Article in International Migration Review

Immigrunt Incolporution und Politicul
Purticipution in the United States'
S. Karthick Ramakrishnan
Princeton University
Thomas J. Espenshade
Princeton University
This article examines several factors related to immigrant incorporation
that have been ignored in previous studies of voting participation. We
add various immigrant-related variables to a model that controls for individual resources, social incorporation, institutional barriers and contexts
of political mobilization. We find little support for straight-line assimilationist theories of immigrant adaptation. We also find that coming from
a repressive regime has no significant effect on voting and that living in
areas with Spanish-language ballots does not increase the likelihood of
voting among first generation Latinos. Our results also suggest that antiimmigrant legislation has a positive effect on participation among first
and second generation immigrants. Overall, the immigrant-related variables introduced in our analysis add significantly to the existing theoretical knowledge on voting participation in the United States.
In the past decade, there has been a substantial amount of research on the
demographic, economic and social incorporation of immigrants in the United States. In the demographic arena, researchers have investigated how
migrants' patterns of fertility, health and mortality, and settlement compare
with those of natives (Kahn, 1994; Guendelman et al., 1990; Frey, 1996;
White and Omer, 1997).
Studies of the economic and social incorporation of immigrants have
analyzed immigrants' earnings and employment opportunities in relation to
those of the native-born and how immigrants' education, occupational
mobility, and English-language ability change with the passage of time in the
'An earlier version of this paper was presented at the annual meeting of the Population Association of America, Los Angeles, CA, March 23-25, 2000. We thank the anonymous reviewers for their valuable suggestions. We also thank R. Douglas Arnold, Larry Bartels, Jennifer
Hochschild, Jane Junn, Tali Mendelberg, Debbie Schildkraut and Min Zhou for their feedback and criticisms. Finally, we thank Sipra Roy for her help with creating the graphs and
tables for this paper, and the Office of Population Research Center Grant #P30HD32030 for
the use of facilities at the Office of Population Research at Princeton University.
0 2001 by the Center for Migration Studies of New York. All rights reserved.
0 198-9183/01/3503.0 135
870
ZMR Volume 35 Number 3 (Fall 2001): 870-909
IMMIGRANT INC@KI’ORATI@NAND POLITICAL PARTICIPATIoN IN THE
u. s.
87 1
United States (Borjas, 1999; Stevens, 1992; Portes and MacLeod, 1999;
Espenshade and Fu, 1997; Lbpez, 1999).
Studies of political incorporation have focused primarily on the determinants of naturalization and differences in citizenship acquisition across different nationalities (Portes and MOZO,1985; Liang, 1994; Yang, 1994).
Scholars of immigrant incorporation have paid considerably less attention to
voting and other forms of political participation occur after naturalization.
Although first and second generation constitute nearly 15 percent of eligible
voters in the United States, there have been fay systematic studies of voting
participation across immigrant generations. Many of the studies have had a
limited scope, applying either to particular ethnic populations or to particular localities. Examples of early works on generational differences in participation include articles by Lamar Kellstedt (1974) and James Lamare (1982).
Kellstedt’s study of Buffalo, New York showed that second generation
respondents have higher rates of voting participation than either immigrants
or nonethnic first or higher generation respondents. Lamare noted in his
study of Chicano children in El Paso, Texas that third generation respondents
had the highest levels of psychological identification and affect towards the
political system.2 More recently, there have been a few cross-ethnic studies of
immigrant political incorporation in particular states such as California and
Texas (Uhlaner etal., 1989; Cho, 1999; Junn, 1999) or in metropolitan areas
such as New York and Los Angeles (Mollenkopf et al., 1999; Wong, 2000).
In their study of political participation in California, Uhlaner et al., note that
voting among immigrants is influenced by the duration of their stay in the
United States, as well as by their ability to speak English. Cho (1999) also
finds such factors to be important and suggests that immigrants educated
abroad are less likely to vote than those educated in the United States. Finally, Junn (1999) concludes that immigrants may be less likely to participate in
“system-directed activities such as voting, but they are just as likely as the
native born to participate in “direct” political activities such as protests. While
these studies all offer important insights regarding immigrant political incorporation, they are limited by the fact that they examine only particular states
or metropolitan areas.
There are a few national-level studies of generational status and voting
participation, but these, too, have had their limitations. During the past
decade, there have been three national-level studies on generational status and
voting participation, two of which have been restricted to Latino subpopula2Larnare refers to children whose parents were born in the United States as ‘second generation.’
872
INTERNATIONAL MIGRATION
REVIEW
tions (Bass and Casper, 1999; DeSipio, 1996b; also see Hill and Moreno,
1996).3 None of these recent studies explicitly considers generational status
in their analyses, as they focus solely on differences between native-born and
foreign-born citizens. As we shall see, separating out the second generation
offers some important insights on processes of political adaptation among
immigrants. In addition, these studies are largely demographic in nature,
paying little attention to how political contexts affect individuals’ propensities to vote. For example, this research fails to account for institutional barriers to registering and voting, as well as the intensity of electoral competition.
Our goal in this paper is to build a more complete model of voting participation, drawing on some of the recent theoretical insights on the effects of
political mobilization and macropolitical contexts on electoral participation
(Rosenstone and Hansen, 1993; Shachar and Nalebuff, 1999). After reviewing the standard theoretical literature on voting participation, we introduce
several factors related to immigrant incorporation that have not previously
been considered in studies of voting participation. An analysis of voting participation is conducted using data from three recent Current Population Survey Voter Supplements (CPS). Although we incorporate such factors as individual resources and contexts of political mobilization, our empirical analysis
emphasizes factors related to immigrant incorporation. Finally, the theoretical implications of the results of our model are discussed, especially as they
relate to assimilationist accounts of immigrant adaptation and the impact of
immigrant legislation on political participation.
TRADITIONAL MODELS OF VOTING PARTICIPATION
Theories of electoral turnout have tended to emphasize one or more of the
following sets of factors: 1) demographic characteristics and individual
resources related to socioeconomic status, 2) incorporation into social networks, 3) institutional barriers to registering and voting, 4 ) strategic mobilization by political actors, and 5 ) attitudinal factors such as partisanship,
political interest and political efficacy.4 Our theoretical model incorporates
the first four sets of factors. Factors related to individual attitudes and orien3Although Hill and Moreno are interested in the question of whether second-generation
Cubans are more likely to participate in politics, they do not explicitly include generational
status in their multivariate model. Instead, they use a continuous variable that measures the
percent of one’s life the respondent has spent in the United States.
*In the standard theoretical literature on voting participation, “external efficacy” refers to the
extent to which respondents believe that their participation matters and “internal effrcacy”
refers to the extent to which respondents feel that they understand government and politics.
IMMIGRANTINCORPORATION
AND POLITICAL
PARTICIPATION
IN THE U. S.
873
tations are set aside because the CPS does not provide any meaningful measures for such variables. The issue of individual-level data on political attitudes is addressed later in the discussion of data sources.
Individual Resources and Demographic Characteristics
Theorists have devoted considerable attention to individual resources such as
education and income on electoral participation. Studies both of the Current
Population Survey and the National Election Study (NES) have shown that
educational attainment has the strongest effects on voting, serving both as an
indicator of political skills and the likelihood of being mobilized (Wolfinger
and Rosenstone, 1980; Rosenstone and Hansen, 1993). Income has also
been shown to be a significant predictor of voting participation, although its
effects are generally weaker than those of education.
In addition to resources, theorists have also drawn attention to the
importance of individual demographic characteristics such as age and race.
Age has a positive, curvilinear relationship to voter turnout: older citizens
tend to have higher levels of political knowledge and partisan identification
(Rosenstone and Hansen, 1993). Racial gaps in voting have been shown to
be significant for Latinos and Asian-Americans, but not for African-Americans. Since the passage of the National Voting Rights Act in 1965, racial gaps
in registration and voting between blacks and whites have narrowed considerably. Even as early as 1980, scholars began to note that, after controlling
for other demographic factors, whites did not have a significant advantage in
participation over blacks (Wolfinger and Rosenstone, 1980). By contrast,
studies of Latino political participation have shown a persistent gap in voting
participation between Latinos and non-Latinos (Calvo and Rosenstone,
1984; Hero and Campbell, 1996). There have been few national studies of
voting participation among Asian-Americans; most studies have focused on
particular states such as California or particular metropolitan areas such as
Los Angeles (Lien, 1997; Uhlaner et a l , 1989; Nakanishi, 1986). However,
one recent analysis of CPS data reveals that Asian-Americans were the group
least likely to vote in the 1992 general election (Lien, 1998).
Social Incorporation
In addition to individual-level factors, theorists have also paid attention to the
incorporation of individuals into communities or social contexts that encourage or discourage political participation. Chief among the social-contextual
variables mentioned are employment status, residential stability and marital
874
INTERNATIONAL
MIGRATION
REVIEW
status. The unemployed are less likely to participate in politics, not only
because they tend to have lower incomes, but also because they do not participate in social networks in the workplace that reward political participation
(Wolfinger and Rosenstone, 1980; Rosenstone and Hansen, 1993). People
with low levels of residential stability are less likely to participate because they
are less likely to be vested in their communities and are less likely to have stable networks of friends and neighbors. These networks are considered to be
important in fostering political participation because they subsidize the costs
of obtaining political information. Furthermore, peer influences among
spouses and social networks are believed to foster participation by applying
selective sanctions for noncompliance (Rosenstone and Hansen, 1993).
While peer influences can indeed operate in the opposite direction (towards
political apathy and abstention from voting), past evidence from both the
CPS and the NES indicate otherwise; those who are more embedded in social
networks are indeed more likely to vote (Wolfinger and Rosenstone, 1980;
Rosenstone and Hansen, 1993).
Institutional Barriers
While an individual’s resources and degree of “social connectedness” have a
strong influence on his or her likelihood to vote, macropolitical contexts also
matter. Institutional barriers to voting can have a substantial impact on voter
turnout, as witnessed by low levels of black turnout during the Jim Crow era
of poll taxes and literacy tests. After the elimination of such restrictions during the 1960s, the most significant barriers that remain are state-level rules on
registration and absentee voting. States with restrictive eligibility for absentee voting decrease the probability of voting among registered citizens (Oliver, 1996). Similarly, citizens who live in states with early closing dates for registration are less likely to register in time for elections.
Political Mobilization
The level of political mobilization in a given state or congressional district can
have a significant influence on voter turnout. From the early works of V.O.
Key (1949) and E. E. Schattschneider (1960),5 political scientists have noted
the importance of political mobilization and parry competition in inducing
5In his comparative study of Southern politics, Key (1747) noted that states with higher partisan or factional competition had higher levels of voter turnout. Indeed, Key attributed the
low turnout among poor Southern whites, not as much to poll taxes or literacy tests, but to
the lack of two-party competition in the South.
IMMIGRANT INCORPORATION AND POLITICAL PARTICIPATION IN THE
u. s.
875
higher levels of voter turnout. Recent studies of electoral turnout have revived
concerns about party competition and political mobilization in general
(Rosenstone and Hansen, 1993; Schachar and Nalebuff, 1999). They show
that individuals who live in states where presidential contests are close are
more likely to vote than those living in states where there is a clear favorite.
This is due largely to differences in state-level mobilization by political actors.
As Rosenstone and Hansen put it, “campaigns, interest groups, and the
media.. .contest every inch in campaigns that stand to be decided by tenths
of percentage points, and they tacitly concede campaigns that look to be
blowouts” ( 179).
Political mobilization also varies by the number of major electoral contests
in a given election. States that have a governor‘s race or a senate race concurrent
with presidential or house elections have higher levels of political mobilization
and voter turnout. The effects can be especially strong during midterm elections, when there is no presidential contest to generate interest among potential voters (Rosenstone and Hansen: 184). Finally, the historical legacy of mobilization and participation in a particular state or region can have a significant
influence on individuals’ propensities to register and vote. Indeed, political scientists in the 1970s and 1980s noted a kind of regional “political culture” in the
South, a vestige of one-party competition and racial disenfranchisement that
led to lower turnout among Southern blacks and whites.
I M M r ~ I ~N C
TO~O~TIU
- AN
N EXRANLIED MODEL OF
VOTING PARTICIPATION
When considering political participation, as well as other processes of immigrant adaptation, it is useful to think of such processes as having at least two
temporal dimensions - across time within a particular immigrant generation
and across generations. For example, researchers have typically found that
earnings and English-language ability improve the longer immigrants live in
the United States. At the same time, there are also differences in language
acquisition across generations. Foreign-born individuals frequently do not
speak English as well as their native-born children, and second generation
immigrants often retained some understanding of their parents’ mother
tongue. Those in the third generation or higher typically lose all proficiency
in the original migrants’ language unless they make some special effort
to regain it (Fishman, 1972; Lbpez, 1999).This study incorporates both these
temporal dimensions and other new features as predictors of political
participation.
876
INTERNATIONAL MIGRATION
REVIEW
Generational Status
Traditional theories of immigrant adaptation considered assimilation to be a
unilinear process, whereby the economic and social conditions of individuals
and ethnic groups improve over each succeeding generation. Over the past
decade, several studies have challenged the applicability of “straight-line’’theories of immigrant adaptation to the “new” second generation (Gans, 1992;
Portes and Zhou, 1993; Rumbaut, 1997; Zhou, 1997). Many of the revisionist perspectives present a model of “segmented assimilation,” in which
different group characteristics and modes of incorporation lead to diverse
outcomes that can include second generation decline. In order to test the
applicability of the straight-line and segmented assimilation models of immigrant incorporation, we analyze generational patterns in participation across
different racial/ethnic groups. We expect to find different generational patterns across racial groups because of stark differences in the historical reception given to immigrants from Europe versus those from Asia and Latin
America (Espiritu, 1992). While urban machines and mobilization under the
New Deal ensured the political incorporation of immigrants from Europe,
exclusionary laws and discriminatory practices prevented many Asian and
Mexican immigrants from exercising their political voice until the 1970s (de
la Garza, 1996; Espiritu, 1992).6
Within each raciallethnic group, we expect first generation citizens to be
the least likely to vote because of linguistic and cultural barriers that make it
more difficult for them to obtain political information or vote in areas with
English-only ballots. First generation citizens are also more likely than those
in higher generations to retain ties to political institutions in the home country. O n the other hand, there may indeed be selection effects whereby those
who choose to naturalize have unusually high levels of commitment to political participation or high levels of political awareness after preparing for the
citizenship exam. Even with these selection effects, however, we still expect
participation to be lowest among first generation respondents because of the
overriding importance of linguistic barriers, cultural barriers and immigrant
political orientations.
Duration of Stay in the United States
For the foreign-born, duration of stay in the United States can be an important indicator of the extent to which immigrants are incorporated into
%deed, most Asian immigrants were not able to naturalize until 1952.
IMMIGRANT
INCORPORATION
AND POLITICAL
PARTICIPATION
IN THE U. S.
877
domestic political institutions. Based on previous findings among Latino and
Asian-American populations (Uhlaner et al., 1989, Hill and Moreno, 1996),
we expect longer residence in the United States to lead to greater political participation among first generation citizens. Just as immigrants who have lived
in the United States for a longer period of time are more fluent in English,
they also tend to have greater contact with, and stronger commitments to, the
mainstream political system. For example, Cain, Kiewiet and Uhlaner (1991)
note in their study of California residents that partisan identification among
Latinos, Chinese and Korean immigrants becomes stronger the longer they
stay in the United States.
Political Socialization in the Home Country
For naturalized citizens, prior experience with repressive or democratic
regimes (hereafter referred to as “prior political experience”) can have a significant impact on the propensity to vote. Those fleeing regimes with long
histories of political repression may be mistrustful of the political system and
therefore be less likely to vote in elections. O n the other hand, they may relish the freedom of choosing among competing candidates for political office
and therefore be more likely to vote. So far, only comparative studies of
Cuban and Mexican immigrants have examined whether prior political experience has a significant impact on subsequent participation. Based on
precinct-level data from Miami, Portes and Mozo (1985) argue that CubanAmericans have higher levels of turnout than other Hispanic subgroups.
Multivariate analyses of the Latino National Political Survey add further credence to the prior political experience hypothesis - Cuban immigrants are
more likely to vote than Mexican immigrants, even after controlling for age,
education, income and length of stay in the United States (Arvizu and Garcia, 1996). However, given the small number of countries of origin in such
studies (Cuba and Mexico), the effect of prior political experience on voting
participation has not been systematically tested. It remains to be seen whether
the voting participation of Cuban Americans is consistent with that of immigrants from other communist or repressive regimes.
Finally, just as prior political experience may influence voting participation among foreign-born citizens, it may also influence participation among
children of the first generation. Studies of political socialization have shown
that childhood exposure to political activities or discussions by parents has a
significant impact on subsequent adult participation (Jennings and Niemi,
1981; Verba, Schlozman and Brady, 1995). Thus, if an immigrant family’s
878
INTERNATIONAL
MIGRATION
REVIEW
“prior political experience” does indeed have a significant impact on political
participation in the United States, we would expect the impact to be present
among the second generation as well, albeit perhaps to a lesser degree.
Linguistic Barriers to Participation
The traditional literature on barriers to voting participation has focused the
effects of absentee ballot restrictions, registration requirements and racial gerrymandering on voter turnout. When considering the political incorporation
of immigrants, linguistic barriers to participation can play just as important
a role as some of the other barriers to participation. The presence or absence
of multilingual ballots should have their greatest impact on individuals with
the lowest levels of English proficiency. Thus, for example, living in a Spanish-ballot area versus an English-only-ballot area should increase turnout
among Latino immigrants, and the difference should be greater for the first
generation than for higher generation Latinos.
Ethnic Residential Concentration
Those who live in states or metropolitan areas with high concentrations of coethnics are more likely to have contacts with, or exposure to, ethnic media and
community organizations. Having higher concentrations of co-ethnics therefore lowers the per-capita cost of ethnic mobilization by interest groups, party
organizations and candidate organizations. At the same time, previous studies
of Latino participation have shown that areas of high ethnic concentrations can
also be areas of lower mobilization and participation (DeSipio, 1996a; de la
Garza, 1996). This may be due to the effect of other social contextual factors
associated with high Latino and immigrant concentrations: higher residential
poverty, greater proportions of noncitizens, and peers with low English proficiency (Cho, 1999; Espenshade and Fu, 1997). If proximity to co-ethnics does
indeed lead to higher electoral mobilization, we would expect such mobilization
to have a stronger effect on the voting participation of those in the second generation or higher. While first generation citizens may have deeper ties to their
co-ethnic communities, such ties may not lead to greater participation in the
United States because first generation ethnic organizations tend to orient themselves more towards homeland politics than U.S. politics (Jones-Correa, 1998).
Mobilization over Anti-Immigrant Legishtion
While mobilization among residential co-ethnics may have their strongest
effects among those in higher generations, we expect anti-immigrant legislation
IMMIGRANT
INCORPORATION
AND POLITICAL
PARTICIPATION
I N THE U. S.
879
to have a distinct mobilizing effect on first and second generation immigrants.
Those in the first or second generation are more likely to mobilize against laws
that threaten to cut off public benefits to immigrants because they are more likely to have relatives that stand to lose those benefits. Media reports of the 1996
election indicate that immigrant legislation in 1996 was indeed more salient for
immigrants than for the rest of the population (Glastris et al, 1997; McDonnell and Ramos, 1996). This article assesses the impact of anti-immigration legislation on turnout by examining whether mobilization over Proposition 187 in
California led to a disproportionate increase in voting participation among first
and second generation immigrants.
To summarize, traditional theories of political participation have examined the effects of individual resources, social incorporation and political
mobilization on voter turnout. In this study, we introduce additional variables related to immigrant incorporation that may influence voting participation. Some of these variables have been included in previous studies of
political participation that have been limited to particular ethnic groups or to
particular regions. We examine the effects of these new immigrantrelated variables across racial/ethnic groups by using national data on voting
by participation.
DATA AND METHODS
This analysis relies on the Current Population Survey November Voter Supplements (CPS) from 1994, 1996 and 1998. The CPS offers several advantages in the study of voting participation across immigrant generations. First,
it is the only national survey since 1994 that includes questions on voting as
well as questions on the nativity of respondents and their parents. Similar
questions have been absent from the National Black Election Study and from
the National Election Study (NES) since 1994. As for Latinos, a 1996 survey by the Tomis Rivera Policy Institute (TRPI) includes questions on the
generational status of respondents, but the sample is limited to three states
(California, Florida and Texas). The last national survey of Latinos on voting
participation that included generational status was the Latino National Political Survey in 1990. The CPS, then, remains the only source of information
on voting that has consistently asked questions on the generational status of
respondents in recent elections.
There are several other advantages to using the CPS to analyze immigrant
voting participation. One is the size of the sample for each election year,
which yields large subsamples of immigrant generations across raciaUethnic
880
INTERNATIONAL
MIGRATION
REVIEW
groups. In 1998, for instance, the sample of first generation citizens induded
1,817 whites, 263 blacks, 1,051 Asian-Americans and 1,046 Latinos.’ Another advantage to the CPS is the richness of the data on immigrant characteristics. For instance, the NES contains information on whether respondents and
their parents were born in the United States or abroad, but it does not include
country-of-origin information for first and second generation immigrants. The
lack of country-of-origin data prevents us from examining the effect of repressive-regime origins on voting participation. Furthermore, the national data in
the CPS has greater variation than the three-state T W I survey on factors such
as state concentration of Latinos and access to Spanish-language ballots. Finally, neither the NES nor the TWI datasets include information such as the duration of stay in the United States for first generation citizens.
At the same time, there are two limitations to utilizing Current Population
Survey data to analyze voting participation. First, the CPS does not include any
individual-level information on political attitudes or party contact that have
been known to lead to greater participation. It has been well documented that
individual contact with party organizations, strong party identification, and
feelings of political efficacy increase the likelihood that a person will vote
(Rosenstone and Hansen, 1993; Miller and Shanks, 1996; Verba, Schlozman
and Brady, 1995). Another limitation is that the CPS does not include validated measures of voting that confirm whether or not an individual’s reported
vote matches public records of registration and voting. Although misreporting
in the CPS is considerably lower than misreporting in other surveys such as the
National Election Study (Presser etal, 1990), past studies have shown that levels of misreporting among blacks and Latinos are higher than those among
whites (Abramson and Claggett, 1992; Shaw et aL, 1999).8
We address the first limitation by including contextual measures of political mobilization and state legacies of voting participation, but we cannot find
a similar solution for other attitudinal factors such as political efficacy and
’These figures are based on valid responses among adult U.S. citizens to the registration and
voting questions in the CPS Voter Supplement. Those who identify as “American Indian,
Aleut, Eskimo” are dropped from the analysis because their generational subsamples are too
small to provide reliable estimates.
8There is some disagreement as to whether the higher level of mismatch between reported and
validated votes among blacks is due to the higher propensity of black to rnisreport or to the
fact that blacks are more likely to live in areas with poor record-keeping of registrations and
votes. Abramson and Claggett ( 1 992) do not find any substantial differences in the quality of
record-keeping by race, based on self-reports by voter registrars. Presser et al. (1990),based on
requests for information by investigators, find that blacks in the NES are twice as likely as
whites to come from areas with poor record-keeping.
IMMIGRANT lNCORPORATION AND POLITICAL PARTICIPATION IN THE
u. s.
881
party identification.’ Next, in order to address the problem of vote misreporting, we examined validated data in the TRPI and NES to determine whether
there are any generational differences in the propensity to misreport the vote.
There were no significant generational differences in vote misreporting among
Latinos in the TRPI survey. We used the NES to examine vote misreporting
only among white respondents because the size of the black, Asian- American
and Latino subsamples were too small for our analysis. Second generation
whites in the NES were less likely to misreport their votes than those in other
immigrant generations. As we shall see, if a similar pattern of misreporting
occurs for whites in the CPS as it does in the NES, then the true generational
differences in voting participation would be even larger than the difference
found in our analysis. O n balance, then, we consider the Current Population
Survey to be a valuable dataset with which to examine the various questions that
we have raised related to immigrant incorporation.
Our sample for the CPS is limited to U.S. citizens who are at least 18
years old. We designate those born outside the United States as “first generation,” those born in the United States with at least one foreign-born parent
as “second generation,” and those born in the United States with U.S.-born
parents as ‘3+
generation.” We fit a logistic regression model to the data
because the dependent variable - voted or did not vote in the November elections - is dichotomous. Our independent variables include not only individual-level data from the CPS, but also several contextual variables from
other data sources. We obtained information on registration requirements
from the League of Women Voters and data on absentee ballot requirements
from a recent study ofvoting participation (Oliver, 1996). We also use information on the closeness of senatorial, gubernatorial and presidential races
from pre-election surveys in Congressional Quarterly Weekly Reports and
Associated Press Research (Cranford, 1994; Babson and Groppe, 1994;
Greenblatt, 1996; Greenblatt and Wells, 1996; Yoachum, 1996; Cassata,
1998; Foerstel, 1998). Finally, we obtained aggregate state-level data on voter
turnout from the Federal Election Commission. Instead of using the dummy
variable “South” to capture the effect of historical legacies of voting participation on contemporary turnout, we use an average of voter turnout from
1972-1992 for each state. We find this measure of state political culture to
9Admittedly, contextual measures of electoral mobilization have a much weaker effect on individual voting than individual-level measures of party contact. Part of the reason is the endogeneity between individual-level contact and participation. Parties are more likely to contact
individuals who have participated in the past, or who are pre-disposed to participate because
of their socioeconomic status.
882
INTERNATIONAL
MIGRATION
REVIEW
be superior for two reasons. First, it looks at contemporary political culture,
taking into account changes in registration and turnout in the South after the
Civil Rights era and the Voting Rights Act of 1965. Furthermore, the measure allows for state-level variation and does not limit the effect of “political
culture” to one regional bloc - namely the Confederate South. We provide a
more complete description of the variables and their sources in the Appendix.
EMPIRICAL RESULTS
There are considerable differences in voting participation, both within the first
generation as well as across immigrant generations (Figure 1). Looking first at
the tabulated data on voter turnout within the first generation, we see a progressive increase in voting participation as immigrants spend more of their lives
in the United States. This pattern appears for all raciallethnic groups. Across
immigrant generations, however, there are considerable differences in voter
turnout for members of different racial groups. Among whites, we do not find
a linear increase in participation from the first generation to higher generations.
Instead, we see a pattern of “second generation advantage” - a progressive
increase in participation within the first generation by duration of stay in the
United States, a peak in participation among the second generation, followed
by a decline among “3+generation” respondents.
Among blacks, we do not find a second generation advantage. Voting
among blacks is highest among those first generation respondents who have
lived in the United States for 20 years or more. Similar results hold true for
Latinos, while for Asian- Americans turnout is highest among those in the third
generation or higher. The bivariate results for generational status and voter
turnout therefore indicate that the second generation advantage applies only to
whites. Among Asian-Americans, a more straight-line pattern seems to hold,
where there is a progressive increase in participation by immigrant generation.
Blacks and Latinos experience a similar monotonic increase in participation,
except that the highest level of turnout is among long-term immigrant citizens.
Since there are many factors related to individual resources, social incorporation, institutional barriers and political mobilization that may account
for differences in turnout, we need to re-examine the issue of generational differences in a multivariate model that controls for such factors. In our multivariate analyses, we fit our model to the entire sample of adult citizens and
then apply the same model to each immigrant generation separately. Many
of the traditional variables mentioned in the background literature are significant across all immigrant generations (see Appendix Tables A-1 to A-3).
IMMIGRANTINCORPORATIONAND POLITICAL
PARTICIPATION
I N THE U. S.
Figure 1
883
Generational Differences in Voting Participation, Midterm
Elections 1994-1998'
Source: Current Population Survey, November 1994, 1998.
Note: 'Tabulated results are only for the midterm elections of 1994 and 1998 because the general level of turnout in
presidential elections is much higher. Generational differences in turnout are still present among all racial grnups in
1996, although the differences are smaller for all groups other than non-Hispanic whites.
Regardless of generation, the likelihood of voting increases with age and
reaches a peak among 70-79 year olds before decreasing among the very old.
Education also has similar effects across all immigrant generations, with voting highest among college graduates and those with graduate degrees.10 Higher income consistently leads to greater voting participation, except for first
generation respondents for whom the effect is weaker and varies from year to
year. Among factors related to social incorporation, marital status and residential stability show a robust relationship to voting participation across
immigrant generations. Married individuals are consistently more likely to
vote than those who are not married, while those who have lived three or more
'OWhile the findings for age and education are in line with the standard accounts of voting
participation, they run contrary to the finding by Cho (1999) that older Asian and Latino
immigrants are less likely to vote than younger ones, and that higher education decreases the
likelihood of participation among Asian immigrants. Even when we restrict the sample to only
Asians and Latinos, the traditional patterns hold - age and education increase the likelihood
of participation. It is possible that the difference is due to the fact that the dataset that Cho
utilizes is from the mid 1980s and is limited to California.
884
INTERNATIONAL MIGRATION
&VIEW
years in a given address are more likely to vote than recent movers. Employment status has an inconsistent relationship to voting across generations and
over time. The coefficients for the variable display inconsistent signs and are frequently statistically insignificant. Finally, contextual variables such as the closeness of presidential races and eligibility for absentee ballots show weak and
inconsistent relationships to voting. O n the other hand, restrictive registration
requirements are consistently associated with lower voting participation and the
closeness of senate/governor races with higher participation in midterm elections. In sum, our empirical results confirm findings from previous studies that
age, education, marital status and residential stability all have a significant effect
on voting participation that is consistent across immigrant generations. Contextual factors such as registration requirements and the closeness of senate/governor races also have a significant impact on voting participation. O n the other
hand, factors such as employment status, absentee voting eligibility and the
closeness of presidential races do not have a significant impact on voter turnout.
Even after controlling for traditional factors related to voting participation,
many of the new variables of immigrant incorporation prove to be significant
in all three election years (see Figures 2-4 below). These new variables are also
jointly significant. The null hypothesis that the newly-included variables are
jointly insignificant was rejected based on chi-squared likelihood-ratio tests in
both the full samples and in the generational subsamples.11 For the rest of our
discussion, we shall focus on the new set of variables introduced in the analysis.
We do so primarily because many of these variables have not received systematic attention in studies of voting participation and because they shed new light
on the political incorporation of different immigrant generations.
IMPACTS OF THE NEWEXPLANATORY VARLABLES
Generational Patterns across RuciaL/Ethnic Groups
Even after controlling for individual resources, social incorporation and contexts of political mobilization, generational differences in the likelihood of
voting remain significant for all raciallethnic groups. Looking across immigrant generations, the full model reveals that the “second generation advanllWe reject the null hypothesis by using the likelihood-ratio test. For example, in 1996 the
likelihood-ratio statistic, -2*(log likelihood of traditional model - log likelihood of full
model), is 836.35. This is well beyond the .01 chi-squared critical value of 49.59 for 29
degrees of freedom. The likelihood-ratio statistics for the generational sub-samples in 1996 are
164.64 for the first generation (.01 chi-squared=41.63 for 23 d.f.), 85.20 for the second generation (.01 chi-squared=29.14 for 14 d.f.) and 521.80 for the 3+ generation (.01 chisquared= 27.69 for 13 d.E).
IMMIGRANT INCORPORATION AND POLITICAL PARTICIPATION IN THE
Figure 2
u.s.
885
Immigrant Incorporation and Voting Participation in 1994
Logistic Regression Estimates of Selected Variables in Full Sample Model
White,
1st t e n 4 O y s
1stGen 10-19ys
1st t e n 19ys
Second Generation
Black,
1st Gen 4 0 ys
1st Gen 10-19ys
1st Gen 19ys
Second Generation
Third and Higher Gen
Asian,
1st Gen c10 ys
l s t t e n 10-19ys
1s t Gen 19 ys
Second Generation
Third and Higher Gen
Latino,
1st Gen c l o y s
1st Gen 10-19ys
1st Gen 19 ys
Second Generation
Third and Higher Gen
From Repressive Regime
Parent from Repressive Regime
In Spanish Ballot Area
Latino x in Spanish Ballot Area
State % Black
Black x State K Black
State % Asian
Asian x State K Asian
State K Latino
Latino x State K Latino
State Voting History
California
-2.00
-1.50
-1.00
-0.50
0.00
0.50
1.oo
COEFFICIENT
‘p<O.l. **p<0.05. +**p<O.Ol
For full model. see Table A-2 in the Appendix.
tage” still holds among whites. Depending on the election year, second generation whites are 22 percent to 34 percent more likely to vote than those in
higher generations and 30 percent to 40 percent more likely to vote than
long-term immigrant residents.’* The fact that second generation whites are
12A word about terminolog. If p is the probability of voting, then pl(1-p) represents
the “odds” of voting. If f3 is a logistic regression coefficient, then the exponentiated form
[exp(P)-1]x100equals the percentage change in the odds (or, loosely, the “likelihood) ofvoting when the associated explanatory variable is increased by one unit. For example, 0.2217 is
the coefficient on “White, Second Generation” in the full model for 1994. Here the odds (or
likelihood) ofvoting are expected to be [exp(0.2217)-1]~100=25percent higher than the odds
for whites in the 3+ generation.
886
INTERNATIONAL
MIGRATION
REVIEW
Immigrant Incorporation and Voting Participation in 1996
Logistic Regression Estimates of Selected Variables in Full Sample Model
Figure 3
White,
1st Gen < l o ys
1st Gen 10-19ys
1st Gen 19ys
Second Generation
Black,
1st Gen 4 0 y s
1stGen 10-19ys
1st Gen 19 ys
Second Generation
Third and Higher Gen
Asian,
IstGen<lOys
1st Gen 10-19ys
1st Gen 19 ys
Second Generation
Third and Higher Gen
Latino,
1 s t Gen < I 0 ys
1st Gen 10-19ys
1st Gen 19 ys
Second Generation
Third and Higher Gen
From Repressive Regime
Parent from Repressive Regime
In Spanish Ballot Area
Latino x in Spanish Ballot Area
State % Black
Black x State 70Black
State % Asian
Asian x State % Asian
State
70Latino
Latino x State % Latino
State Voting History
California
-2.00
-1.50
-1 .oo
-0.50
0.00
0 50
1.oo
COEFFICIENT
*p<O.l, **pc0.05, ***p<O.Ol
For full model, see Table A-2 in the Appendix.
less likely to misreport the vote in the NES than whites in other immigrant
generations does not diminish the significance of the second generation
advantage. Indeed, if a similar generational pattern of misreporting occurred
in the CPS, the magnitude of the second generation advantage would be even
greater with validated data than with reported data.13
For blacks, voting participation increases in a stepwise manner from the
first generation to higher generations. A similar straight-line pattern can be
'3Since those in the second generation would be less likely to misreport the vote, using validated data would lead to an even higher likelihood of voting among second generation respondents when compared to those in the first generation or the 3+ generation.
IMMIGRANT
INCORPORATION
AND POLITICAL
PARTICIPATION
IN THE U. S.
Figure 4
887
Immigrant Incorporation and Voting Participation in 1998
Logistic Regression Estimates of Selected Variables in Full Sample Model
White,
1st Gen 4 0 ys
1stGen 10-19ys
1st Gen 19 ys
Second Generation
Black,
1st Gen 4 0 ys
1st Gen 10-19ys
1st Gen 19 ys
Second Generation
Third and Higher Gen
Asian,
1st t e n 4 0 ys
1st Gen 10-19 ys
1st Gen 19 ys
Second Generation
Third and Higher Gen
Latino,
1st Gen <lOys
1st Gen 10-19 ys
1st Gen 19 ys
Second Generation
Third and Higher Gen
From Repressive Regime
Parent from Repressive Regime
In Spanish Ballot Area
Latino x in Spanish Ballot Area
State % Black
Black x State % Black
State % Asian
Asian x State % Asian
State % Latino
Latino x State % Latino
State Voting History
California
-2.00
-1.50
-1.oo
-0.50
0.00
0.50
1.oo
COEFFICIENT
*p<O.t. **p<0.05. ***pcO.Ol
For full model, see Table A-3 in the Appendix
found among Asian-Americans, although the increase in participation seems
to taper off after the second generation.14 This tapering off in participation
means that racial gaps in participation persist for Asian-Americans in the
third generation and higher. Even after being in the United States for three
or more generations, Asian-Americans are considerably less likely to vote than
their white, black and Latino counterparts. Finally, among Latinos, the highest likelihood of participation is among first generation citizens who have
'41998 was the only year in which the likelihood of participation among him-Americans
increased from the second generation to the 3+ generation.
888
INTERNATIONAL MIGRATION
REVIEW
lived in the United States for 20 years or more.15 The likelihood of participation actually decreases for second generation Latinos and remains low for
those in the third generation and higher. So, we see that generational patterns
in participation vary across racial/ethnic groups - highest participation
among long-term immigrants for Latinos, a straight-line pattern among
blacks, a tapering off in participation for Asian-Americans, and a second generation advantage among whites.
What accounts for these different generational patterns in participation?
Looking first at whites, we see that conventional assimilationist accounts of
cultural and political adaptation may explain why participation among second generation citizens is higher than that among first generation citizens.
Second generation respondents may be less likely to face societal discrimination and linguistic barriers to participation and may be less likely than first
generation citizens to be oriented towards homeland politics. However,
“straight-line” assimilationist theories cannot account for why second generation whites are consistently more likely to participate than those in higher
generations. One plausible explanation for the second generation advantage
among whites is that those in the “3+ generation” may be less attached to
political institutions than their parents’ generation.16 The second generation
advantage may also be driven by a sense of relative deprivation based on discrimination against the first generation. In their book Immigrant America,
Alejandro Portes and Rubtn Rumbaut (1996) note that hostile reactions to
first generation immigrants stir “ethnic militancy among subsequent generations.. .descendants of the first immigrants have gained ‘voice’ and have used
15Pantoja and Segura (2000) find, in their analysis of the 1996 TRPI survey, that California
residents who recently naturalized had the highest rates of voting participation among Latinos. Their findings do not necessarily contradict ours. First, even the TRPI survey shows that
immigrants other than those recently naturalized in CaIifornia were less likely to vote than
higher-generation respondents. Furthermore, the date of naturalization among Latinos is not
strongly related to the date of arrival in the United States (Portes and Rumbaut, 1996).
Indeed, during the mid-1990s, many of those who naturalized were long-term residents who
gained eligibility after being legalized under the Immigration Reform and Control Act (IRCA)
of 1986 (Immigration and Naturalization Services, 1997). We cannot have an exact comparison between the CPS and the TRPI surveys because the latter do not ask questions on the
duration of stay in the United States.
‘6In his pioneering work on immigrant adaptation, Marcus Lee Hansen (1938, 1987) suggests
that second-generation enthusiasm for American institutions may spring from a desire to
escape from the immigrant generation’s cultural influences. According to Hansen, this desire
wanes by the third generation: “What the son wishes to forget the grandson wishes to remember” (1987, p. 15). It is unclear, however, whether this resurgence in ethnicity by the third
generation leads to a decrease in attachment to domestic politics.
IMMIGRANT INCORPORATION AND POLITICAL PARTICIPATION IN THE
u. s.
889
it to reaffirm identities attacked previously with so much impunity” (95).
Whatever the reason, it is clear that straight-line assimilationist theories cannot account for the generational differences in participation among whites.
Just as straight-line assimilationist theories are unable to account for the
generational pattern in participation among whites, they are of limited value
in accounting for the patterns among non-whites. First, it is important to
note that, unlike their white counterparts, many blacks, Asian-Americans and
Latinos in the third generation and higher confronted significant barriers to
social incorporation and political participation until the 1970s. Despite this
legacy of disenfranchisement, blacks in the third generation and higher have
been able to reach parity in participation with whites, thanks largely to the
mobilizing effects of the civil rights movement. Several studies have shown
that, even though blacks faced enormous barriers to participation until the
1960s, the civil rights movement was a powerful mobilizer that enhanced citizens‘ sense of group consciousness and political efficacy (Verba and Nie,
1972; Rosenstone and Hansen, 1993). Asian-Americans and Latinos also
faced several institutional barriers to participation until the 1970s, but they
did not experience a social movement on the scale of the civil rights movement.” As other scholars have noted, the absence of such broad-scale movements among third and higher generation Asian-Americans and Latinos may
account for the continued racial gap in participation for such groups (DeSipio, 1996a; Uhlaner et a l , 1989).1* The persistence of this racial gap thus
challenges conventional assimilationist accounts of immigrant incorporation
that would suggest convergence in participation by the third generation.
Duration of Stay in the United States
For some racial/ethnic groups, longer stay in the United States among first generation citizens dramatically increases the likelihood of voting. Among first generation Latinos, those living in the United States for 20 years or more (“longterm residents”) were 39 percent more likely to have voted in the 1996 election
than those living in the United States for 10 to 19 years (“medium-term residents”). The difference was even greater in the 1994 midterm election, where
long-term Latino residents were twice as likely to have voted as medium-term
residents. Similarly, among first generation whites, long-term residents were 43
~7Institutionaibarriers were especially strong for Asian immigrants, most of whom were not
able to naturalize until 1952 (Espiritu, 1992).
‘*Indeed, Pantoja and Segura’s study of Latino participation (2000) suggests that immigrants
who naturalized in California during the 1990s were more likely than native-born Latinos to
be mobilized in the 1996 election.
890
INTERNATIONAL
MIGUTION&VIEW
percent to 70 percent more likely to vote than medium-term residents, depending on the election year.19 For black and Asian-American immigrants, however,
there is no clear relationship between duration of stay in the United States and
the likelihood of voting. So, even though duration of stay is associated with
higher turnout for all racial/ethnic groups in the bivariate analysis (Figure l),
the relationship has a positive slope only for Latinos and whites when controlling for the various other factors in our full model.
Political Experiences in the Home Country
Prior political experience with repressive regimes has no consistent effect on
voting participation. Looking at the first generation subsamples within each
election year, we see that those who emigrate from repressive states do not have
higher rates of voting than those who come from democratic states. While
some may argue that the experience of immigrants from communist countries
is distinct from those that come from other repressive regimes, the results were
almost identical when we ran a separate model using “Communist Regime Origin” instead of “Repressive Regime Origin.” Just as fleeing from communism
may increase the motivation to participate among some groups such as CubanAmericans, it can also lower the likelihood of participation among groups such
as Vietnamese-Americans, many of whom harbor a distrust of the political system (Collett, 2000). Thus, on balance, coming from a communist regime or
repressive regime has no net significant effect on voting participation. Finally,
given the weak effects of repressive regime origins on participation among first
generation respondents, it is perhaps not surprising that such experiences have
even weaker effects on the second generation.
Linguistic Barriers to Participation
As indicated earlier, studies have shown that first generation respondents have
lower levels of English proficiency than those in higher generations. We would
‘9It is possible that part of the low voter turnout among short-term immigrant residents is due
to the misreporting of citizenship status, especially by undocumented migrants who may
report being citizens when in fact they are not. However, our finding that longer duration of
stay leads to greater political participation holds for white immigrants as well as Latino irnmigrants (the former are much less likely to be undocumented than the latter). Finally, even
among Latinos, the relationship between longer duration and greater participation holds
across medium-term and long-term immigrants for all election years. This finding strongly
suggests that political socialization is occurring and that the difference in participation
between short-term and medium-term immigrants is not due solely to the misreporting of citizenship status by recent arrivals.
IMMIGRANT
INCORPORAT~ON
AND POLITICAL
PARTICIPATION
I N THE U. S.
891
therefore expect Spanish-language ballots to increase turnout among first generation Latinos. Our analysis reveals that the presence of Spanish ballots is not
sufficient to ensure higher voting among first generation Latinos.20 Some may
argue that the coefficient is insignificant because the CPS is conducted entirely
in English and therefore does not include those who are most likely to benefit
directly from Spanish-language ballots. While the CPS is conducted in English, non-English speakers are still included in the survey because respondents
who can speak English can report data for household members who do not
speak English well. First generation respondents who have another household
member report for them tend to have a lower level of English proficiency than
those who responses are self-reported (Espenshade and Fu, 1997). We performed an alternative test of the effects of Spanish-language ballots by restricting the sample to first generation Latinos who had lower levels of English proficiency by excluding self-reported responses. In results not reported here, we
found that even in this restricted sample an individual’s residence in areas with
Spanish ballots does not increase the likelihood of voting. Our evidence therefore casts doubt on the importance of Spanish language ballots in stimulating
turnout among first generation Latinos.21
Ethnic Residential Concentration
Proximity to co-ethnics has weak effects on voting participation, both in the
full sample and across immigrant generations. The only exception is among
“3+ generation” Asian-Americans, who are consistently more likely to vote in
states with high proportions of co-ethnics. It is possible that the proportion
of co-ethnics in a respondent‘s metropolitan area is a better measure of ethnic
mobilization than the proportion of co-ethnics in a respondent’s state.22
However, when we substitute metropolitan concentrations for state concentrations of co-ethnics, we find the results to be similar. The lack of any significant effect of ethnic concentration on voting participation thus seems to
support the findings of earlier studies among Latinos, in which factors such
as residential poverty and in high proportions of noncitizens produce lower
than expected mobilization states with high concentration of Latinos (DeSipio, 1996a).
20Even when we modified the interaction term to account for the percent of one’s life a Latino first generation respondent had spent in the US., living in Spanish-ballot areas had no
effect on the likelihood of voting.
*‘This is consistent with tabulated reports from the Latino National Political Survey, which
indicate that the presence of Spanish-language ballots has a modest effect on turnout among
Spanish-dominant speakers (de la Garza eta[., 1992).
892
INTERNATIONAL
MIGRATION
REVIEW
State History of Voting Participation
A state’s political culture, as measured by its recent history of voting participation, leads to a significant increase in an individual’s likelihood to vote.
Although the coefficients for the variable are relatively small, their impact is
significant when one considers the wide variation in the legacy of voting participation across states. Depending on the election year, those living in
Delaware, a state one standard deviation above the mean level of participation, were 49 percent to 58 percent more likely to have voted than those living in Tennessee, whose history of voting participation is one standard deviation below the mean. What is significant for our study of immigrant voting
participation is that this increase is similar across all immigrant generations.23
If, as traditional assimilation theories would suggest, first generation citizens are less likely to be incorporated into social and political institutions in
the United States, then one would expect them to be the group least affected
by their state’s political culture. Our results indicate otherwise - even first
generation citizens are influenced by the legacy of voting participation in
their state of residence.
Mobilization over Anti-Immigrant Legislation
We hypothesized that anti-immigrant legislation would have a differential
impact on voting participation across immigrant generations. We do not
have individual-level data on whether respondents were aware of anti-immigrant legislation or were motivated to vote because of it. However, we do
know that anti-immigrant legislation emerged in California in 1994, two
years before the passage of national legislation restricting immigrant access to
welfare benefits. Proposition 187, which sought to deny public benefits to
undocumented immigrants, was a prominent election issue in 1994, and
Governor Pete Wilson made it a major part of his re-election campaign. By
contrast, anti-immigrant legislation did not enter the national legislative
agenda until after the Republican takeover of Congress in 1994 and did not
become law until the summer of 1996. Thus, mobilization over anti-immigrant legislation began in California months before the 1994 general election,
state-wide contests and presidential elections, we assume that ethnic groups, party organizations and candidate organizations attempt to mobilize ethnic voters from across the state.
However, it is likely that the per-capita cost of ethnic mobilizationwill be lower in cities where
there are higher concentrations of co-ethnics.
23The only exception is for first generation immigrants in 1994.
22For
IMMIGRANT INCORPORATION AND POLITICAL PARTICIPATION IN THE
u. s.
893
whereas in the rest of the country mobilization did not occur until well after
the election. Given this difference in the exposure to anti-immigrant legislation, we would expect first and second generation immigrants in California
to be much more likely to vote in 1994 than their generational peers elsewhere. Furthermore, since anti-immigration legislation was no longer confined to California by 1996, we would not expect California residence to have
a unique impact on immigrant participation after the 1994 election.24
The evidence from 1994 through 1998 supports our hypothesis regarding anti-immigrant legislation and voting behavior. In 1994, first generation
immigrants in California were twice as likely to have voted as their generational counterparts in other states that did not have similar measures. Similarly, second generation immigrants in California were 83 percent more likely to have voted as their generational peers elsewhere. By contrast, “3+ generation” citizens in California were only 32 percent more likely to have voted
than their generational peers in other states. By 1996, the unique mobilization among first and second generation citizens in California seems to have
dissipated. After the passage of national legislation in 1996, the enormous
“California effect” for first and second generation immigrants disappeared,
whereas it remained roughly the same for “3+ generation” respondents.25
Similar results held true for 1998, an election that had a governor’s race in
California like in 1994, but that did not have a ballot initiative related to
immigration.26 During that year, the “California effect” was significant only
among first and 3+ generation respondents. Even among first generation
respondents, the magnitude of the California advantage paled in comparison
to that in 1994 - 33 percent in 1998 versus 105 percent in 1994. The results
therefore suggest that anti-immigrant legislation in 1994 did indeed have a
significant and selective impact on voting among first and second generation
citizens in California. While a true test of our hypothesis would require individual-level data on political knowledge and opinions, these findings suggest
a strong link between anti-immigrant legislation and immigrant participation.
1998, the initiative on bilingual education in California (Proposition 227) captured the
attention of many immigrant voters, but the measure was part of the June primary, not the
November election.
25Based on their andysis of theTRP1 survey, Pantoja and Segura (2000) argue that anti-immigrant legislation continued to have a positive impact on voting in 1796 among recently-naturalized Latinos in California. Our evidence indicates that such an impact was not present
among the rest of the naturalized population.
ZGUnlike Proposition 187, the 1978 initiative on bilingual education (Proposition 227) was
held in the June primary and not the November election.
2%
894
INTERNATIONAL
MIGRATION
REww
DISCUSSION
This article has examined the influence of immigrant generational status on
voting participation in U.S. elections. Adding to the accumulating evidence
from other recent studies of immigrant incorporation, our analysis casts
doubt on the applicability of straight-line assimilationist theories to contemporary immigrant adaptation. Immigrant political incorporation proceeds in
a different manner for members of different racial/ethnic groups. Blacks are
the only group for whom participation increases in a linear manner from the
first generation to higher generations. For whites, we find that voting participation is highest among second generation respondents. As we had suggested earlier, this second generation advantage may be due to higher levels of
cynicism among those in the third generation or higher or to a greater sense
of relative deprivation among second generation respondents. Further examination of surveys such as the National Election Study or historical studies of
particular communities may shed additional light on the reasons why second
generation whites have the highest rates of voting participation. For AsianAmericans, there is an increase in voting participation after the first generation, but it tapers off after the second generation. For Latinos, the likelihood
of voting is lower among second and “3+ generation” respondents than
among long-term immigrant residents. The persistence of a racial gap in participation among Latinos and Asian-Americans challenges assimilationist
accounts of immigrant incorporation that would suggest a convergence in
participation by the third generation. What we find instead is evidence of
segmented assimilation, whereby different processes of political socialization
produce different generational patterns in participation that do not lead to
convergence by the third generation.
The durability of racial differences in political participation is not the
only limitation of straight-line assimilationist theories of immigrant incorporation. We also find that first generation citizens are just as likely to be influenced by their state “political culture” as citizens from higher generations.
Furthermore, linguistic barriers may not be the primary obstacles to voting
among first generation Latinos, because Spanish-language ballots do not
increase the likelihood of participation among first generation citizens. Finally, longer stay in the United States leads to greater participation for whites
and Latinos of the first generation, but not for blacks or Asian-Americans.
Thus, the evidence presented in this study raises serious doubts about the
applicability of straight-line assimilationist theories across immigrant generations and across racial/ethnic groups.
IMMIGRANT
INCORPORATION
AND POLITICAL
PARTICIPATION
IN THE U. S.
895
Our analysis also indicates that prior political experience is not as important to voting participation as has been previously suggested. Immigrants
from communist states or other repressive regimes do not have higher rates of
participation than immigrants from democratic countries. Furthermore, second generation citizens whose parents emigrated from repressive regimes are
not more likely to vote than those whose parents came from democratic
regimes. Thus, the high level of voting participation among first and second
generation Cuban-Americans seems to be an exception that does not hold for
immigrants from other repressive regimes.
What does lead to higher voting participation among first and second
generation immigrants is the presence of anti-immigrant legislation. Immigrant legislation in 1994 had profound impacts on turnout among first and
second generation immigrants in California. Some may argue that the loss of
a “California effect” among first and second generation immigrants in 1996
is indicative of the generalized mobilization of a national election and not of
immigration issues in particular. However, the fact that the California advantage persisted among “3+ generation” respondents through 1998 while it
largely disappeared among first and second generation respondents suggests
that anti-immigrant legislation in California did indeed have a selective
impact on voting among first and second generation immigrants.
Taken together, these findings add substantially to our knowledge of the
determinants of voting participation and extend in important ways the
lessons from more traditional models of voting behavior. Our findings also
point to the need for more systematic research on immigrant incorporation
and political participation. We need more studies of generational differences
in political participation for activities other than voting - attending rallies,
writing legislators, working for a campaign, and contributing money for
political causes.27 There also needs to be more research on the attitudinal factors and historical processes that have led to a higher level of participation
among second generation whites. It remains to be seen whether the patterns
and trends noted in this analysis will continue into the 2000 elections and
beyond. With record numbers of immigrants applying for naturalization,
with the “new” second generation reaching voting age, and with parties and
interest groups attempting to mobilize ethnic voters, generational differences
in voting participation may not remain the same in the next decade as they
have during the 1990s.
27Louis Desipio (1999) provides a preliminary analysis of differences in such activities as
reflected in the Latino National Political Survey and the Citizen Participation Survey.
896
INTERNATIONAL
MIGRATION
REVIEW
APPENDIX
Data on individual characteristics and voting participation were obtained
from the Voter Supplement to the November Current Population Survey in
1994, 1996 and 1998. Data sources for other variables are noted below.
INDIVIDUAL RESOURCES
Age: Series of dummy variables with “Age 18-29” as the omitted category.
Income: Series of dummy variables with Family Income below $10,000 as the
omitted category.
Education: Less than High School as the omitted category.
High School Grad 1 if respondent graduated or obtained GED or equivalent, 0 otherwise.
Some College: 1 if respondent received associate’s degree or did not complete college, 0 otherwise.
College Grad 1 if respondent received a bachelor’s degree, 0 otherwise.
Post Graduate: 1 if respondent received a graduate degree, 0 otherwise.
RACE/ETHNICITYAND OTHER DEMOGRAPHIC VARIABLES
RacelEtbnicity: Non-His panic white as the omitted category.
Black 1 if black, non-Hispanic, 0 otherwise.
Asian: 1 if Asian, non-Hispanic, 0 otherwise.
Latino: 1 if respondent of Hispanic Origin (any race), 0 otherwise.
Female: 1 if female, 0 otherwise.
First Generation: 1 if born outside the United States or outlying territories, 0
otherwise.
First Generation 0-9 ys. in the United States: First Generation and lived in the
United States less than ten years.
First Generation 10-19ys. in the United States: First Generation and lived in
the United States 10 to 19 years.
First Generation20+ys. in the United States: First Generation, and lived in the
United States more than 20 years.
Second Generation: 1 if born in the United States or outlying territories and
either parent not born in the United States or outlying territories, O otherwise.
3+ generation: 1 if respondent born in the United States with U.S.-born parents, 0 otherwise.
INTERNATIONAL
MIGRATION
&VIEW
897
SOCIAL INCORPORATION
Long-Term Resident 1 if respondent has lived in the same address for five years
or more, 0 otherwise.
Married 1 if married, 0 if unmarried, widowed, divorced or separated.
Employed: 1 if currently employed, 0 otherwise.
Self-Report:1if response if self-reported, 0 otherwise.
INSTITUTIONRC BARRIERS
Liberal Absentee Eligibility 1 if Universal Eligibility (anyone can vote absentee) or Expanded Eligibility (automatic eligibility for citizens of a certain age
or who live a certain distance from the polls), 0 otherwise. Source: Oliver,
1996.
Early Registration Deadline: Individual lives in state where one has to register
at least 21 days prior to election day. Data obtained from the League of
Women Voters (www.lwv.org)
MOBILIZATION CONTEXTS
Presidentid Toss-up:1 if state ranked as “Clinton Slightly Favored,” “Dole
Slightly Favored,” or “Toss-up” = 1, 0 otherwise. Source: Associated Press
Research, as reported in the San Francisco Chronicle (Yoachum, 1996).
Senate or Governor Toss-up: 1 if state ranked as Highly Vulnerable of Vulnerable by the Congressional Quarterly. During each election season, the Congressional Quarterly comes up with a survey of Senate and Governor Races
based on opinion polls and surveys of candidate and party organizations
(Sources: Babson and Groppe, 1994; Cassata, 1998; Cranford, 1994; Foerstel, 1998; Greenblatt, 1996; Greenblatt and Wells, 1996).
OTHER VARIABLES
From Repressive Regime: 1 if respondent came from a repressive regime at the
time of entry to the United States, 0 otherwise. Regime coded as “repressive”
if it was ranked as ‘partly free’ or ‘not free’ by Freedom House. Freedom
House rankings of countries began in 1972. For years prior to 1972, countries were coded as repressive if they were communist countries or if they were
ranked as repressive in 1972, and continued to be repressive for more than 75
percent of the years since 1972. Source: Freedom House (http://www.freedomhouse.org/ratings)
898
INTERNATIONAL
MIGRATION
REVIEW
Parentsfrom Repressive Regime: 1 if one or both of respondent’s parents are
from Repressive Regime, 0 otherwise.
In Spanish Ballot Area: 1 if respondent lives in a County, State or Metropolitan Area with Spanish language ballots. Source: United States General
Accounting Offtce, 1997.
% State Black, % State Asian, % State Latino: State ethnic proportions are
based on Census reports that take into account the total state population.
Source: U.S. Bureau of the Census, “1990 to 1998 Annual Time Series of
State Population Estimates By Race and Hispanic Origin.” http://www.census.gov/population/www/estimates/st-srh.
html
State Voting History: Average of percent voting in each state from 1972-1 992.
Source: Federal Election Commission (http://www.fec.gov/ votregis/turn/
turn.htm)
California: 1 if respondent is resident of California, 0 otherwise.
30-39
40-49
50-59
60-69
70-79
80 +
$10,000-$25,000
$25,000-$50,000
$50,000-$75,000
$75,000 +
Not Reported
High School Grad
Some College
College Grad
Post Graduate
Long-Term Resident
Married
Employed
Female
Liberal Absentee Eligibility
Early Registration Deadline
SenatelGovernor Toss-up
White,
1st Gen < 10 ys.
1st Gen 10 - 19 ys.
1st Gen > 19 ys.
Second Generation
Education:
Income:
Age:
Beta
0.1446<
0.4881'
0.7812"
1.3030'
1.5333'
1.0854c
0.272F
0.4079'
0.6202'
0.6243"
0.3928"
0.71 17'
1.2303'
1.7460'
2.0776'
0.7731"
0.4988"
0.0658'
0.0491"
0.0671"
-0.0519b
0.1922'
-1.7606'
-0.4320b
-0.0346
0.221 7'
Std. Err.
0.0254
0.0266
0.0310
0.0353
0.0395
0.0486
0.0329
0.0337
0.0386
0.0423
0.0426
0.0274
0.0300
0.0358
0.0473
0.0185
0.0191
0.0222
0.0176
0.0214
0.0254
0.0188
0.4137
0.1717
0.0650
0.0370
0.6837'
0.6726'
0.6224'
0.8394'
0.3499'
0.4367'
0.8426'
1.2331'
1.8116'
0.5917'
0.3255'
0.2434b
0.1709b
-0.1903
-0.0456
0.1268
-1.6857'
-0.4292b
0.1504
0.1541
0.1751
0.1785
0.1809
0.1174
0.1283
0.1456
0.1779
0.0848
0.0931
0.1036
0.0817
0.1176
0.1417
0.0996
0.4074
0.1706
First Generation
Beta
Std. Err.
-0.3717'
0.1353
0.1336
0.1628
0.2344
0.1462
0.9721'
0.1629
1.0225'
0.1772
0.9350'
0.1941
Full Sample
0.3376'
0.4261'
0.5591'
0.4663'
0.3735'
0.5574'
1.0045'
1.3641'
1.7799'
0.7131'
0.5756'
-0.1360"
0.0868
-0.0518
0.0292
0.3628'
0.0975
0.1034
0.1229
0.1325
0.1267
0.0783
0.0889
0.1112
0.1410
0.0607
0.0597
0.0717
0.0560
0.0770
0.0864
0.0645
Second Generation
Beta
Std. Err.
-0.0905
0.1013
0.2543b
0.1027
0.5901'
0.1044
0.9391'
0.0957
1.1757'
0.0987
0.8923'
0.1126
1994 ELECTION
OF V m N G IN THE
LOGISnC REGRESSION bkTIh4A''
TABLE 1
0.2408'
0.3908'
0.6221'
0.6272'
0.3980'
0.7594'
1.2881'
1.8299'
2.1406'
0.7907'
0.4990'
0.0815'
0.0392b
0.0849'
-0.0495"
0.1727'
0.0361
0.0368
0.0420
0.0462
0.0468
0.0305
0.0332
0.0393
0.0524
0.0201
0.0207
0.0241
0.0191
0.0229
0.0272
0.0202
3+ Generation
Beta
Std. Err.
0.1823'
0.0269
0.5155'
0.0283
0.8285'
0.0334
1.3660'
0.0392
1.5934'
0.0450
1.0494'
0.0575
Data on individual characteristics and voting participation were obtained from the Voter Supplement to the November Current Population Survey in 1994, 1996 and 1998. Data sources for other variables are noted below.
APPENDIX
00
LD
W
2
zs
2:
5
$
$
z5
E
#5
F
p < 0.1, p < 0.05, p < .01
n
a
- b.
1st Gen < 10 vs.
1st Gen 10 19
1st Gen z 19 ys.
Second Generation
Third and Higher Gen
Asian,
1st Gen < 10 ys.
1st Gen 10 - 19 ys.
1st Gen > 19 ys.
Second Generation
Third and Higher Gen
Latino,
1st Gen 10 ys.
1st Gen 10 - 19 ys.
1st Gen > 19 ys.
Second Generation
Third and Higher Gen
From Repressive Regime
Parent from Repressive Regime
In Spanish Ballot Area
Latino x In Spanish Ballot Area
State % Black
Black x State % Black
State % Asian
Asian x State Yo Asian
State % Latino
Latino x State % Latino
State Voting History
California
Constant
pseudo R2
Black.
Full Sample
Beta
Std. Err.
0.1986
0.8448
0.3423
-0.1285
0.2927
-0.2248
0.2590
0.0484
0.0641
0.1621b
0.3681
-1.5244'
0.1511
-1.1598'
0.1530
-0.9199'
-0.7442'
0.1699
0.1821
-0.7847'
0.3889
-1.4736'
0.2129
-0.7452'
0.1565
-0.0538
0.1301
-0.4430'
0.1001
-0.5037'
0.1147
-0.2447b
0.0640
-0.0433
0.0268
-0.0485'
0.0851
0.2521'
0.0015
0.0023
0.0032
0.0044
0.0025
0.0056b
0.0038
0.0133'
0.0014
0.0053'
0.0039
0.0056
0.0022
0.0267'
0.0473
0.3303'
0.1475
-4.1964'
0.1556
87,596
0.4177
0.2663
0.2249
-1.3628'
-0.5735a
0.1998
3.952
",,,-
-0.1092
0.1728
-0.1641
0.1683
0.0213
0.1078
0.1426
0.2044
-0.0360'
0.0096
-0.0305
0.0384
-0.0196
0.0128
0.0241"
0.0138
-0.0158b
0.0078
0.0104
-0.0093
-0.0436'
0.0130
0.1821
0.7185'
0.8698
0.8791
0.1488
0.3654
0.1707
0.1702
-1.5680'
-1.1504'
-0.8541'
First Generation
Beta
Std. Err.
0.5536
0.8901
0.2862
0.6860
0.3226
0.6578
-0.0224
-0.1429'
0.0858
0.0005
0.0175
0.0038
0.0132
0.0017
0.0087
0.0310'
0.6057'
-3.8464'
0.1502
8,895
-0.7104'
-0.9387'
-0.3488
0.0664
0.0733
0.1677
0.0060
0.0379
0.0076
0.0095
0.0056
0.0082
0.0086
0.1385
0.5877
0.1823
0.2087
0.5638
Second Generation
Beta
Std. Err.
TABLE 1 (CONTINUED)
L€)GlSTIC REGRI?SSIONl?.STlMAlES OF vCYI"G IN THE 1994 ELECTION
-0.0336
0.3670'
0.0035b
0.0043
0.0072'
0.0133b
0.0062"
0.0084
0.0283"
0.2759"
-4.3813'
0.1571
74,749
-0.6034'
-0.8294"
0.1754'
0.0299
0.1 169
0.0016
0.0033
0.0027
0.0056
0.0015
0.0052
0.0023
0.0537
0.1568
0.1317
0.2402
0.0652
3+ Generation
Beta
Std. Err.
30-39
$10,000-$25,000
$25,000-$50,000
$50,000-$75,000
$75,000 +
Not Reported
Education:
High School Grad
Some College
College Grad
Post Graduate
Long-Term Resident
Married
Employed
Female
Liberal Absentee Eligibility
Early Registration Deadline
Presidential Toss-up
SenatelGovernor Toss-up
White,
1st Gen < 10 ys.
1st Gen 10 - 19 ys.
1st Gen > 19 ys.
Second Generation
1st Gen < 10 ys.
Black,
1st Gen 10 - 19 ys.
1st Gen > 19 ys.
Second Generation
Third and Higher Gen
Age:
40-49
50-59
60-69
70-79
80 +
Income:
TABLE 2
Full Sam~le
Beta
s.e.
0.1023'
0.0267
0.0282
0.3791'
0.0328
0.6897'
1.2274'
0.0393
0.0432
1.3438'
0.0532
0.9179'
0.1681'
0.0357
0.0364
0.3481'
0.0413
0.5905'
0.0455
0.7401'
0.0440
0.3918'
0.0283
0.6948'
0.0314
1.3348'
1.8946'
0.0383
0.0551
2.3005'
0.0199
0.6064'
0.0205
0.4105'
0.0236
0.0536b
0.1620'
0.0188
0.0237
0.0432'
0.0278
0.0000
0.0245
0.041 1
0.0213
-0.0330
0.2696
-1.4789'
-0.4998'
0.1557
0.0769
0.0317
0.0446
0.2906'
-0.6928
0.6803
0.4612'
0.2859
0.1131
0.2932
0.2470
0.0668
0.3771'
0.0654
-1.3653'
0.0431
-0.1367
0.8005
0.4756
0.5453
First Generation
Beta
s.e.
0.1627
0.1219
0.2005
0.1236
0.2501'
0.1436
0.7711'
0.1559
0.1820
0.8897'
0.7212'
0.1993
0.0366
0.1527
0.1509
0.0763
0.1689
0.1754
0.4969'
0.1 826
-0.4115b
0.1795
0.3561'
0.1150
0.1255
0.9757'
0.1351
1.0749'
1.6098'
0.1872
0.3202'
0.0854
0.1194
0.0900
0.0981
-0.0199
0.0802
0.0755
0.1277
0.1230
-0.1417
0.1496
-0.1604
0.1201
0.1059
-0.1555
0.2687
-1.5110'
0.1542
-0.6278'
0.3762
0.4886
Second Generation
Beta
s.e.
-0.1316
0.1092
0.2743b
0.1207
0.6416'
0.1295
1.1659'
0.1198
1.4029'
0.1142
0.7683'
0.1254
0.1200
0.3409'
0.4653'
0.1276
0.4834'
0.1445
0.6986'
0.1579
0.3101b
0.1454
0.0903
0.5989'
1.2097'
0.1061
1.7848'
0.1329
2.2 131
0.1774
0.0697
0.7785'
0.0696
0.5070'
0.0830
0.0382
0.0659
0.1670b
0.0181
0.0959
0.1087
0.0378
0.1263
0.0959
0.0820
0.0971
h3GlSTIC REGResslON &ITMATES OF VOTING IN THE 19% ELECTION
0.3858'
0.0667
3+ Generation
Beta
s.e.
0.1135'
0.0283
0.0300
0.3975'
0.7249'
0.0351
1.2644'
0.0434
1.3385'
0.0487
0.0629
0.9327'
0.1502'
0.0387
0.3470'
0.0393
0.6205'
0.0447
0.7546'
0.0493
0.4416'
0.0478
0.7389'
0.0310
1.3811'
0.0342
1.9833'
0.0421
2.3881'
0.0616
0.6096'
0.0215
0.4174'
0.0221
0.0255
0.0588b
0.1671'
0.0203
0.0251
0.0435'
0.0068
0.0295
0.0262
0.0497'
-0.0405"
0.0229
1st Gen < 10 ys.
1st Gen 10 - 19 ys.
1st Gen > 19 ys.
a
p < 0.1, p < 0.05, p < .01
In Spanish Ballot Area
Latino x In Spanish Ballot Area
State % Black
Black x State % Black
State Yo Asian
Asian x State % Asian
State Yo Latino
Latino x State % Latino
State Voting History
California
Constant
pseudo R2
n
Second Generation
Third and Higher Gen
Latino,
1st Gen < 10 ys.
1st Gen 10 - 19 ys.
1st Gen > 19 ys.
Second Generation
Third and Higher Gen
From Repressive Regime
Parent from Repressive Regime
Asian,
TABLE 2 (CONTINUED)
-0.0042a
0.0114'
0.0028
0.0003
0.0267'
0.2987'
-3.4555'
0.1435
76,656
0.0040
-0.0691'
0.1141
0.0058'
0.0320
0.1065
0.0017
0.0032
0.0025
0.0038
0.001 8
0.0045
0.0023
0.0493
0.1530
Full Sample
Beta
s.e.
-0.6250'
0.2965
-0.8858'
0.1410
0.1642
-1.0908'
-0.7747'
0.1547
-0.8257'
0.1890
-0.1983
0.2892
-0.0130
0.1818
0.3157'
0.1613
-0.1594
0.1245
-0.2159'
0.0947
-0.1850
0.1127
-0.0051
0.0660
0.1205
0.2296
0.0080
0.0295
0.0140
0.0148
0.0086
0.01 10
0.0124
0.1682
0.7919
0.1584
0.1573
-0.0078
-0.1481
-0.0871
0.0513
0.0020
0.0 156
0.0019
0.0001
0.0102
-0.0230'
0.0295b
0.0984
-2.1104'
0.1047
3,811
0.2968
0.2457
0.2322
-0.3041
0.0357
0.6156'
First Generation
Beta
s.e.
0.2800
-0.6707b
-0.9160'
0.1636
0.1762
-0.8672'
0.0731
0.1943
0.1963
-0.1643"
0.0957
0.2236
-0.1575
0.0072
0.0046
-0.0281
0.0255
0.0079
-0.0015
0.0002
0.0093
0.0103
0.0066
0.0102
0.0073
0.0255'
0.0095
0.1954
0.1454
-3.2787'
0.6250
0.1619
7,245
0.0081
0.4331'
-0.8065'
Second Generation
Beta
s.e.
b G I S T I C &GREssrON &TIMAT€S OF VOTING IN THE 1996 ELECTION
-0.0517
0.3244b
0.0061'
0.0043
-0.0043
0.0171'
0.0012
-0.0031
0.0266'
0.3330'
-3.5307'
0.1456
65,600
-0.2634b
-1.0320'
0.0356
0.1458
0.0018
0.0032
0.0027
0.0057
0.0020
0.0058
0.0024
0.0569
0.1629
0.1 157
0.2647
3 t Generation
Beta
s.e.
2:
28
c
z5
$
%
5
3
P
2
Not Reported
High School Grad
Some College
College Grad
Post Graduate
$10,000-$25,000
$25,000-$50,000
$50,000-$75.000
$75,000 +
80 t
30-39
40-49
50-59
60-69
70-79
Long-Term Resident
Married
Employed
Female
Liberal Absentee Eligibility
Early Registration Deadline
SenatelGovernorToss-up
White,
1st Gen < 10 ys.
1st Gen 10 - 19 ys.
1st Gen > 19 ys.
Second Generation
Black,
1st Gen c 10 ys.
1st Gen 10 - 19 ys.
1st Gen > 19 ys.
Second Generation
Third and Higher Gen
Education:
Income:
Age:
TABLE 3
0.1405'
0.4687'
0.7648'
1.3428'
1.5512'
1.0961'
0.1805'
0.3310'
0.4559'
0.4974'
0.3361'
0.6266'
1.1819'
1.5780'
1.967 1'
0.7284'
0.4894'
-0.0047
0.0522'
0.0920'
-0.1 158'
0.3204'
-1.0001'
-0.5186'
-0.1581b
0.2016'
-0.4889
-0.4487
-0.2341
-0.1213
0.2247'
0.0279
0.0276
0.0308
0.0371
0.0412
0.0516
0.0418
0.0413
0.0448
0.0463
0.0465
0.0300
0.0321
0.0369
0.0468
0.0195
0.0202
0.0236
0.0183
0.0220
0.0269
0.0194
0.3 130
0.1810
0.0692
0.0414
0.4934
0.2558
0.1970
0.2015
0.0645
Full Sample
Beta
s.e.
0.1282
0.1209
0.1324
0.1415
0.1653
0.2218
0.1728
0.1661
0.1827
0.1850
0.1861
0.1141
0.1241
0.1366
0.1674
0.0835
0.0869
0.0988
0.0784
0.1120
0.1516
0.0933
0.3043
0.1838
0.5766
0.3674
0.3451
-0.3421
-0.1585
0.0659
-0.6300
-0.0089
0.2446b
0.3803'
1.1771'
1.2925'
0.6925'
0.3248b
0.3865'
0.4714'
0.4969'
0.2217
0.4716'
1.02 12'
1.2572'
1.7274'
0.7185'
0.5156'
-0.2023'
0.0110
0.0496
-0.0507
0.2281'
0.4226
0.1065
0.1106
0.1105
0.1 123
0.1063
0.1182
0.1334
0.1321
0.1470
0.1478
0.1440
0.0876
0.0979
0.1 147
0.1435
0.0653
0.0655
0.0779
0.0604
0.0807
0.1005
0.0672
1998 ELECTION
Second Generation
Beta
s.e.
-0.0004
0.2628b
0.7314c
0.7824'
0.9869'
0.4414b
-0.2930'
-0.1095
-0.1273
0.0436
-0.2210
0.5166'
0.776G'
1.1182'
1.4518'
0.6709'
0.3377'
-0.1818'
0.0648
0.1136
-0.1505
0.4235'
-0.9554'
-0.4355b
First Generation
Beta
s.e.
LOGISnC &GREsslON ESTIMATES OF VOTING IN THE
0.2418'
0.1646'
0.4987'
0.8015'
1.4041'
1.6041'
1.1883'
0.1856'
0.3441'
0.4796'
0.5207'
0.3727'
0.6676'
1.2401'
1.6566'
2.0556'
0.7316'
0.4943'
0.0300
0.0526'
0.0976'
-0.12 18'
0.3240'
0.0664
0.0298
0.0295
0.0332
0.0412
0.0466
0.0604
0.0456
0.0451
0.0488
0.0507
0.0510
0.0334
0.0356
0.0409
0.0521
0.021 1
0.0219
0.0256
0.0199
0.0235
0.0286
0.0210
3+ Generation
Beta
s.e.
n
1 s t Gen <
= p < 0.1, p < 0.05, ' p < .01
1st
10 vs.
Gen 10 - 19 ys.
1st Gen > 19 ys.
Second Generation
Third and Higher Gen
Latino,
1st Gen < 10 ys.
1st Gen 10 - 19 ys.
1st Gen > 19 ys.
Second Generation
Third and Higher Gen
From Repressive Regime
Parent from Repressive Regime
In Spanish Ballot Area
Latino x In Spanish Ballot Area
State % Black
Black x State % Black
State % Asian
Asian x State % Asian
State % Latino
Latino x State % Latino
State Voting History
California
Constant
pseudo R2
Asian.
Full Sample
Beta
s.e.
-1.3368'
0.3490
-1.5461'
0.1588
-1.2426'
0.1291
-1.1016'
0.1569
-0.6367'
0.1795
-1.4179'
0.3778
-0.4828b
0.1933
-0.1 132
0.1376
-0.3978'
0.1253
-0.3060'
0.0903
-0.0051
0.1020
-0.1363b
0.0618
0.0406
0.0277
0.1875'
0.0991
0.0022
0.0016
0.0171'
0.0031
0.0037
0.0027
0.0212'
0.0038
0.0038b
0.0016
0.0048
0.0039
0.0325'
0.0021
0.1825'
0.0486
-4.4790'
0.1444
0.1473
75,485
0.4005
0.2483
0.2062
0.1470
0.1471
0.1084
0.2087
0.0077
0.0185
0.0114
0.0125
0.0077
0.0092
0.0104
0.1716
0.7026
-1.4966'
-0.5240b
-0.0913
0.2071
-0.3829'
0.0584
0.1742
0.0161b
0.0007
0.0282b
0.0025
-0.0028
0.0046
0.0364'
0.2827"
-3.8551'
0.1375
4,177
0.1982
-0.0967
0.0662
0.1090
0.0802
0.0543
0.2037
0.0053
0.0063
0.0297
0.0238
0.0098
0.0081
0.0178"
0.0096
0.0075
0.0056
0.0023
0.0083
0.0375'
0.0078
0.1711
0.1401
-4.2159'
0.5315
0.1542
7,180
-0.6181'
1998 ELECTION
First Generation
Second Generation
Beta
s.e.
Beta
s.e.
-1.1327'
0.3598
-1.2921'
0.1815
-0.9528'
0.1 542
-1.3362'
0.1979
TABLE 3 (CONTINUD)
LOGISTIC REGRESSION ESI'IMATFS OF VOTING IN THE
0.1065
0.2349
0.0166
0.0309
0.2658"
0.1373
0.0013
0.0017
0.0175'
0.0032
0.0005
0.0030
0.0185'
0.0053
0.0044'
0.0017
0.0014
0.0052
0.0317'
0.0022
0.1749'
0.0560
-4.5455'
0.1549
0.1478
64,128
-0.2627b
-0.490Gb
3+Generation
Beta
s.e.
2
a75
z
5E;
E
5:
E
E;
i!5
2
IMMIGRANT
INCORPORATION
AND POLITICAL
PARTICIPATION
IN THE U. S.
905
Abramson, l? and W. Claggett
1992 “The Quality of Record Keeping and Racial Differences in Validated Turnout,” Journal
ofPolitics, 54(3):871-882.
Arvizu, J. R and F. C. Garcia
1996 “Latino Voting Participation: Explaining and Differentiating Latino Voting Turnout,”
Hispanic Journal OfBehauioral Sciences, 18(2):104-128.
Babson, J. and M. Groppe
1994 “Governors’ Races May End in November Surprise,” Congressional Quarterly Weekly
Report, 52(30):2165-2168.
Bass, L. and L. Casper
1999 “Are There Differences in Registration and Voting Behavior between Naturalized and
Native-Born Americans?” Population Division Working Paper No. 28. Washington,
DC: U.S. Bureau of the Census.
Bobo, L. and E Gilliam, Jr.
1990 “Race, Socioeconomic Status and Black Empowerment,” American Political Science
Review, 84(2):377-394.
Borjas, G .
1999 Heaven; Door: Immigration Policy and the American Economy. Princeton, NJ: Princeton
University Press.
Cain, B. E., R. D. Kiewiet and C. J. Uhlaner
1991 “The Acquisition of Partisanship by Latinos and hian-Americans,” American Journal
of Political Science, 35(2):390-422.
Calvo, M. A. and S. J. Rosenstone
1989 Hispanic Political Participation. San Antonio, TX: Southwest Voter Research Institute.
Camarota, S.
1999 “Immigrants in the United States - 1998.” January 1999 Backgrounder. Washington,
DC: Center for Immigration Studies.
Carnegie Center for International Peace
1997 “New Americans and Co-Ethnic Voting,” Research Perspectives on Immigration, 1:3.
Cassata, D.
1998 “California, The Coveted
56(28):1865-1868.
Prize,”
Congressional Quarter4 Weekly Report,
Cho, W. K. T.
1999 ”Naturalization, Socialization, Participation: Immigrants and (Non) Voting,”Journal of
Politics, 61(4):1140-1155.
Collett, C.
2000 “The Determinants of Vietnamese American Political Participation: Findings from the
January 2000 Orange County Register Poll.” Paper prepared for presentation at the
annual meeting of the Association of Asian American Studies. Scottsdale, Arizona.
Cranford, J.
1994 “The Senate: Election Slate Full of Close Calls Puts Democrats on the Edge,” Congressional Quarterly Weekly Report, 52(41):2996-2999.
906
INTERNATIONAL MIGRATION
REVIEW
de la Garza, R. 0.
1996 “The Effects of Primordial Claims, Immigration, and the Voting Rights Act on Mexican-American Sociopolitical Incorporation.” In The Politics ofMinority Coalitions:Race,
Ethnicity and Shared Uncertainties. Ed. W. C. Rich. Westport, CT: Greenwood Press.
de la Gaaa, R. 0. et al.
1992 Latino Voices: Mexican, Puerto Rican and Cuban Perspectives on American Politics. Boulder, CO: Westview Press.
DeSipio, L.
1999 “The Second Generation: Political Behaviors of Adult Children of Immigrants in the
United States.” Paper prepared for presentation at the annual meeting of the American
Political Science Association. Atlanta.
1996a Counting on the Latino Voice: Latinos as a New Electorate. Charlottesville: University
Press of Virginia.
1996b “Making Citizens or Good Citizens? Naturalization as a Predictor of Organizational
and Electoral Behavior among Latino Immigrants,” Hispanic Journal of Behavioral Sciences, 18(2):194-21 3.
Espenshade, T. J. and H. Fu
1997 “An Analysis of English-Language Proficiency among U.S. Immigrants,” American Son’ological Review, 62(2) :288-305.
Espiritu, Y. L.
1992 Asian-American Panethnicity. Philadelphia, PA: Temple University Press.
Fishman, J.
1972 The Sociology of Language. Rowley, MA: Newbury House.
Foerstel, K.
1998 “Senate Hopefuls’ Trial By Fire,” Congressional Quarterly Weekly Report,
56(28): 1860-1 864.
Frey, W. H.
1996 “Immigration, Domestic Migration and Demographic Balkanization in America: New
Evidence for the 1990s,” Population and Development Review, 22(4):74 1-763.
Fulwood, S.
1998 “Decision ’98: The Final Count,” Los Angelps Emes, November 5, p. 2.
Gans, H.
1992 “Second-Generation Decline: Scenarios for the Economic and Ethnic Futures of the
Post 1965 American Immigrants,” Ethnic and Racial Studies, 15(2): 173-1 92.
Glastris, P. et al.
1997 “Immigration Boomerang,” U S . News and World Report. March 17
Greenblatt, A.
1996 “Open Governorships Provide a Measure of Suspense,” Congressional Quarterly Weekly
Report, 54(40):2961-2963.
Greenblatt, A. and R. Wells
1996 “Senate Elections: Still Anybody’s Call,” Congressional Quarterly Weekly Report,
54(40):2954-2959.
IMMIGRANT INCORPORATION AND POLITICAL PARTICIPATION IN THE
u. S.
907
Guendelman, S. et al.
1990 “Generational Differences in Perinatal Health among the Mexican American Population: Findings from HHANES 1982-84,” American JournalofPublic Health, 80:61-65.
Hansen, M. L.
1987 The Problem of the Third Generation Immigrant: A Republication of the 1937A&ress with
Introductions ly Peter Kivisto and Oscar Handlin. Rock Island, I L Swenson Swedish
Immigration Research Center and Augustana College Library.
1938 The Problem
Society.
of
Third Generation Immigrants. Rock Island, IL: Augustana Historical
Hill, K. and D. Moreno
1996 “Second-Generation Cubans,” Hispanic journal of Behavioral Sciences, 18(2): 175- 193.
Hero, R. E. and A. G. Campbell
1996 “Understanding Latino Political Participation.” Hispanic journal of Behavioral Sciences,
18(2):1129-141.
Jennings, K. and R. Niemi
1981 Generationsand Politics: A Panel Study of YoungAdults and Their Parents. Princeton, NJ:
Princeton University Press.
Jones-Correa, M.
1998 Between Two Nations: The Political Life of Latin American Immigrants in New York City
Ithaca, Ny: Cornell University Press.
Junn, J.
1999 “Participation in Liberal Democracy: The Political Assimilation of Immigrants and Ethnic Minorities in the United States,” American BehavioralScientist, 42(9):1417-1438.
Kahn, J. R.
1994 “Immigrant and Native Fertility during the 1980s: Adaptation and Expectations for the
Future,” International Migration Review, 28(3):501-5 19.
Kellstedt, L.
1974 “Ethnicity and Political Behavior: Inter-Group and Inter-Generational Differences,”
Ethnici3 ( 1) :393-4 15.
Key, V. 0.
1949 Southern Politics in State and Nation. New York: A. A. Knopf.
Lamare, J. W.
1982 “The Political Integration of Mexican-American Children: A Generational Analysis,”
International Migration Review, 16(1):169-188.
Liang, Z.
1994 “Social Contact, Social Capital and the Naturalization Process: Evidence from Six
Immigrant Groups,” Social Science Research, 23(4):407437.
Lien, P.
1998 “Does the Gender Gap in Attitudes and Behavior Vary across Racial Groups?” Political
Research Quarter& 51 (4):869-895.
1997 The Political Participation of Asian-Americans: W i n g Behavior in Southern California.
New York: Garland Publishing.
908
INTERNATIONAL
MIGRATION
REVIEW
M p a , D. E.
1999 “Social and Linguistic Aspects of Assimilation Today.” In The Handbook of International Migration - The American Eiperience. Ed. C. Hirschman, P.Kasinia and J. DeWind.
New York: Russell Sage Foundation.
McDonnell, I? and G. Ramos
1996 “Latinos Make Strong Showing at the Polls,” The Los Angehs Times, November 8.
Mollenkopf, J., T. Ross and D. Olson
1999 “Immigrant Political Participation in New York and Los Angeles.” New York: Center for
Urban Research, City University of New York.
Miller, W. E. and J. M. Shanks
1996 The New American Eter Cambridge, MA: Harvard University Press.
Nakanishi, D. T.
1986 “The UCLA Asian Pacific American Voter Registration Study.” Los Angeles: Asian
Pacific American Legal Center.
Neal, T. and R. Morin
1998 “For Voters, It’s Back toward the Middle,” The Washington Post, November 5.
Oliver, J. E.
1996 “The Effects of Eligibility Restrictions and Party Activity on Absentee Voting and Overall Turnout,” American Journal of Political Science, 40(2):498-5 13.
Pantoja, A. and G. Segura
2000 “Citizens by Choice, Voters by Necessity: Patterns in Political Mobilization by Naturalized Latinos.” Paper prepared for presentation at the annual meeting of the Midwest
Political Science Association. Chicago.
Portes, A. and D. MacLeod
1999 “Educating the Second Generation: Determinants of Academic Achievement among
Children of Immigrants in the United States,” Journal of Ethnic and Migration Studies,
25(3):373-396.
Portes, A. and R. Mozo
1985 “The Political Adaptation Process of Cubans and Other Ethnic Minorities in the United Stares: A Preliminary Analysis,’’ International Migration Review, 19(1):35-63.
Portes, A. and R. G. Rumbaut
1996 Immigrant America: A Portrait, Second Edition. Berkeley, C A University of California
Press.
Portes, A. and M. Zhou
1993 “The New Second Generation: Segmented Assimilation and Its Variants,” Annals of the
America Academy of Political and Social Science, 530:7&96
Presser, S., M. W. Traugott and S. Traugott
1990 “Vote ‘Over’ Reporting in Surveys: The Records or the Respondents.” Technical Report
No. 39, National Election Study. <http:llwww.umich.edu/-nes/>
Rosenstone, S. J. and M. Hansen
1993 Mobilization, Participation and Democracy in America. New York: Macmillan Publishing
Company.
Rumbaut, R.
1997 “Assimilationand Its Discontents: Between Rhetoric and Reality,” International Migration Rpview, 31(4):923-960.
IMMIGRANT
INCORPORATION
AND POLITICAL.
PARTICIPATION
IN THE U. S.
909
Scales, A.
1998 “Blacks, Especially in South, Answered the Democrats‘ Call,” Boston Globe, November
5, p. A34.
Schachar, R. and B. Nalebuff
1999 “Follow the Leader: Theory and Evidence on Political Participation,” TheAmerican Economic Review, 89(3):525-547.
Schattschneider, E. E.
1960 The Semi-Sovereign Peopk: A Realist’s View OfDmocrary. New York: Holt, Rinehart and
Winston.
Shaw, D., R. 0. de la Garza and J. Lee
2000 “Examining Latino Turnout in 1996: A Three-State, Validated Survey Approach,”
American Journal of Political Science, 44(2) :338-346.
Stevens, G.
1992 “The Social and Demographic Context of Language Use in the United States,”American Sociological Review, 57(2):171-185.
Uhlaner, C., B. Cain and R. Kiewiet
1989 “Political Participation of Ethnic Minorities in the 1980s,” Political Behavior,
1l(3): 195-23 1.
United States General Accounting Offke
1997 “Bilingual Voting Assistance - Assistance Provided and Costs.” GAO-GGD-97-8 1.
May.
United States Immigration and Naturalization Service
1999 September FY 1999 Ear End Report. Washington, DC: United States Immigration and
Naturalization Service.
Verba, S. and N. H. Nie
1972 Participation in America. Chicago: University of Chicago Press.
Verba, S., K. L. Schlozman and H. E. Brady
1995 Voice and Equality. Cambridge, MA: Harvard University Press.
White, M. J. and A. Omer
1997 “Segregation by Ethnicity and Immigrant Status in New Jersey.” In Kqs to Success&l
Immigration. Ed. T. J. Espenshade. Washington, DC: Urban Institute Press.
Wolfinger, R. E. and Ss. J. Rosenstone
1980 Who Votes? New Haven: Yale University Press.
Wong, J.
2000 “Institutional Context and Political Mobilization among Mexican and Chinese Immigrants in New York and Los Angeles.” Paper presented for presentation at the annual
meeting of the American Political Science Association. Washington, DC.
Yang, P.Q.
1994 “Examining Immigrant Naturalization,” International Migration Review,
28(3):449477.
Yoachum, S.
1996 “Clinton Seems to Have Enough Electoral Votes,” San Francisco Chronicle. November
1, p. A3.
Zhou, M.
1997 “Segmented Assimilation: Issues, Controversies and Recent Research on the New Second Generation,” International Migration Review, 31(4):975-1008.