar irrespective of national origin.
Second and third generation immigrants and those who have lived in the US the
longest are more likely to vote than ﬁrst-generation immigrants. Among the foreign-
born, time spent in the US. boosts the level of political participation. Moreover,
as socio—economic status improves, participation increases. Although these ﬁndings
strongly support the straightline model of immigrant assimilation, the role of discrim-
ination in motivating political activity and the psychological aspects that accompany
identity transformations as immigrants assimilate must not be overlooked.

Table 8.1 summarizes the theoretical predictions and empirical results with respect

to the hypotheses laid out in Chapter 5.

308

 

 

808$ 808$ 808$ 030200005 030200005 030200005 + 03000800:
300030000$ 000300384 020.

808$ 808$ 808$ - 030200005 030200005 + 030008050
0000000300000 00000 000$.

808$ 808$ 808$ 030200005 030200005 - + 0300500wa
000003000 205m 020.

808$ 808$ 800006 030200005 030200005 - - 0300509n$
003389000 005$ 02$.

808$ 808$ 808$ 030200005 030200005 030200005 - 0300280»:
08020062 2:.

80096 80096 800006 + + + + 0300200003
888$ 30008006 000.

808$ 808$ 808$ - 030200005 030200005 + 030000000$
3303050 03800$ 000.
80096 80096 80096 + + + + 0300000003 0038330030000
3000380030005 00H.

808$ 808$ 808$ - - - + 030008030
0000:300$ 00200 000.

80096 80096 800006 + + + + 030050003
0038:0000-4- 380200

000.000 00003$ 08000 0000902 000000 00003$ 0800.0 0000002
003020000 08000.0 0038:0000 0300000003
00.80000 3003800009

003.2030m 3003200 0038.5 00 00300< 020 000.00 02000$ 300830050 3000 00030800me 3003800009 A6 0308

309

000.80

808$

800006
80096

80 8$

800006

80 8$

808$

800006
800006

808$

800006

80.80

80.80

800006
800006

808$

800006

030200005

030200005

0>W6ASUA$OOA$H

+

030200005

IT

0>~65~UA$OUA$H

+

00365802006;

...—I

0300280»:

00383002 3000200 00$.
0300080m$

00000005 030802 00$.

030008050 00000003“.— 00$.
030000000$

30003800000 00$.
030008002

2005 880$: 00$.
03000000m$

00380308005 00$.

310

8. 1 Policy Implications

From a policy perspective, the question becomes, what do we do, as a society, to
increase/support the probability that immigrants will politically act in a way that
supports US ideals concerning democratic processes? According to Schmid (2004),
“political participation is often the result of emotion and people feel that their iden-
tity is at stake. In this context, they may not calculate individual cost and beneﬁt.
When people feel deprivation and injustice, they may bear large costs to reform in-
stitutions, a kind of reform utility that differs from ordinary goods.” Group—related
resources such as perceived discrimination and group based identiﬁcation should be
considered a political resource that shifts the distribution of costs and beneﬁts and
motivates Latino political behavior. The fact that the mere perception of group-based
discrimination motivates Latino political behavior suggests that when discriminatory
barriers block an individualistic pattern of social or economic mobility (Alba and Nee,
2003), political assimilation depends on collectivist strategies that may or may not
be beneﬁcial. Clearly, discrimination acts to encourage Latinos to rectify a perceived
wrong and eradicate inequalities in opportunity. Attitudinally, however, discrimina-
tion tends to discourage immigrants from adopting American values and attitudes,
so to the extent that American society discriminates against Latinos, we “shoot our-
selves in the foot.” The controversy concerns the fact that many people believe that
immigrants will vote in self serving ways and, that given their increasing numbers
that the “American way” will be compromised. To the extent that I could, in this

dissertation, I provided evidence that immigrants are motivated to participate in

311

self-serving ways. Nevertheless, immigrants are attitudinally oriented to American
political life. Since immigrants want to participate, policies should be developed to
promote participation in positive ways and discrimination should be discouraged. In
the absence of discrimination, immigrants will still participate, but the nature of the
participation will change.

The results also suggest that Americans should encourage immigrants to assimi-
late (Chavez, 1995). For example, immigration policies must be expanded to include
how immigrants are incorporated into American society. While the evidence suggests
that immigrants are assimilating into American society, measures can be undertaken
to facilitate it. For example, if immigrants were given the resources and opportunity
to assimilate, one consequence would be the “preservation of American political cul-
ture.” For example, speaking English is a key ingredient in forging a sense of national
identity. This suggests that in schools, English immersion programs would help Lati-
nos, who appear eager to learn English, learn it at faster rates and this would in turn
facilitate American national identity and unity. This would also effectively facilitate
structural assimilation, which is clearly related to political attitudes and behavior.
Affirmative action policies would also help Latinos assimilate more quickly. Segre-
gated schools tend to reify ethnic based conceptions of identity and hence suppress

political activity.

312

Appendices

313

APPENDIX A

The model and its component parts are as follows:

1. There are ﬁve assimilation factors, as indicated by the four circles labaled F 1 -

F5.

2. The ﬁve factors are intercorrelated, as indicated by the two-headed arrows.

3. There are 35 manifest (observed) variables, as indicated by 35 rectangles. They
each represent components of scales, subscales or questions derived from the

2002 National Survey of Latinos (see Figures 1.1-1.4).

4. The observed variables load on the factors in the following pattern: struc-
tural assimilation loads on F1; Cultural and linguistic retention load on F2;
attitudinal behavioral and reception assimilation load on F3; identiﬁcational

assimilation loads on F4 and political assimilation loads on F5.

5. Each observed variable loads on one and only one factor.

6. Measurement errors associated with each observed variable are uncorrelated.

The primary task with respect to this model is to determine the goodness of ﬁt
between the hypothesized dimensions of assimilation and the observed data. The
null hypothesis is that the model ﬁts the data. Any discrepancy between the sample
covariance matrix and the population matrix is captured by the residual covariance

matrix (Byrnes, 2006).

314

In examining the standardized residual information it is apparent that the av-
erage off—diagonal value is .0255, whereas the largest off-diagonal value is .0319,
both of which reflect a very good ﬁt to the data. A review of the frequency dis-
tribution reveals that all residual values fall between -.1 and .1. The independence
model is soundly rejected (2223.222,df = 36). From this information, we can con-
clude that although there may be some minimal discrepancy in ﬁt between the
hypothesized model and the sample data, overall the model as a whole appears
to ﬁt the data very well. In addition, all three indexes discussed in chapter 3
(NF! 2 .940, NNFI = .912,CFI = .950) were consistent in suggesting that the
hypothesized model represented an adequate ﬁt to the data. Finally, the RMSEA
value for the hypothesized model is .048 with a 90% conﬁdence interval ranging from
(.041 — .056), which represents a good degree of precision. Given that (a) the RMSEA
point estimate is < .05 and (b) the upper bound of the 90% conﬁdence interval is
< .06, we conclude that the estimated model adequately describes the datal.

In reviewing the unstandardized solution, it is apparent that the estimates to be
reasonable and the majority are statistically signiﬁcant; all standard errors are fairly
small. The R—squared value represents the proportion of the factor variance that
is explained by its corresponding measurement. Again, in reviewing the output, it
is clear that parameter values are consistent with the literature. For example, lan-
guage acquisition and retention explains 87.7% of the variance in cultural assimilation

while primary mode of identiﬁcation explains 53.7% of the variance in identiﬁcational

 

1This is not the ﬁnal model that was tested, but a pictorial representation of the
model. The software I used was not able to provide me with a graphical representa-
tion.

315

assimilation.

316

Bibliography

317

Bibliography

Aguirre Jr., A., and Jonathan H. Turner. American Ethnicity: The Dynamics and
Consequences of Discrimination, 3rd ed. Boston, MA: McGraw-Hill, 2001.

Aguirre, B.E., Saenz, Rogelio, and Hwang, Sean-Shong. (1989). “Discrimination and
the Assimilation and Ethnic Competition Perspectives.” Social Science Quarterly,
70(3): 595-606.

Aguirre, A. and Turner J. American Ethnicity: The Dynamics and Consequences of
Discrimination. Mayﬁeld, 2001.

Alba, Richard. (2006). “Mexican Americans and the American Dream.” Perspectives
on Politics, 4(1): 289—296.

Alba, RD, and Chamlin, MB. (1983). “A Preliminary Examination of Ethnic Iden-
tiﬁcation Among Whites.” American Sociological Review, 48, 240-247.

Alba, R. and Nee, V. 1997 “Rethinking assimilation theory for a new era of immigra-
tion”, International Migration Review, 31(4), p. 826-74.

Alex-Assensoh, Y., and Assensoh, A. B. (2001). “Inner-city contexts, church at-
tendance, and African-American political participation.” Journal of Politics, 63,
886-901.

Alvarez, R. M., and Lisa Garcia Bedolla. (2003). “The Foundations of Latino Voter
Partisanship: Evidence from the 2000 Election.” Journal of Politics, 65(1):31-49.

Alvarez, R. Michael and John Brehm. (2002). Hard Choices, Easy Answers. Prince—
ton.

Anderson, B., (1983). Imagined Communities: Reﬂections on the Origin and Spread
of Nationalism. London: Verso.

Andersen, Kristi and Elizabeth F. Cohen. (2002). “Political Institutions and the
Incorporation of Immigrants.” Program in American Democracy Working Paper,
University of Notre Dame.

Antunes, G., and Gaitz, C. M. (1975). “Ethnicity and participationzAstudy of
Mexican—Americans, Blacks and Whites.” American Journal of Sociology, 80(5),
1192-1211.

Barboza, Gia and Rohan Williams. (2005). “The Effect of Family Structure and
Family Capital on the Political Behavior of Latino Immigrants in the United States,”

318

paper prepared for presentation at the 2005 Annual Meeting of the lvlidwest Political
Science Association.

Barretto, Matt. (2004). “The Role of Latino Candidates in Mobilizing Latino Voters:
Revisiting Latino Vote Choice.” Prepared for Latino Politics: State of the Discipline,
Texas A&M University, College Station TX, April 2004.

Barreto, Matt A., and Jose A. Munoz. 2003. “Reexamining the ‘Politics of In-
Between’: Political Participation Among Mexican Immigrants in the United States.”
Hispanic Journal of Behavioral Sciences, 25, p. 42747.

Bautista del Demanico,Y., Crawford, 1., and DeWolfe, A. S. (1994). “Ethnic identity
and self-concept in Mexican-American adolescents: Is bicultural identity related to
stress or better adjustment?” Child and Youth Care Forum, 23, 197-206.

Bean, Frank, Susan K. Brown, and Ruben G. Rumbaut. (2006). “Mexican Immigrant
Political and Economic Incorporation.” Perspectives on Politics, 4(1): 309-313.

Bean, Frank and Gillian Stevens. (2003). America’s Newcomers and the Dynamics
of Diversity, Russell Sage Foundation Publications.

Bedolla, L. (2003). “The Identity Paradox: Latino Language, Politics, and Selective
Dissociation.” Latino Studies, 1: 264-283.

Berry, J. W. (1990). “Psychology of acculturation.” In J. Berman (Ed.), Nebraska
Symposium on Motivation, 1989: Cross-cultural perspectives (pp. 201 - 234). Lin-
coln: Nebraska University Press.

Berry, J. W. (1997). “Immigration, acculturation and adaptation.” Applied Psychol-
ogy: An International Review, 46 (1), 5 - 68.

Birman, D., 'I‘rickett, E. and Vinokurov, A. (2002). “Acculturation and Adaptation of
Soviet Jewish Refugee Adolescents: Predictors of Adjustment Across Life Domains.”
Journal of Community Psychology, 30(5): p. 585-607.

Bobo, Lawrence and Frank Gilliam. (1990). “Race, Sociopolitical Participation and
Black Empowerment.” American Political Science Review, 84(2): 379-393.

Boyd, Monica. (2006). “Educational Attainments of Immigrant Offspring: Success
or Segmented Assimilation?” Unpublished Manuscript.

Brant, R. (1990). “Assessing Proportionality in the Proporitonal Odds Model for
Ordinal Logistic Regression.” Biometrical Journal, 42(6): 677-699.

319

Browning, Rufus, Dale Rogers Marshall and David Tabb. (1984). Protest is Not
Enough: The Struggle of Blacks and Hispanics for Equality in Urban Politics. Berke-
ley: University of California Press.

Calderon, Jose. (1992). “’Hispanic’ and ’Latino’: The Viability of Categories for
Panethnic Unity.” Latin American Perspectives, 19(4) 37-44.

Calhoun-Brown, A. (1996). African American churches and political mobilization:
The psychological impact of organizational resources. The Journal of Politics, 58,
935-953.

Cain, Bruce E., D. Roderick Kiewiet, and Carole J. Uhlaner. (1991). “The Acquisi-
tion of Partisanship by Latinos and Asian Americans.” American Journal of Political
Science, 35:390-422.

Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes.
(1960). The American Voter. Chicago: University of Chicago Press.

Castles, S. and Miller, M. J. (1998). The Age of Migration: International Population
Movements in the Modern World, Second Edition, London: Macmillan.

Chong, Dennis, and Rene] Rogers. 2005. “Racial Solidarity and Political Participa-
tion,” Political Behavior, 27:4 (December): 347-374.

Chunrong, Ai, and Norton, Edward C., “Interaction Terms in Nonlinear Models”
(July 2001). Triangle Health Economics Working Paper No. 2. Available at SSRN:
http://ssrn.com/abstract=303862 or DOI: 10.2139/ssrn.303862

Cisneros, H. Respuesta A Huntington, Financial Times, 24 de Mayo, 2004. Avail-
able online at http: / / rosasl46.blogspot.com / 2004 / 06/ respuesta—huntington—ﬁnancial-
timeshtml.

Citrin, Jack and Donald. P. Green. (1986). “Presidential Leadership and the Resur-
gence of Political Trust.” British Journal of Political Science, vol. 16.

Citrin, Jack, Beth Reingold, and Donald P. Green. (1990). “American Identity and
the Politics of Ethnic Change.” Journal of Politics.

Citrin, Jack, Kathryn Pearson, John Sides, and David O. Sears. (2002). “National
and Ethnic Identities: Competing or Complementary,” paper prepared for presenta-
tion at the 2002 Annual Meeting of the Midwest Political Science Association.

Citrin Jack and Highton B. (2002). How Race, Ethnicity, and Immigration Shape the
California Electorate. San Francisco: Public Policy Inst. Calif.

320

Citrin, Jack, Kathryn Pearson, Jonathan Cohen. (2003). “Assimilation of the Fourth
Wave: Identity, Attitudes, and Participation.” paper prepared for presentation at
the 2003 Annual Meeting of the Midwest Political Science Association.

Claassen, Ryan L. (2004). “Political Opinion and Distinctiveness: The Case of His-
panic Ethnicity.” Political Research Quarterly, 57:609-620.

Connaughton, S. L. and Jarvis, S. E. (2004). Invitations for partisan identiﬁcation:
Attempts to court Latino voters through televised Latino—oriented political advertise-
ments, 1984-2000. Journal of Communication, 54, 38—54.

Conover, Pamela Johnston. (1984). “The Inﬂuence of Group Identiﬁcation on Polit-
ical Perception and Evaluation.” Journal of Politics, 46: 760-785.

Conover, Pamela Johnston. (1988). “The Role of Social Groups in Political Think-
ing.” British Journal of Political Science, 18:51-76.

Converse, Philip E. and Roy Pierce. (1985). “Measuring Partisanship.” Political
Methodology, 11:143-166.

Cullar, 1., Arnold, B.,and Maldonado, R. (1995). “Acculturation rating scale for
Mexican Americans-ILA revision of the original ARSMA scale.” Hispanic Journal of
Behavioral Sciences, 17(3), 275-304.

Dahl, Robert. Who Governs? New Haven: Yale University Press, 1962.

Dawson, Michael C. (1994). Behind the Mule: Race and Class in African-American
Politics. Princeton, NJ: Princeton University Press.

de la Garza, Rodolfo, Louis DeSipio, F. Chris Garcia, John A. Garcia, and Angelo
Falcon. (1992). Latino Voices: Mexican, Puerto Rican, and Cuban Perspectives on
American Politics. Boulder, CO: Westview Press.

de la Garza, R.O., A. Falcon, and RC. Garcia. (1996). “Will the Real Americans
Please Stand Up: Anglo and Mexican—American Support of Core American Political
Values.” American Journal of Political Science, 40: 335-351.

de la Garza, Rodolfo, Angelo Falcon, F. Chris Garcia, and John A. Garcia. (1998).
Latino National Political Survey, 1989-1990. Ann Arbor, MI: Inter-university Con-
sortium for Political and Social Research.

DeSipio, Louis. (1995). Counting on the Latino Vote: Latinos As a New Electorate.
Charlottesville, VA: University of Virginia Press.

321

DeSipio, Louis. (2002). Immigrant Organizing, Civic Outcomes: Civic Engagement,
Political Activity, National Attachment and Identity in Latino Communities. Uni-
versity of California at Irvine, Center for the Study of Democracy. Working Paper
02-08. http://repositories.cdlib.org/csd/02-08.

DeSipio, Louis. (2006). Transnational Politics and Civic Engagement: Do Home
Country Political Ties Limit Latino Immigrant Pursuit of US. Civic Engagement
and Citizenship. In Transforming Politics, Transforming America: The Political and
Civic Incorporation of Immigrants in the United States, University of Virginia Press.

De Toqueville, Alexis. (1956). Democracy in America. New York: New American
Library.

Diaz-Cotto, Juanita. (1996). Gender, Ethnicity and the State: Latina and Latino
Prison Politics. Albany, NY: SUNY Press.

Dominguez, G. (1996). “Do Latinos Exist?,” Contemporary Sociology, 23(3), p. 354-
356.

Dwyer, C. (1991). The Dominican Americans, Chelsea House, New York.

77

Eckstein, Susan. 2006 “Cuban Emigres and the American Dream. Perspectives on

Politics, 4(1): 289-296.

Edmonston, Barry and Jeffrey S. Passe] (eds). (1994). Immigration and Ethnic-
ity: The Integration of America’s Newest Arrivals. Washington, DC: The Urban
Institute Press.

Finifter, Ada W. and Bernard M. F inifter. (1989) “Party Identiﬁcation and Political
Adaptation of American Migrants in Australia.” The Journal of Politics, 51(4): 599-
630.

Fox, G. (1996). Hispanic nation: culture, politics, and the constructing of identity.
Secaucus, NJ: Carol Publishing Group.

Fraga, Luis Ricardo and Gary M. Segura. (2006). “Culture Clash? Contesting
Notions of American Identity and the Effects of Latin American Immigration.” Per-
spectives on Politics, 4(June): 279-287.

Fuchs, Lawrence. (1990). The American Kaleidoscope. London: Wesleyan University
Press.

Gans, Herbert. (1992). Second generation decline: scenarios for the economic and
ethnic futures of the post-1965 American immigrants. Ethnic and Racial Studies, 15
(April): 173-191.

322

Gay, Claudine. (2001). “The Effect of Black Congressional Representation on Polit-
ical Participation,” American Political Science Review 95(3): 589-602.

Glazer, Nathan. (1994). “Immigration and the American Future,” The Public Inter-
est.

Glazer, Nathan. (1997). We are all lVIulticulturalists Now. Cambridge, Mass: Har-
vard University Press.

Glazer, Nathan. and Daniel Moynihan. 1963. Beyond the Melting Pot. Cambridge,
Mass: Harvard University Press.

Gleason, Phillip. 1980. “Americanization,” in Harvard Encyclopedia of American
Ethnic Groups, S. Thernstrom ed. Cambridge, Mass: Harvard University Press.

Gordon, Milton M. 1964. Assimilation in American Life. New York: Oxford Univer-
sity Press.

Gowan, M., and Trevio, M. (1998). “An examination of gender differences in Mexican-
American attitudes toward family and career roles.” Sea: Roles, 38( 11 / 12), 1079-1093.

Graham, PM. (199). “Reimagining the nation and deﬁning the district: Dominican
migration and transnational politics”, in P. Pessar (Ed), Caribbean Circuits, New
Directions in the Study of Caribbean Migration, Center for Migration Studies, New
York, p. 91-125.

Gurin, Patricia, Arthur Miller, and Gerald Gurin. 1980. “Stratum Identiﬁcation and
Consciousness.” Social Psychology Quarterly, 43: 30—47.

Guzmn, Betsy and Eileen Diaz McConnell. (2001). “The Hispanic Population: 1990-
2000 Growth and Change.” Population Research and Policy Review, 21 (1-2): 109—
128.

Hagenaars, Jacques. (1993). Loglinear Models with Latent Variables. Sage University
Press.

Hajnal, Zoltan and Taeku Lee. (2003). “The Multidimensionality of African Amer-
ican Party Identiﬁcation.” Paper presented at the 3rd Minnesota Symposium on
Political Psychology Perspectives on Race and Politics. Twin Cities, MN.

Harris, F. (1994). “Something within: Religion as a mobilizer of AfricanAmerican
political activism.” The Journal of Politics, 56, 42-68.

Haynie, L. (2001). African American Legislators in the American States. New York:
Columbia University Press, 2001.

323

Hernndez, R., and G. Jacobs (2001). “Beyond homeland politics, Dominicans in
Massachusetts”, in C. Hardy-Fanta and J .N . Gerson (Eds), Latino Politics in Massa-
chusetts, Struggles, Strategies, and Prospects, Routledge, New York, p. 277-295.

Hertel, Bradley R., and Michael G. Hughes. (2001). “The Signiﬁcance of Color
Remains: A Study of Life Chances, Mate Selection, and Ethnic Consciousness Among
Black Americans” Social Forces, 68(1): 1105 - 1121.

Hetherington, Marc. (1998). “The Political Relevance of Political Trust.” The Amer-
ican Political Science Review, 92(4): 791-808.

Highton, B. and Burris, A. (2002). “New Perspectives on Latino Voter Turnout in
the United States.” American Politics Research, 30(1):285-306.

Hirschman, Charles. (2003). “The Role of Religion in the Origin and Adaptation of
Immigrant Groups.” Paper presented at the conference on ”Conceptual and Method-
ological Developments in the Study of International Migration.” Center for Migration
and Development, Princeton University. May 23—24.

Hu, L., and Bentler, P. M. (1999). “Cutoff criteria for ﬁt indexes in covariance struc-
ture analysis: Conventional criteria versus new alternatives.” Structural Equation
Modeling, 6, 1—55.

Huddy, Leonie. (2001). “From Social to Political Identity: A Critical Examination
of Social Identity Theory.” Political Psychology, 22(1): 127-156.

Huntington SP. (1996). The Clash of Civilizations and the Remaking of World Order.
New York: Simon and Schuster. 367

Huntington SP. (1997). “The Erosion of American National Interests.” Foreign
A ﬂairs, 76: 28—49

Huntington SP. (2000). Reconsidering immigration: Is Mexico a special case? Cent.
Immigration Stud. http://www.cis.org/articles/2000/back1100.html

Hurtado, Aida. (1994). “Does Similarity Breed Respect? Interviewer Evaluations
of Mexican-Descent Respondents in a Bilingual Survey.” Public Opinion Quarterly,
58:77-95.

Hwang, S. S. and Murdock, S. H. (1991). “Ethnic Enclosure or Ethnic Competition:
Ethnic Identiﬁcation among Hispanics in Texas.” Sociological Quarterly, 32, 469-476.

324

Jackman, M. R., and Jackman, R.W. (1973). “An interpretation of the relation
between objective and subjective social status.” American Sociological Review, 38,
569-582.

Jackson, Byran 0. (1987). “The Effects of Racial Group Consciousness on Political
Mobilization in American Cities” Western Political Quarterly, 40(Decernber 1987):
631—647.

Jones-Correa, Michael. (1998). Between two nations : the political predicament of
Latinos in New York City. Cornell University Press.

Jones-Correa, Michael and David L. Leal. (1996). “Becoming ’Hispanic’: Secondary
Panethnic Identiﬁcation among Latin American-Origin Populations in the United
States.” Hispanic Journal of Behavioral Sciences, 18(2):214—254.

Jones-Correa, Michael and David L. Leal. (2001). “Political Participation: Does
Religion Matter?” Political Research Quarterly, 54:751-770.

Jones-Correa, Michael. (2001). “Under two ﬂags: dual nationality in Latin America
and its consequences for naturalization in the United States”, International Migration
Review, 35(4), p. 997-1029.

Kaufmann, Karen M. (2003). “Cracks in the Rainbow: Perceived Commonality as
a Basis for Latino and African-American Political Coalitions.” Political Research
Quarterly, 56:2:199-210.

Keefe, S. and A. M. Padilla. (1987). Chicano Ethnicity. Albuquerque: University of
New Mexico Press.

Kenski, K. and Tisinger, R. (2006). “Hispanic Voters in the 2000 and 2004- Presiden-
tial and General Elections.” Presidential Studies Quarterly, 36(2):189

King, Gary, Michael Tomz, and Jason Wittenberg. (2000). “Making the Most of
Statistical Analyses: Improving Interpretation and Presentation.” American Journal
of Political Science, 44 (2)1341-355.

LaFromboise, T., Coleman, H., and Gerton, J. (1993). “Psychological impact of
biculturalism: Evidence and theory.” Psychological Bulletin, 114, p. 395—412.

Lee, T. and Ramakrishnan, S. K. (2002) “What’s in a Name? Eth-
nic Names, Immigrant Acculturation, and Latino Mass Opinion.” Paper
presented at the annual meeting of the American Political Science Associ—
ation, Boston Marriott Copley Place, Sheraton Boston and Hynes Conven-
tion Center, Boston, Massachusetts Online i.PDF;J Retrieved 2006-10-05 from
http: / / www.allacademic.com / meta / 1366249,,ndeazhtml

325

Levitt, P. (1998). “Social remittances: migration driven local-level forms of cultural
diffusion”, International Migration Review, 32(4): 926-948. 2001 The Transnational
Villagers, University of California Press, Berkeley.

Liebkind, K. (2001). Acculturation. In R. Brown and S. Gaertner (Eds), Blackwell '
Handbook of Social Psychology: Intergroup Processes, p. 386-406. Oxford, U.K.:
Blackwell.

Lien, Pei-te. (1994). “Ethnicity and political participationzAcomparison between
Asian andMexican Americans.” Political Behavior, 16, 237-264.

Lien, Pei-te. (2001). The Making of Asian America through Political Participation.
Temple University Press.

Lien, Pei-te, M. Margaret Conway, and Janelle Wong. (2003). Understanding Asian
Americans. Routledge.

Lipset, Seymour. (1964). The First New Nation. New York: Basic Books.

Long, Scott. (1997). Regression Models for Categorical and Limited Dependent
Variables. Advanced Quantitative Techniques in the Social Sciences, Volume 7. Sage
Publications.

Lopez, Alfred. (1973). The Puerto Rican Papers. Indianapolis, IN: Bobbs—Merrill.

Lubke G. H. and Muthen B. (2005). “Investigating Population Heterogeneity With
Factor Mixture Models.” Psychological Methods, 10 ( 1): 21-39.

Markstrom-Adams, C. (1992).A consideration of intervening factors in adolescent
identity formation. In G. R. Adams, T. P. Gullota, and R. Montemayor (Eds),
Adolescent identity formation (pp. 173-192). Newbury Park, CA: Sage.

Marschall, Melissa. (2001). “Does the Shoe Fit?: Testing Models of Participation for
African-American and Latino Involvement in Local Politics .” Urban A ﬁairs Review,
37(2): 227-48.

Marschall, Melissa. (2004). “Citizen Participation and the Neighborhood Context: A
New Look at the Coproduction of Local Public Goods.” Political Research Quarterly,
57(June, No. 2): 231-44.

Marquez, Benjamin. (2004). “Latino Identity Politics Research: Problems and Op—
portunities.” Available online at http://perg.tamu.edu/lpc/Marquez.pdf.

Massey, Doug. (2000). The Residential Segregation of Blacks, Hispanics, and Asians:
1970—1990. In Gerald D. Jaynes, ed., Immigration and Race. New Haven: Yale
University Press.

326

Mayo, Y. Q., and Resnick, R. P. (1996). “The Impact of Machismo on Hispanic
women.” Journal of Women and Social Work, 11(3), 257-277.

McCloskey, H. and John Zaller. (1984). The American Ethos. Cambridge, MA:
Ilarvard University Press.

McCutclieon, Allan. (1987). Latent Class Analysis. Sage University Press.

McClain, Paula D. and John A. Garcia. 1993. “Expanding Disciplinary Boundaries:
Black, Latino, and Racial Minority Group Politics in Political Science,” In Ada W.
F inifter, Political Science: The State of the Discipline, 11. Washington, DC: Ameri-
can Political Science Association.

I\~‘Iichelson, M. (2001). “Political Trust Among Chicago Latinos.” Journal of Urban
A ﬂairs, 23(3): 323-334.

Miles, R. (1989). Racism. London: Routledge.

Milligan, S. Hispanic Vote Gains Inﬂuence for 2000 Vote. The Boston Globe,
Septermber, 13 1999.

Miller, A., Patricia Gurin, Gerald Gurin, and Oksana Malanchuk. (1981). “Group
Consciousness and Political Participation.” American Journal of Political Science,
25(3): 494-511.

Miller, Arthur H., and Martin P. Wattenberg. (1983). “Measuring Party Identi-
ﬁcation: Independent or No Partisan Preference?” American Journal of Political
Science, 27 (1)2106-121.

Miller, Warren E., and National Election Studies. (1994). American National Election
Studies Cumulative Data File, 1948-2000. Ann Arbor, MI: University of Michigan,
Center for Political Studies.

I\Iontoya, L.J., (2002). “Expanding Disciplinary Boundaries: Black, Latino, and
Racial Minority Group Politics in Political Science,” In Ada W. F inifter, Political
Science: The State of the Discipline, II. Washington, DC: American Political Science
Association.

Moore, J. and H. Panchon. (1985). Hispanics in the United States. Englewood Cliff,
New Jersey: Prentice-Ilall.

Muoz, Carlos. (1989). Youth, Identity, and Power: The Chicano Movement. New
York: Verso.

327

Nguyen, H., Messe, L., and Sollak, G. (1999). “Toward a more complex understanding
of acculturation and adjustment: Cultural involvements and psychosocial functioning
in Vietnamese youth.” Journal of Cross-Cultural Psychology, 30, p. 5—31.

Olsen, Marvin. ( 1970). Social and Political Participation of Blacks. American Soci-
ological Review, 35: 682-696.

Olzak, Susan (1983). “Contemporary Ethnic Mobilization.” Annual Review of Soci-
ology, 9: 355-374.

Ong, N. and David S. Meyer. (2004). “Protest and Political Incorporation: Viet-
namese American Protests, 1975-2001.” Center for the Study of Democracy. Paper
04-08. htt p: / / repositories.cd1ib.org / csd / 04-08

Pachon, Harry, and Louis DeSipio. (1994). New Americans By Choice: Political
Perspectives of Latino Immigrants. Boulder: Westview Press.

Padilla, Felix M. (1985). Latino Ethnic Consciousness: The Case of Mexican Ameri-
cans and Puerto Ricans in Chicago. Notre Dame: University of Notre Dame Press.

Pantoja, A. (2005). “Transnational Ties and Immigrant Political Incorporation: The
Case of Dominicans in Washington Heights, New York.” International Migration,
43(4), p. 123-146.

Pardo, Mary. (1998). Mexican American Women Activists: Identity and Resistance
in Two Los Angeles Communities. Philadelphia: Temple University Press.

Park, R. E., Burgess, E. W., McKenzie, R. D., and Wirth, L. (1925). The city.
Chicago: University of Chicago Press.

Pearson, Kathryn and Jack Citrin. (2002). “Identity Choice and Language Politics
in the United States,” unpublished paper presented at the 2002 Conference of the
International Society for Political Psychology, Berlin, Germany.

Peppas, S. (2001). “Subcultural Similarities and Differences: An Examination of
Core U.S. Values.” Cross Cultural Management, 8(1): 59-70.

Petrocik, John R. (1974). “An Analysis of Intransitivities in the Index of Party
Identiﬁcation.” Political Methodology, 1:31-47.

Phinney, Jean and Rotheram, M. J. (1986). Children’s Ethnic Socialization: Themes
and Implications. In J.S. Phinney and M.J. Rotheram (Eds) Children’s Ethnic
Socialization: Pluralism and Development (pp. 274-283). Newbury Park: Sage Pub—
lications.

328

Phinney, J. S. (1990). “Ethnic Identity in Adolescents and Adults: Review of Re-
search.” Psychological Bulletin, 108, 499—514.

Phinney, J. S., and Alipuria, L. (1990). “Ethnic identity in college students from four
ethnic groups.” Journal of Adolescence, 13, 171-184.

Phinney, J. S. (1992). “The Multigroup Ethnic Identity IVIeasure: A new scale for

73

use with adolescents and youth adults from diverse groups. Journal of Adolescent

Research, 7, p. 156-176.

Phinney, J. S. (1996). “Understanding Ethnic Diversity: The Role of Ethnic Identity.”
American Behavioral Scientist, 40, p. 143-152.

Phinney, J. S. and Devich-Navarro, M. (1997). “Variation in bicultural identiﬁcation
among African American and Mexican American adolescents.” Journal of Research
on Adolescence, 7, p. 3—32.

Phinney, J. S., Cantu, C., and Kurtz, D. (1997). “Ethnic and American identity as
predictors of self-esteem among African American, Latino, and White adolescents.”
Journal of Youth and Adolescence, 26, 165-185.

Phinney, J. S., Gabriel Horenczyk, Karmela Liebkind, and Paul Vedder. (2001).
“Ethnic Identity, Immigration, and Well—Being: An Interaction Perspective.” Journal
of Social Issues, 57(3): 493-510.

Phinney, J. S., and Juana Flores. (2002). “’Unpacking’ Acculturation: Aspects of
Acculturation as Predictors of Traditional Sex Role Attitudes.” Journal of Cross-
Cultural Psychology, 33(3): 320-331.

Portes , Alejandro, Robert Nash Parker, and Jose A. Cobas. (1980). “Assimilation
or Consciousness: Perceptions of the US. Society Among Recent Latin American
Immigrants to the United States.” Social Forces, 59: 201-224.

Portes, Alejandro and Min Zhou. (1993). “The New Second Generation: Segmented
Assimilation and its Variants.” Annals of American Political and Social Sciences,
voL 530.

Portes, Alejandro and Ruben Rumbaut. (1996). (2nd ed.). Immigrant America.
Berkeley: University of California Press.

Portes, Alejandro and Ruben Rumbaut. (2001). Legacies: The Story of the Immi-
grant Second Generation. Berkeley: University of California Press.

329

 

Ramakrishnan, Karthick and Thomas Espenshade. (2001). “Immigrant Incorpo—
ration and Political Participation in the United States.” International Migration
Review, vol. 36.

Rogers, Reuel. (2000). Afro—Caribbean Immigrants, African Americans, and the
Politics of Group Identity. In Black and Multiracial Politics in America, edited by Y.
M. Alex-Assensoh and L. J . Hanks. New York: New York University Press.

Rosenstone, Steven J ., Roy L. Behr and Edward H. Lazarus. (1984). Third Parties in
America. Princeton: Princeton University Press. Saito, Leland. 1998. Race and P01—
itics. Indiana University Press. Santa Ana, Otto. 2002. Brown Tide Rising. Austin,
TX: University of Texas Press. Schattschneider, BE. 1942. Party government. New
York: Rinehart.

Rotheram-Borus, M. J ., Dopkins, S., Sabate, N ., and Lightfoot, M. (1996). Personal
and ethnic identity, values, and self-esteem among Black and Latino adolescent girls.
In B. J. Ross Leadbeater N. Way (Eds), Urban girls (pp. 35—52). New York: New
York University Press.

Rumbaut, Ruben. (1984). “The Crucible Within: Ethnic Identity, Self-esteem, and
Segmented Assimilation Among Children of Immigrants.” International Migration
Review, 28(4), p. 748- 794.

Sapiro, Virginia. (1981). When Are Interests Interesting? The Problem of Political
Representation of Women, The American Political Science Review, 75(3), p. 701-716.

Sapiro, Virginia. (1983). The Political Integration of Women. Urbana, IL: University
of Illinois Press.

Schildkraut, Deborah J. (2003). “American Identity and Attitudes toward Ofﬁcial-
English Policies.” Political Psychology, 24: 469-99.

Schmid, A. Allan. (2004). Conﬂict and Cooperation: Institutional and Behavioral
Economics. Blackwell Publishers, Malden, MA.

Segura, Gary M., Dennis Falcon, and Harry Pachon. (1996). Dynamics of Latino Par-
tisanship in California: Immigration, Issue Salience, and Their Implications. Harvard
Journal of Hispanic Policy 10:62-80.

Segura. Gary M. (2006). “Immigration and National Identity.” Perspectives on
Politics, 4(June): 276-278.

Shaw, D., de la Garza, R. O.,and Lee, J. (2000). “Examining Latino turnout in 1996:
A three-state, validated survey approach.” American Journal of Political Science,
44(2), 338-346.

330

Shingles, Richard. (1981). “Black Consciousness and Political Participation: The
Missing Link.” American Political Science Review, 75(1): 76-91.

Shively, W. Phillips. (1979). “The Development of Party Identiﬁcation among
Adults.” American Political Science Review, 73, 4 (Dec) 1039-1054.

Simpson, George and J. Milton Yinger. (1972). Racial and Cultural Minorities: An
Analysis of Prejudice and Discrimination. New York: Harper and Row.

Smith, R. (1988). “The ’American Creed’ and American identity: The limits of liberal
citizenship in the United States.” Western Political Quarterly, 41: 225-51.

Smith, R. (1993). “Beyond Tocqueville, Myrdal, and Hartz: The multiple traditions
in America.” American Political Science Review, 87: 549-66.

Smith, R. (1997). Civic Ideals: Conﬂicting Visions of Citizenship in US. History.
New Haven: Yale University Press.

Smith, R. 2003. Stories of Peoplehood: The Politics and Morals of Political Member-
ship. Cambridge: Cambridge University Press.

Smith, R. 2004. “Identities, interests, and the future of political science.” Perspectives
on Politics, 2; 301-312.

Sniderman, P.M., Brody, R., and Tetlock, RE. (1991). Reasoning and choice: explo-
rations in political psychology. New York: Cambridge University Press.

Sniderman, P. M. and Hagen, M. (1985). Race and inequality: a study in American
values. Chatham, NJ: Chatham Hourse.

Soto, E., and Shaver, P. (1982). “Sex-role traditionalism, assertiveness, and symp-
toms of Puerto Rican women living in the United States.” Hispanic Journal of Be-
havioral Sciences, 4(1), 1-19.

Staton, Jeffery K., Damarys Canache and Robert Jackson. (Forthcoming 2007).
“Dual Nationality Among Latinos: What Are the Implications for Political Connect-
edness?” Journal of Politics,

Stokes, Akai. (2003). “Latino Group Consciousness and Political Participation.”
American Politics Research, 31:4 (July):361-378.

Tajfel, Henri, and John C. Turner. (1979). “An Integrative Theory of Intergroup
Conflict.” In William G. Austin and Stephen Worchel, eds, The Social Psychology
of Intergroup Relations. Monterey, CA: Brooks/ Cole.

331

Tajfel, Henri, and John C. Turner. (1979). “An Integrative Theory of Intergroup
Conﬂict.” In William G. Austin and Stephen Worchel, eds, The Social Psychology
of Intergroup Relations. Monterey, CA: Brooks/ Cole.

Tajfel, H. (1981). Human Groups and Social Categories. Cambridge University Press,
Cambridge.

Tate, Katherine. (1993). From Protest to Politics: The New Black Voters in American
Elections. Cambridge: Harvard University Press.

Torres, Andres. (2001). “Latino Cultural Identity and Political Par-
ticipation: Scanning the Recent Literature.” Available online at
ht tp: / / www. gaston. umb.edu / publications / gr / 8xxnl / torrescornell.html.

Torres-Saillant, S. (1989). “Dominicans as a New York community: a social ap-
praisal”, A Journal of Marginal Discourse, 2(1), p. 7-25.

Uhlaner, Carole J. (1989). “Rational Turnout: The Neglected Role of Groups.”
American Journal of Political Science, 33: 390-422.

Uhlaner, Carole J ., and F. Chris Garcia. (1998.) Foundations of Latino Party Iden-
tiﬁcation: Learning, Ethnicity, and Demographic Factors Among Mexicans, Puerto
Ricans, Cubans, and Anglos in the United States. Irvine: Center for the Study of
Democracy Research Monograph Series.

Uhlaner, Carole Jean, Mark M. Gray, and F. Chris Garcia. (2000). Ideology, Issues,
And Partisanship Among Latinos. Presented at the Annual Meeting of the Western
Political Science Association, March 24-26, at San Jose, CA.

Umana-Taylor, Adriana, Marcelo Diversi and Mark A. Fine. (2002). “Ethnic Identity
and Self-Esteem of Latino Adolescents: Distinctions Among the Latino Populations.”
Journal of Adolescent Research, 17(3): 303-327.

US Census Bureau, Population Projection Division, March, 2004.

Verba, Sidney and Norman Nie. (1972). Participation in America: Political Democ-
racy and Social Equality. New York: Harper and Row.

Verba, Sidney, Kay Lehman Schlozman, and Henry Brady. (1995.) Voice and Equal-
ity. Cambridge: Harvard University Press.

Verba, Sidney, Kay Lehman Schlozman, Henry Brady, and Norman H. Nie. (1993).
“Race, Ethnicity, and Political Resources: Participation in the United States.”
British Journal of Political Science, 23(4): 453- 497.

332

Verba, Sidney, Nancy Burns, and Kay Lehman Schlozman. (1997). “Knowing and
Caring about Politics: Gender and Political Engagement.” The Journal of Politics,
59(4): 1051- 1072.

Wagenheim, Kal (ed). (1973). Puerto Ricans: A Documentary History. Garden
City, NY: Doubleday, Anchor.

Warner, W. Lloyd, and Leo Srole (1945). The Social Systems of American Ethnic
Groups (New Haven, CT: Yale University Press).

Waters, Mary. (1994). “Ethnic and Racial Identities of Second-Generation Black
Immigrants in New York City.” International Migration Review, 28 (Winter): 795-
820.

Waters, Mary. (1997). Immigrant Families at Risk: Factors that Undermine Chances
for Success. Alan Booth, Ann C. Crouter, and Nancy Landale (eds) Immigration
and the Family: Research and Policy on US. Immigrants. Mahwah, NJ: Lawrence
Erlbaum associates, Inc., Publishers. pp. 79—87.

Wattenberg, Martin P. (1990). The Decline of American Political Parties: 1952-1988.
Cambridge: Harvard University Press.

Wilcox, Clyde and Gomez, Leopoldo. (1990). “Religion, Group Identiﬁcation and
Politics Among American Blacks.” Sociological Analysis, 51: 271-287. 1990.

W’ong, Janelle S. (2000). “The Effects of Age and Political Exposure on the Devel-
opment of Party Identiﬁcation Among Asian Amerian and Latino Immigrants in the
United States.” Political Behavior, 22 (4):341-371.

Wong, Janelle. (2001). “The New Dynamics of Immigrants’ Political Incorporation.”
Doctoral dissertation. Yale University.

Wong, Janelle, Pei-ti Lien, and M. Margaret Conway. (2005). “Group-Based Re-

3,

sources and the Political Participation of Asian Americans. American Politics Re-

search, 33: 545-76.

Xie, Yu and Emily Greenman. (2005). “Segmented Assimilation Theory: A Refor-
mulation and Empirical Test.” PSC Research Report No. 05-581. August 2005.

Yang, Phillip. (1994). “Explaining Immigrant Naturalization.” International Migra-
tion Review, 28(3): 449-477.

Zaller, John. (1992). The Nature and Origins of Mass Opinion. Cambridge, UK:
Cambridge University Press.

333

 

Zhou, Min and Carl L. Bankston. 1998. Growing Up American: How Vietnamese
Children Adapt to Life in the United States. New York: Russell Sage Foundation.

Zuo, J ., and Tang, S. (2000). “Breadwinner Status and Gender Ideologies of Men
and Women Regarding Family Roles.” Sociological Perspectives, 43(1), 29—43.

334

IlﬂlljjlﬂjijijjljjlIljjjlill

 

i . _‘_—-