Group Title: political participation of Asian Americans in the early 1990s
Title: The political participation of Asian Americans in the early 1990s
Full Citation
Permanent Link:
 Material Information
Title: The political participation of Asian Americans in the early 1990s
Physical Description: Book
Language: English
Creator: Lien, Pei-te, 1957-
Copyright Date: 1995
 Record Information
Bibliographic ID: UF00102717
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 33486698
ltuf - AKP2552

Full Text








Being a "non-traditional" student in many senses, I am

deeply indebted to all the faculty members, classmates,
office staff, friends, and family members surrounding me

who, at one time or another, have come to my help and

contributed in different ways to the completion of this


My sincere appreciation goes first to my committee

chairwoman, Dr. M. Margaret Conway, who, despite many

unanticipated professional and personal burdens, never

turned down any of my numerous requests for advice and

support throughout the years. I thank her for believing in

my ability when I was doubtful; I thank her for listening
when I was confused; and I thank her for being patient and

understanding when my writing was awkward and my sentences

cumbersome. For similar reasons, I am very grateful to

other members of my committee: Dr. Wayne Francis, Dr.

Michael Martinez, Dr. James Button, and Dr. Leonard Tipton.

Without the constant guidance and encouragement from these

professors, this dissertation would not have been possible.
The Pacific Cultural Foundation of Taipei, Taiwan,

deserves high praise for :providing the money necessary to

purchase books and equipment, analyze data, and present

results in a national conference. Special thanks go to

Dolores Jenkins, reference librarian and a dear friend, for

helping me locate and acquire useful survey data. I also
want to thank Dr. Don Nakanishi of UCLA Asian American

Studies Center as well as Rob Cioe, Jill Milbuna, and

Claudia Vaughn of the Los Angeles Times Poll for providing

information of the datasets analyzed here.

Thanks are extended to those mothers and friends--

Martha Samborski, Connie Ayers, Sharon Pence, Eileen Kao,

Ruth Sheng, Maggie Runnels, Laurie Davis, and many many

more--who sacrificed their leisure or otherwise productive

time to volunteer in schools and in field trips so that I

could go to school without much fear or guilt.

Last but not least, my deepest appreciation goes to my

parents, my brother and his family, and my two daring

children, Albert and Alice, for their unfailing faith, love,

and understanding.




LIST OF TABLES . . . . . . . . . .

ABSTRACT. . . . . . . . . . . .


1 INTRODUCTION. . . . . . . . . .

The Current Status: A Two-tiered Picture....
The Puzzle and Some Macro Answers.......
An Alternative Approach: Studying the Individuals

Notes . . . . . . . . . .



Definition of Key Concepts. .......
Ethnicity . . . . . . . .
Asian American (Pan)ethnicity .....
Political Participation ........
Models of Ethnic Political Participation
Ethnic Culture Model .........
Socioeconomic Model ..........
Demographic ]Model ...........
Socio-Psychological Model .......
Legal Constraint Model ........
Summary . . . . . . . . .
Notes . . . . . . . . .

3 METEDDOLOGY ........... ...

Studying Asian American Politics ....
The Need for Survey Data ........
Limitations of Secondary Survey Analysis
Data . . . . . . . . .
Method . . . . . . . . .
Operational Definitions .........
Political Participation ........

. 20
. 20
. 24
. 25
. 28
. 28
. 30
. 33
. 38
. 40
. 43
. 44

. 46
. 47
. 49
. 50
. 53
. 54
. 54

(Pan)ethnicity--Function of Objective Background
. . . . . . . . .56
(Pan ethnicity- -Function of Subj ective Factors. 57
The Structures of Panethnicity ......... 60
Sumary.................... 65


Comparing Asian American to Other Ethnic Groups in
the Aggregate ................ 68
Explaining Asian Participation in a Comparative
Perspective . . . . . . . . 7
Sorting Out Sources of Influence: Comparatively kand6
Internally .. .. .. .. .. .. .. .. 82
Summary.................... 97
Rotes .. .. .. .. .. .. .. .. .. 98


Preliminary Observations from Summary Statistics. 106
The Structure of Korean Immigrant Ethnicity .. 112
Results from Multiple Regression Analysis ... 116
Summairy.................. .. 120
Notes . . . . . . . . . . . 121


The Meaning of Participation Within Asians ... 127
The Meaning of Asian Participation in a Comparative
Perspective .. .. . ... .... 137
Summary.. ................. 142
Notes . . . . . . . . . . . 145

7 CONCLUSION ........ ~........... 147
Notes . . . . . . . . . . . 162


1992 ... .. .. .. . ... ... . 163


. . . . . . . . . 165



REFERENCES................... ... 174

BIOGRAPHICAL SKETCH ................. 187



3-1 Principal Component Analysis of Ethnicity Among
Citizens and All Respondents in the Southern
California Survey. ................62

3-2 Principal Component Analysis of Ethnicity Among
Asian American Citizens in the Southern California
Survey . ... 63

3-3 Principal Component Analysis of Ethnicity Among
All Asian Respondents in the Southern California
Survey . . 64

4-1 Percentage Distributions of Political Participation
and Its Possible Causes Across Ethnic Groups in
Southern California, 1993 . .. .. .. ... 69

4-2 Logistic Regression Estimations of Citizens'
Registration and Voting Participation in Southern
California, 1993 ... .. .. . ... 77

4-3 Logistic Regression Estimations of Participation Other
Than Voting in Southern California, 1993 ....78

4-4 Multiple Regression Estimations of Participation Other
Than Voting in Southern California, 1993 ....79

4-5 Logistic Regression Estimations of Asian American
Citizens' Registration and Voting Participation in
Southern California, 1993 . ... ... .. 86

4-6 Logistic Regression Estimations of Asian American
Participation Other Than Voting in Southern
California, 1993 .. .. .. ... .. .. .. 87

4-7 Multiple Regression Estimations of Asian American
Participation Other Than Voting in Southern
California, 1993 ... .... .. .. .. .. 89

5-1 Percentage Distributions of Political Participation
and Its Possible Causes Among Koreans in Los Angeles,
1992 .. .. ... .. ... .. . 101


5-2 Principal Component Analysis of Ethnicity Among
Koreans in Los Angeles, 1992 .......... 113

5-3 Multiple Regression Estimations of the Political
Participation of Koreans in Los Angeles, 1992. .. 117

5-4 A Simplified Multiple Regression Model of the
Political Participation of Koreans in Los Angeles,
1992 .. .. . .... .. .. .. .. . . 122

6-1 Percentage Differences Between Asian American
Participants and Nonparticipants in terms of
Sociodemographic Background ........... 128

6-2 Percentage Differences Between Asian American
Participants and Nonparticipants in terms of M~inority
Group Experience and Information Level ..... 131

6-3 Percentage Differences Between Asian American
Participants and Nonparticipants in terms of Policy
Preferences and Other Political Orientations .. 134

6-4 Percentage Difference Between Voters and Nonvoters in
terms of Policy and Other Orientations Across Four
Ethnic Groups . . . . . . . . . 138


Abstract of Dissertation Presented to the Graduate School
of the Uhiversity of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

Pei-te Lien

May 1995

Chairperson: M. Margaret Conway
Major Department: Political Science

Defying the image of success, the participation of
Asian Americans in the U.S. electoral process is extremely

limited. What explains the participation patterns of

individuals with Asian ancestry in voting and election-

related activities other than voting? Using two sets of

recent survey data collected in southern California, the

study examines the meanings of being Asian in political

participation at three levels--across panethnic groups,
inside the multi-ethnic Asian group, and within a specific

Asian nationality group. The results confirm previous

findings that, compared to other major ethnic groups, Asian

ethnicity as indicated by objective culture background

depresses participation. Despite controls over several sets
of factors commonly related to the political participation

of mainstream and minority groups, being Asian means less

politically active. Yet, within the Asian sample, one's
national origin usually has no independent impact on the

likelihood or extent of participation.

When the meaning of being Asian is measured with

subjective socio-psychological factors underlying the
construction of ethnic identity, the results reveal a

different dimension of the relationship between ethnicity

and political participation. The concept of ethnicity cast
either at the panethnic or specific nationality group level,

first of all, involves a multi-faceted process that can be

manifested in such components as group consciousness,

cultural and social integration, and ethnic attachment.

Second, being more identified with the panethnic or specific

Asian group can increase participation. Although

insufficient to compensate for the participation disparity

between Asians and Anglo whites, indicators of subjective

identity are most useful of all models to explain turnout or

other participation among Asians. Third, for a group of

foreign-born Korean Americans, however, length of stay best

predicts political integration. For Koreans, the different

impact of being a victim of hate crimes in the two surveys
conducted before and after the Rodney King riot also

highlights the importance of socio-political context in

shaping the impact of ethnicity on participation.
The study concludes by discussing the meaning of voting

participation for Asians as well as the roles of political

parties, interest groups, and media in the mobilization of
Asians into American politics.


'The Current Status: A ?two-Tiered Picture

Asian Pacific Americans (Asians hereafter) have a long

history of immigration to the United States which not only

does not dissipate over time but has accelerated in recent

decades due to changes in U. S. immigration laws and

economic, social, and political conditions within and

outside of the country (Reimers 1985; Chan 1991; Hing 1993).

The 1990 Census provides one of the strongest evidences of

the Asian emergence; the Asian Pacific population rose from
3.5 million to 7.3 million over the decade, with about 71%

of the growth coming from immigration. The Asian rate of
increase in the decennial census--108% or about twice that

of Latinos, six times that of blacks, and 20 times that of

non-Hispanic whites--made Asians the fastest growing

minority group between 1980 and 1990. In five states

(Hawaii, Maine, New Hampshire, North Dakota, and Vermont),

the population of Asians has equaled or outnumbered Latinos
and blacks combined. And it has been projected that, if the

current rate of growth holds--which is a likely scenario

short of a major change in immigration policy and global

political economy, Asians will comprise about one-tenth of
the nation's population by 2050. In an electoral democracy

where numbers count, the explosive increase in group size

may promise a great boost to the group's political leverage.
Moreover, the phenomenal growth in population has been

accompanied by what looks like a remarkably high level of

socio-economic achievement. For instance, in March 1991,

among persons of 25 and over, the percentage of Asians with
four or more years of college (40%) was almost twice the

proportion for whites (23%); their median family income of

$42,250 was $5,300 higher than that earned by all of the
white families (Bennett 1992). Most studies on American

political participation often emphasize the defining impact
of socioeconomic status--especially the role of education

(Ver~ba and Nie 1972; Milbrath and Goel 1977; Wolfinger and

Rosenstone 1980; Conway 1991a). Judging from the high

scores in education and family income, Americans of Asian

origin would be expected to have high rates of participation
in American politics.
In California, a state which accounts for 40% of the

nation's Asians, and in Los Angeles County and its vicinity,

home to the largest settlement of Asians in the mainland,

the story of boom and prosperity is familiar. From 1980 to

1990, the state's Asian population grew from 1.3 million to
2.8 million (+127%), and it is projected that by 2020,
Asians in California will number 8.5 million or be about 20%

of the state's population ((144 and Hee 1993). The

percentage of persons 25 or over that have at least a

bachelor's degree was higher for Asians (34%) than for non-

Hispanic whites (28%) in 1990. The Asian median family
income ($39,769) in 1989 was slightly higher than that for

the non-Hispanic whites ($39,032). In Los Angeles County,

home to one-third of the state's Asian population, Asians

again surpassed all whites in the percentage of those with

higher education or family income.2 And some have not been

keeping the wealth just for themselves.
A few studies reported that the Asian American

community has been donating funds to political campaigns in

much greater proportion than their numbers. Lew (1987)

observed that whereas Asian Americans might constitute for

one-tenth of the population in California, they often

contributed to about 20-30% of a candidate's campaign fund.

This figure was much higher for contributions made to Asian

candidates. Tachibana (1986) reported, for example, that

75% of March Fong Eu's campaign money and 70% of Delaware

Lieutenant Governor S. B. Woo's fund-raising coffer came

from Asian contributors. Because of the demonstrated

financial strength, major political candidates now campaign

in Asian American neighborhoods, headlining Asian American

concerns (Espiritu 1992, 61-65). And the financial

contributions received by the national campaigns have been

remarkable. Nakanishi (1991) reported that, in 1988, both

George Bush and Michael Dukakis received a total of over $10
million dollars from the Asian American community, making it

second only to the American Jewish community in terms of the

amount of campaign money raised by an ethnic or minority

These figures help to support the "model minority"

stereotype established first by the media in the wake of the

Civil Rights Era when the nation was enjoying a relative

economic prosperity.3 Yet the current state of Asian

Americans is a controversial issue. There is no question

that the status of Asians has drastically improved since the

19th century and the earlier part of this century when

formal and blatant discrimination and exploitation were the

norm of practice (Kitano and Daniels 1988; Chan 1991; Feagin

and Feagin 1993). Relative to other minority or even

majority groups, Asians in the aggregate have clearly

experienced a remarkable degree of economic success ((hag and
Hee 1994). There is also little doubt that the election and

appointment of Asians to federal, state, and local positions
in recent years have become less novel and they have become

viable participants in many small mainland cities such as

Gardena and Monterey Park in California, Mesa in Arizona,

West Windsor in New Jersey, or even, to a much lesser

extent, major urban areas such as Seattle, San Francisco,

Los Angeles, Houston, and New York (Nakanishi 1985-1986;

Tachibana 1986; Erie and Brackman 1993; and numerous reports
in ethnic media such as Asian Week and World Journal). If

naturalization is an indicator of one's desire to become

integrated into the American system, scholars concur that

Asians at the aggregate have the strongest commitment to do

so. In the last three decades, Asian immigrants petitioned

to become American citizens much earlier and at a higher

rate than their counterparts from any other part of the

world (Barkan 1983; Portes and M~ozo 1985; Portes and Rumbaut

1990). And, to speed up the new citizens' process of

involvement in electoral politics, community organizations

and political parties have launched voter registration
drives and hired Asian recruiters in major urban areas in

California, Illinois, New York, and Texas (Espiritu 1992).

But those who expect to see a corresponding growth of Asian

participation in the American political power structure have
so far been mostly disappointed.

In terms of officeholding, except in Hawaii,4 there

have only been a handful of Asians--mainly of Chinese or

Japanese descent and from California--who were able to hold

elected positions at higher levels of governments (Espiritu

1992). Among the nine Asian members in the current U.S.

Congress, for instance, three are from California. Results
of the 1990 Bureau of Census Civilian Labor Force data

(Equal Employment Opportunity File) revealed that only 1.4%

of all legislators surveyed in the nation were Asians.s At

the local level, a 1987 Census of Governments reported that

less than .5% of all elected officials were Asians (1988).

An analysis using the 1991 "Form of Government" Survey

indicated that only .2% of the nation's city council members

or mayors are of Asian ethnicity (Mad~anus and Bullock

1993). Federal appointed offices have also been
inaccessible. In early 1990s, only 45 of the 8,200 staff

positions in Congress were held by Asians (Feagin and Feagin

In California, after the mid-1960s and throughout the

1970s, there were usually two or three Asians seated in the

state legislature. This did not happen in the entire decade

of the 1980s, leaving Secretary of State March Fong Eu

(1974-1994) the lone Asian in an elected position at the
state level until the election of Nao Taksugi to the State

Assembly in 1992.6 Similar fluctuation existed at the local

level. For instance, a strong Chinese American incumbent on

the city council of Monterey Park (pop. 60,000, 57% Asian)
was defeated in his reelection bid in April 1994, along with

two other aspirants of Chinese origin, amidst reported fear

of an Asian majority in the city council where two of the

five seats have been served by Asians since 1992.7 In early

1990s, in a state where one out of ten residents were

Asians, only 2% of the state's top elected officials could

claim their origin as Asians; only three out of the state's

53-member delegation to Congress were Asian; and only 1% of

city council and school board seats were held by Asians
(Efron 1990). Even the highest ratio of political

representation found in the 1990 Census EEO File--4.4% of

all legislators in California--pales when compared to the

group's 9.6% share of population in the state.
The lack of elected or appointed Asian officials is

mirrored by the low levels of voter registration and turnout

by the Asian public. A 1984 analysis of the voter

registration lists for three areas of high Asian
concentration in San Francisco revealed that people with

Chinese and Japanese American surnames registered to vote at
far lower levels (31% and 37%, respectively) than the 60%

registration rate found among the general electorate (Din
1984). Similar findings were reported in a Los Angeles

study where 43% of all Japanese were registered, followed by
36% of Chinese, 27% of Filipinos, 17% of Asian Indians, 13%

of Koreans, and 4% of Vietnamese (Nakanishi 1985-1986). In

the city of Monterey Park, Asian Americans' overall rate of

registration was 29% in 1984 and 39% in 1989; the increase
was mostly accounted for by new Chinese registrants

(Nakanishi 1991). Across the state, a California ethnicity

survey found that Asians registered at 55% and voted at 48%
in 1984; these rates were about 30% lower than those for

non-Hispanic whites and blacks (Chlaner, Cain, and Kiewiet

1989). In June 1990 primary, Asian Americans' registration
rate of 39% was found to be the lowest among the four major

groups in California (Field Institute 1990). In the 1992

presidential election, a Los Angeles Times exit poll
indicated that the Asian percentage of the state-wide vote

share was a dismal 3%, despite the group's 7% share of adult

citizens. The situation appeared not to change much in the

1993 mayoral race of Los Angeles where the Democratic

candidate was an Asian (]Michael Woo). Despite their 11%

share of the city's voting-age population, Asians accounted

for only 4% of the votes (Skelton 1993).

Perhaps by far the most accurate estimation of the

Asian electoral participation is the figures reported for

the first time by Cu~rrent Population Survey (CPS) for the

1992 national election (A~ppendix A). Of the entire Asian

population age 18 or over in the nation, only 31% were

registered to vote and 27% actually turned out at the polls.

A major reason for the lack of participation may be that

only about half of the adult population were citizens. When
citizenship status is taken into account, Asian registration

increased to 57% and voting to 50% among eligible adults.

There was still a 13-14% participation gap between Asians
and whites.

Even in a small area of Asian politics--donating money

to political campaigns--where participation may not be
lacking, there is a deficit in the return for Asians.

Although Asians have been found to contribute money in

greater proportion to their share of the population, the
reactions to this emerging view of Asians as the new

political moneybags of American politics have not been
totally positive, particularly when considering the absence

of the types of political benefits and goods--be it greater

access or more high-level decision-making appointments--that

were sought after or promised in the campaigns (Nakanishi

1991).8 Besides, the higher amount of money donated by

Asians as a whole does not necessarily indicate that more

Asians participated in making political contributions.

Comparing the rate of political involvement between Japanese
Americans in a California sample and the general population

in a national sample, Fugita and O'Brien (1991) found that

the rate for Japanese Americans was 26% higher than that for

the general public. Yet, Ulhlaner and her associates (1989)
found that Asian respondents in California did not

contribute money at a higher rate than their non-Hispanic

white counterparts in the 1984 campaign.g

In sum, although progress has been made over the years,

observers noted that "Asian American politics has remained

to an unusual degree 'politics by other means,' i.e., not

direct electoral representation but indirect access through

campaign contributions, lobbying, litigation, and protest"

(Erie and Brackman 1993, 47). And this "asymmetrical

participation" in campaign contribution has not produced

very efficacious results in terms of public office holding.
The Puzzle and Some Macro Answer

The preceding review of the current state of Asian

Americans presents a two-tiered picture. Although Asian

Americans have experienced an explosive increase in


population and have enjoyed the highest levels of education,
family income, and naturalization rate among ethnic groups
in recent decades, the participation of Asian Americans

either in government or in the electoral process has been
extremely limited. Despite some evidence of the Asian gain
in resources and leverage in recent years, the return in

offices held is comparatively low and it remains a puzzle

why the political participation of Asian Americans is very
low in relation to their socioeconomic achievement and

population share.
Numerous explanations, mostly at the aggregate level,

have been proposed for the participation deficit. Some
noted the history of exclusion and discrimination

experienced by Asian immigrants regardless of their national

origins and time of arrival (Daniels 1988; Takaki 1989; Chan
1991; Feagin and Feagin 1993; Hing 1993). A few mentioned

the general antipathy toward government which often grew out
of the unpleasant or fearful experiences dealing with

corrupt regimes in home countries (Nakanishi 1991; Skelton

1993) or the continued interference of homeland government

and politics in the Asian American community (IKim 1981;
Kwong 1987; Chang 1988; Portes and Rumbaut 1990; Wang 1991).
Some emphasized the cultural impediment factor that the
Buddhist-Confusianist values of hierarchy, reverence for

authority, resignation, and passivity are antidemocratic
civic traditions discouraging the participation of Chinese,


Japanese, and Korean Americans in democratic partisan

politics (Sue and Sue 1971; Kitano 1976). Others attributed

low participation to the legal restraints in the electoral

system via the numerous practices of minority vote dilution.
These include the requirements of geographic compactness and

bloc voting component in redistricting and in accessing a

foreign language ballot (Bai 1991; Kwoh and Hui 1993;
Ancheta and Imahara 1993). Still others mentioned the labor

market segmentation which is responsible for the

underemployment, underpayment, and a lack of upward

occupational mobility of group members (Suzuki 1977; Kwong
1987; Light and Bonacich 1988). In his account of the Asian
American Movement, Wei (1993) offered another set of

reasons--the lack of a nationally known leadership, unified

ideology, or even a plan of action.

Most basically, the participation potential of Asians

may be severely discounted by certain demographic
characteristics unique to an emergent multi-ethnic minority

group. First, despite the rapid growth of Asian population
in recent decades, the total number continues to be small

and accounts for only 3% of the entire U.S. population.

Second, the presence of a large stock of the foreign-born

and the young means that members are likely to have limited

proficiency in English and incomplete information about the

political system and democratic processes, and would

possibly not be eligible for citizenship and/or using the


voting privilege for a long while (Rothenberg 1989). Third,

the population is heavily concentrated in the Western region
and mostly dispersed in a number of metropolitan areas where

the costs of living tends to be higher and foreign-born

immigrants and other minority groups also congregate.
Fourth, the population is fragmented as well across

generations, nationalities, religions, languages, and social
classes. In a political system where numbers count, Asians

(and their minority neighbors) are disadvantaged by their
relative small size, limited experiences, and tremendous

internal diversity.

Yet, it is unclear how seriously small population size

or low population share impedes Asian participation. For

unlike other minority groups, there exists a lack of

correspondence between Asian American population size/share

and political participation or representation. Asian
Americans who have been elected often come from districts

with a very small percentage of Asian population. A case in

point is the election of Michael Woo to the Los Angeles city
council from a district that was only 5% Asian in 1985.

Another example is the election of Norman Mineta to the U.S.

House of Representatives from a district with 2.5% Asian in

1975. On the other hand, municipalities having high Asian

concentration such as Monterey Park and San Francisco do not

have a proportion of high officials that comes close to

resemble proportional representation. In the same vein,


Erie and Brackman (1993) found that homogeneous precincts

with higher percentage of Asians did not have a larger

turnout rate among Asians.

Many in the community also contend that the image of
socioeconomic success is more myth than reality, that

economic deprivation, racism, and nativism persist, and that

the gilded image only serves to increase the tension between
Asians and other minority groups as well as within the Asian

community (e.g., Suzuki 1977; Hurh and Kim 1989; Lee 1989;

Takaki 1989; Feagin and Feagin 1993; Ong and Hee 1994).

They claim that the use of median family income and the like
as indicators of wealth may inflate the value of income and

mask the lack of economic participation of the group. A

closer look into information presented in Bennett (1992)

supports this assertion in several ways: 1) Nineteen percent
of Asian families had three or more earners, compared to

fourteen percent of white families. 2) Seventy-four percent

of Asian families consisted of three or more persons as

compared to fifty-seven percent for whites. 3) The high
median income may not hold true for some Asian groups such

as Samoans, Guamanians, and Vietnamese. As a consequence,

the per capital income of Asians in 1990 ($13,420) was about
$2,000 lower than that for whites and a larger proportion of

the Asian population (11%) was below poverty. The higher

tendency for Asians to reside in the Western region and

metropolitan areas may further discount the value of income


for many Asian families. Moreover, there was continued

evidence of under employment and lower educational returns.

In 1990, the proportion of Asian males having 4 or more

years of college in the executive, administrative, and

managerial occupations (23%) was significantly smaller than
that for their white counterparts (31%). The median

earnings of full-time college-educated Asian workers

($34,470) was also lower than that for whites ($36,130).
Perhaps because of the economic segmentation, past research

has not been able to find a strong relationship between

aggregated socioeconomic measures and registration
(Nakanishi 1985-1986).
In sum, an overview of the existing literature using

macro-level analyses suggests that the participation deficit

of the Asians may be attributed more to historical,

cultural, homeland political, legal, economical, and group

organizational factors, and less so to objective
socioeconomic class or demographic indicators. The puzzle

lingers on: Are sociodemographic factors not useful to

predict Asian participation or are they less valid

predictors for Asians than for other groups? Should we
revise our understanding of the defining role of

sociodemographic characteristics in political participation?

WThat better explains the political participation of Asians?


An Alternative Approach: Studying the Individuals
Whereas each of these macro-level theories offers some

ins ig~ht into the roots of the Asilan under-participat ion, and
the conceptualization of the current study certainly

benefits from these predecessors, few give any estimate of

the magnitude of the deficit as compared to other ethnic

groupss. No one has been able to compare in a systematic

way the efficacy of one theory to the others) nor to

explain individual differences in the extent and the
likelihood of participation. Without controlling for

alternative explanations, causality is often difficult to

attribute. And mnamy, by focusing on structural constraints,

seem to explain better the lack of rather than the incidence

of participation. Although qualitative data gathered by

studies using in-depth interviews, participant-observations,

community power, and ethnography provide great insight into

the dynamics of participation in a designated setting, the

narratives are limited to the context under study and cannot

be used to make generalizations beyond the subjects chosen.

Using the macro-level approach is therefore insufficient to
answer the questions proposed above.

An individual-level survey-based approach as adopted by

this study, by contrast, is able to answer questions such

as: Does being an Asian American matter for participation?
and How? What is the role of socioeconomic status and other

factors for Asians as compared to other ethnic groups in


assessing the impact of ethnicity on participation? What

explains the participation patterns within the pan-Asian as
well as the respective nationality group? These are

questions raised in political participation research in the
United States. Chapter 2 reviews five of the major theories

of political participation in the discipline and evaluates
the efficacy of each in the studying of Asian participation.

Key concepts such as ethnicity and political participation
are defined and hypotheses proposed.

One major reason past research has not been focusing on

individuals is the lack of data, particularly large-scale

comparative survey data. The one exception is the
California Ethnicity Survey collected in 1984 by Uhlaner and

others (1989) on which this student has conducted extensive

research (Lien, 1992, 1993, 1994). In light of this

research vacuum, a proposal to launch a large-scale in-

language'" national survey of Asians in America may be

highly appealing. However, the costs involved can be very

high.n An alternative approach, secondary analysis of
surveys previously collected for other uses, is therefore

proposed for this study. An extensive search into major
social science data archives and research institutes in the

U.S. for opinion surveys with a significant number of Asian

American respondents produced four polls collected by the

Los Angeles Times in southern California counties between
1989 and 1993. ?Two of them ("Asians in Southern California,


1993" and "Koreans in Los Angeles, 1992") are used in the

analyses .

Chapter 3 describes the data sets used as well as the

operational definitions of key concepts. Justifications and
limitations of secondary analysis and of survey data are

discussed. A methodological review on the studying of Asian

American politics is also included.

The results of the Southern California study are

reported in Chapter 4. Additional methodological issues and
results of the Korean study are the focus of Chapter 5.

Chapter 6 puts the question of participation deficit in

perspective by asking whether participation matters for
Asians as a group and as compared to other ethnic groups in

the political environment. Differences between
voters /part icipant s and non-voters /-part ic ipants in terms of

sociodemographic outlook, minority group experiences,

political information, policy preferences, and other

political orientations are explored. Comparisons are also
made between Asian and non-Asian voters in terms of the

distributions and distances of their policy and political

Chapter 7 concludes the dissertation by first
revisiting the research questions and summarizing major

findings. It then ponders on the implications of the "non-

significant" findings for models of Asian American political

participation. Recognizing the central role of the

mobilization context, the study ends by discussing some of

the movers and shakers in the on-going Asian American

Movement such as political parties, ethnic community

organizations, and the ethnic media.

1. The census' use of the term "whites" generally includes
persons of Hispanic origin. According to a March 1992
Current Population Survey (CPS), about one tenth of whites
are of Hispanic origin. Although there is a lot of
contention in the Hispanic community about the adequacy and
the choice of a panethnic group label, the term "Latinos" is
used in this study to refer to this diverse ethnic group.

2. Compared to Anglo whites, the percentage of Asians 25 or
older having four or more years of college education (37.2%)
also exceeded that of whites (30.5%). However, the median
income for Asian families in 1989 ($39,296) was lower than
that for Anglos ($41,222).

3. The term first appeared in a New York Times' article in
1966 to describe the success of Japanese Americans. An
analysis of media portrayals of Asian Americans in recent
decades revealed some change of direction. Asian American
issues of crime, school dropout, and poverty were gradually
incorporated into the popular discourse in the 1980s.
However, there was a continued reliance on a culturally-
based explanations of success which stressed Asian American
educational achievement and supported the 'model minority'
thesis (Osajima 1988).

4. The situation in the state of Hawaii is an exception
because even since the beginning of statehood in 1959,
congressional seats in both the U.S. Senate and House have
been dominated by Japanese Americans who were elected by a
population with nearly two-thirds originating from Asia or
Pacific Islands. But the political power of other Asians is
also in the rise. In the mid-term election of 1994, it was
a Filipino American, Ben Cayetano, who filled the Governor's

5. Because of the small population size, results on Asians
from this government- sponsored survey or from any other
public or private survey can only be treated as suggestive.
Large-scale government surveys such as those reported by the
Current Population Survey (CPS) have a much smaller sampling
error than any other surveys.

6. The absence of elected representatives does not
necessarily mean that Asians were totally excluded from
state politics. The Office of Asian/Pacific Affairs (1987-
1991) created by President pro Tem David Roberti in the
California Senate served, along with other Asian Pacific
legislative aides, many of the same functions performed by
black and Latino representatives for their ethnic
constituents (Syer and Culver 1992).

7. Another reason for this as well as many other local
defeats may be that having more than one Asian candidate
running at the same time for the same office can split the
community's votes and resources. Espiritu (1992) cited some
instances of this in the city council elections in 1991.
She perceived this as a potential threat to solidarity of
the pan-Asian community.
8. An author commented that this "asymmetrical
participation" has started to change in recent years after a
group of Chinese Americans formed a bipartisan Interim
Coordinating Committee for Chinese Americans (ICCCA) to ask
for better value of their money (Wei 1993).

9. The rate for Asians and W~hites was 18% and 20%,
respectively. However, the rate for Asian citizens was 24%.
There was also a large difference in terms of participatory
rate among respondents of the five major Asian groups

10. Interviews conducted in the preferred languages of the
respondents rather than in English only may be very much
needed for an internally-diverse ethnic group with a large
proportion of non-English speakers.
11. According to an experienced Asian American pollster, the
estimated cost for a 30-minute telephone survey of 1,000
respondents nationwide randomly selected from a list of
common Asian surnames and interviewed either in English or
their home-country language is at least US$45,000.


Quite a few sets of theories exist in the current

literature to explain the political participation of

Americans in general. Theories on the participation of

ethnic minority groups are few and often target blacks. In

this chapter, five models of ethnic group political

participation are proposed and their relevance to explaining

the political participation of Asian Americans are examined.

Special attention will be paid to the concept of ethnicity
and the roles of socioeconomic status (SES) and socio-

psychological factors in the shaping of a (pan)ethnic group
identity in Asian America.
Definitions of Key Concepts

Before the relationship between ethnicity and political

participation can be meaningfully discussed and analyzed, it
is necessary to define the meanings of ethnicity, Asian

American (pan)ethnicity, and political participation as

adopted by this study.
Ethnicity. Ethnicity can be defined as a sense of

belonging to "an involuntary group of people who share the

same culture" or are perceived by others as sharing the same

culture (Isajiw 1974, 122). Expressions of ethnicity for

minority groups are complex and always occur against a


backdrop of at least two levels of identification--with

one's own ethnic group and with the dominant group (Yinger

1985; Hutnik 1986). For groups in change, the double

boundary is often maintained from within by the

socialization process and from without by the process of

intergroup relations (Barth 1969; Isajiw 1974). For groups

that have a recent history of international migration and

are experiencing rapid changes in their composition and

socio-political position, the concept of culture, however,

refers much less to an unmediated heritage than to socially

constructed boundaries which can be created and re-created

to unite group members (Roosens 1989). Immigrant group

ethnicity, therefore, is an "emergent phenomenon" rather

than a static construct (Yancey, Erickson, and Juliani

1976). Far from being an essence or something fixed,

concrete, or objective, ethnic/racial identity is formed

through the interaction between subjective identification

and objective conditions and can be constantly transformed

by political conflicts (Omi and Winant 1986).
However, ethnicity may be derived not from one

distinctive, integrated culture but from a multiplicity of

cultures coercively lumped together under one supranational

group label, such as the inclusion of Mexicans, Cubans, and
Columbians under the umbrella term, Latinos. These

designations discount class, national, and generational

cleavages. ThI~e equivalent of the term "ethnicity" as is


used to classify blacks or Jews then may be "panethnicity"

or "the generalization of solidarity among ethnic subgroups"

(Espiritu 1992, 6). According to Espiritu, previously
unrelated and marginalized ethnic groups, thrown together at

first by ignorant or insidious panethnic categorization and

later by racial violence, confront the meanings of the

imposed pan-group identity and the deprived group status,
and unite together to protect and promote collective

interests. The result of this process--variously called

"ethnicization" (Sarna 1978), "racialization" (Omi and

Winant 1986), or "ethnic Americanization" (Fuchs 1990)--is

the forging of a multi-tiered, situational, and partly

ascribed panethnic culture. Although summarily called

ethnicity in this study, distinctions between pan-group

ethnicity (panethnicity) and sub-group ethnicity

(nationality) will be made when necessary.
Dn the other hand, ethnicization is also a process of

building up a sense of national identity with the host

country (Garcia 1987; Finifter and Finifter 1989) where the

extent of identification as being an American is at least as

important as being an ethnic minority. The term
"assimilation" or the characterization of immigrants'

responses to the host environment, however, is highly
inadequate to describe the process of becoming Americans

(Feagin and Feagin 1993). Because of differences in

experiences of discrimination and stereotyping, unique group


history, and political/economic structure such as the

presence of urban machines, many noted that the assimilation

experience of European immigrants cannot be transferred to

immigrants from other parts of the world (e.g., Pachon 1985;
Fuchs 1990; Hero 1992).

Beginning with Gordon's (1964) notion that the

adaptation of non-Anglo-Saxon-Protestants may take place in
a number of stages, many scholars perceive the

Americanization of immigrants as an endless and dialectical

process of acculturation (Parenti 1967; Hurh 1980; Keyes
1981; Padilla 1985; Keefe and Padilla 1987; Kitano and

Daniels 1988; Waters 1990). In this process, immigrants may

have adopted certain cultural patterns in public domain but

have maintained a distinct subculture in private domain

(Keefe and Padilla 1987; Hutchison 1988). They may also
have any one of the four combinations of high/low ethnic

identity and high/low acculturation (Huxda 1980; Kitano and

Daniels 1988). The persistence of ethnic culture, according

to Portes and Rumbaut (1991), "has been the rule among

immigrants, old and new, and represents simultaneously a
central part of their process of political incorporation"

(141). They may also become culturally but not

psychologically or structurally adapted to the new identity

(Yinger 1985). Instead of predicting assimilation or the
eventual adoption of a white American identity and the

complete detachment from the ethnic culture over time, this


multi-dimensional concept of ethnicity allows one the

freedom to maintain ethnic loyalty at one level and to

become acculturated to the new identity at another level.

Asian American (pan)ethnicity. Although immigrants

from different Asian nations coming to America at different

stages of history have a shared experience of discrimination
and have long been received with panethnic terms such as
"Asiatic," "Oriental," and "Mongolian," the development of

an unified ethnic group identity or panethnicity among Asian

Americans has a short history. The term "Asian American"

was introduced only in the late 1960s by a group of mainly

native-born Chinese and Japanese college students and later

by professionals in the community-based human service

organizations in a movement aiming to fight for racial

equality, social justice, and political empowerment (Wei
1993). It was a product of the convergence of at least
three forces: internal demographic changes, the anti-Vietnam

War protests as well as the Black Power and other New Left
movements, and the emergence of pan-Asian organizations such

as Asian American Political Alliance (Espiritu 1992).

Despite the inclusive label, Asians seldom think of
themselves as a single people. They often identify

themselves as people from a certain Asian country or even a

certain district or region within a country. Thus, pan-

Asianism has been observed to be primarily the ideology of

native-born, American-educated, and middle-class Asians.


Nevertheless, as indicated above, the construction of a

culture is often not voluntary and ethnicity is multi-

dimensional, contextual, and flexible in nature. Hence,

(pan)ethnicity is not able to be fully indexed or explained

by a respondent's self-identification with the pan-Asian or

a specific Asian ethnic group. Rather, components of
ethnicity can be revealed when multiple indicators measuring

such dimensions as acculturation, ethnic attachment, and

group consciousness are used. More discussion on the layers
of ethnicity can be found in the chapter (3) which

delineates operational definitions. Suffice it to say that

at least when the right context emerges, e.g., the

prevalence of racial violence or the urgent need for social
service funding or to overcome the arbitrary and

inconsistent census classification of persons of Asian

origin, pan-Asian group consciousness can be raised and

political action mobilized (Espiritu 1992).
Political participation. Scholars disagree on the

meaning of "political participation" (Conway 1991b). In the

first chapter, political participation was defined to

include not only actions by individuals to influence the
selection and/or the actions of government officials (Ver~ba

and Nie 1972) but also the outcome of this participation in

the sharing of governing. The purpose is to give a more

inclusive picture about the political status of Asian

America--the distance between political actions and the goal


of full empowerment. For the individual-level analysis

which is the main thrust of this study, the definition

offered by Milbrath and Goel (1977) which covers individual

actions and/or attitudes both to influence and to support

government and politics seems to be more appropriate to
describe the participation process for persons with recent

immigration background. Examples of participation in the
surveys used include citizenship intent and naturalization

for the foreign-borns, voter registration and voting for

citizens, and, for all members of the society, making

campaign contributions, contacting officials, attending

political meetings or fund.-raisers, and volunteering for a

political cause.
Some studies have found political participation to be

multidimensional (Ver~ba and Nie 1972; Milbrath and Goel

1977; Bobo and Gilliam 1990). Because of the different

degrees of demand on information and motivation for each

type of activities, those who vote often do not share the
same level of involvement as those who join with a group or

organization to solve community problems, or work for

political campaigns, or contact elected officials. However,
some note that, at least in the past 20 years, participants

tended to overlap in activities that required the same kinds
of resources but not to the extent of clustering in

identifiable "modes" (Rosenstone and Hansen 1993). This is

supported by a study on Asian and Mexican Americans where


voting and other conventional activities can all be loaded
into one factor (Lien 1994). Thus, the dimensionality of

participation remains an open question.
Voting is often observed to be a blunt indicator of the

overall satisfaction and involvement of the public.

However, it also provides an equality no other types of

participation can afford: each citizen gets one and only one
vote (Verba et al. 1993a). For the majority of Americans,

voting requires few resources (registration and information)

and is easily practiced (Teixeira 1992). Yet, for colored

minorities, the promise of voting equality did not come

close to being carried out until the passing of the Voting

Rights Act in 1965 and its amendment on non-English ballots
in 1975. Before the installment of these legal guarantees,

citizens of Asian origin, like citizens of African or

Hispanic origins, had to overcome the barriers of English-

only elections, literacy tests, racial gerrymandering, and

physical intimidation and violence to register to vote (Wei
1993). For individuals with a recent history of

immigration, there is an additional "cost" to this most
common form of participation--the acquisition of

citizenship, which is itself a process most likely

influenced by proximity to the mother country, fear of

officials from Immigration and Naturalization Service, lack

of information and knowledge, difficulty in meeting language

and civics requirements, and a general lack of a sense of


political efficacy and trust in political institutions of
the mother country where socialization was initiated (Fuchs

1990). Other barriers that increase the costs of

participation for recent immigrant groups include the

regional ly- di spersed and geographical ly- concentrated
distribution of the population, the high proportions of the

young and the new, along with the institutionalized

practices of minority vote dilution (Pachon 1985). Many of
these and other determinants of participation are discussed
in the next section.

Models of Ethnic Political Participation

Ethnic Culture Model

To explain the distinctive patterns of political

participation of ethnic Americans, a common approach in
survey-based research is to attribute the patterns to a

composite ethnic culture variable as denoted by one's self-
identified or ascribed race, language, religion, or national

origin. Assessment of the independent impact of the ethnic
factor is then achieved by adding some control over a

respondent's socio-political background. Although this

approach is useful to capture whatever group-related effect
is left unexplained by other quantifiable measures in the

equation, this cultural definition of ethnicity has
increasingly been criticized in recent years as reductionist

and is unable to reflect the evolving or situational nature

of ethnicity (Patterson 1975).


Nonetheless, as discussed above, the meaning of culture

for any ethnic or panethnic minority group includes more

than primordial ties. Particularly for Asian Americans, the

supranational group label was not created spontaneously by

persons of different Asian nationality origins. Instead, it
was the product of the interaction of external and internal

forces which help create and re-create the political

community. Asian American ethnic culture, in other words,

is a panethnic culture. Perhaps because of this unique

property of ethnicity, a past attempt by Uhlaner et al.

(1989) had very limited success in locating a universal

principle that can account for the participation of Asians
as well as of blacks, Latinos, and Anglo whites. It is

possible that the Asian American culture in and of itself
not only may render differently the meanings of the existing

measures of participation for Asians than for other groups,

but that it may not be fully accounted for by existing

measures which often focus on one's current experience in

the United States (Lien 1994).

For this study focusing on the political participation

of Asians in Southern California, it is hypothesized that

the Asian American panethnic culture will again have

significant impact on participation. The significance of

the participation deficit, however, may be reduced when

variables measuring a respondent's sociodemographic

background and political attitudes structuring one's


ethnicity are controlled. Within each Asian

ethnic/nat ionality group, the role of ethnic culture

associated with each home country to explain participation

may be similar to that for the p~an.-Asian group. However,
the meaning of nationality may also draw from a more or less

distinct cultural heritage. Therefore, consistent with

criticisms of primordialism, the impact of ethnic culture as

indicated by one's country of origin may not be a

significant predictor of participation when measured along
with indicators of current status and attitudes (Lien 1992;

1994) .
Socioeconomic Model

The defining impact of socioeconomic indicators--

especially education--has been well-established in the
American political participation literature (e.g., Verba and

Nie 1972; Milbrath and Goel 1977; Wolfinger and Rosenstone

1980; Conway 1991a; Leighley and Nagler 1992; Verba,
Schlozman, Brady, and Nie 1993b). Generally, evidence

supports the idea that citizens of higher social and
economic status participate more in electoral politics.

This is partly because of the comparative absence of

political cleavage along class lines in the U.S. two-party

system (Verba, Nie, and Kim 1978). This is also because
education may impart democratic values and information about

government and politics, nurture a sense of competence and
efficacy which predisposes an individual to political


involvement, and provide facilitating skills to obtain more

information about politics. Income, on the other hand, may

enable the disposal and conversion of wealth into other

resources or the tradeoff for other opportunities and reduce

the costs of participation (Rosenstone and Hansen 1993).

Education, in addition, can increase one's opportunity to be

placed in social networks through employment, organization
membership, or volunteer work (Verba et al. 1993b). Given
that this conclusion is often derived from observing the

white majority, it remains to be seen if it can be applied

to different ethnic groups in America.
When socioeconomic factors are added to assess the

independent impact of ethnicity, many studies have found
that the difference in voting and registration rates between

blacks (or Latinos) and whites decreased or disappeared

(e.g., Milbrath and Goel 1977; Conway 1991a; Teixeira 1992;
Verba et al. 1993b). This indicates that the impact of

class may be more important than ethnicity. However, given

the empirical relationship between SES and ethnicity, the

impact of class may be overemphasized whereas the role of
ethnicity is underplayed. Using five ethnic/nationality

groups in New York, Nelson (1979) found support for his

hypothesis that ethnicity has an independent influence on

participation over and above socioeconomic class.
Decomposing persons of Hispanic origin into three

nationality groups, Calvo and Rosenstone (1989) showed that


the education-turnout relationship holds for Mexicans and

Puerto Ricans, but not for C~ubans. Higher educated Cubans

did not turnout more than their coethnics who did not finish

grammar school. Similarly, studies on Asians using either

aggregated or individual approach found a lack of

relationship between SES and registration or turnout

(Nakanishi 1985-86; Lien 1992). By comparison, Lien (1994)

found that Mexican Americans in the same survey displayed a

very strong relationship between SES and participation in
voting and other types of activities. Because a larger

proportion of both Asians and Cubans are of higher SES than
are Mexicans and Puerto Ricans, it is possible that the

differential in the impact of SES is smaller for higher

status groups than for lower status groups.

Yet, it may also be possible that the influence of

socioeconomic factors for emergent groups such as Asians and

Latinos on participation is weak as compared to the more

established groups such as Anglo whites and blacks. In the

same study by Calvo and Rosenstone, the impact of education

and income was less significant for Hispanics than for non-

Hispanics. The reason for this may be that members may find
it harder to translate their educational achievement into

resources for participation (i.e., information, time, and

money) because of either language difficulty for the
foreign-borns,2 10wer returns for educational attainment in

income and employment, or that there are many more


compelling demands of time for immigrant group members in
their daily struggles to adapt to the adopted land. Second,

past socialization either in an undemocratic homeland or in
a democracy with blatant legal discrimination against

newcomers may impede the development of political efficacy

and a sense of civic duty which were found to increase

participation. Third, the emergent minority status of the

group and the lack of perceived benefits (via representation

in government) may negatively affect the participation

potential for both the native- and foreign- born members.
Demographic Model
In addition to socioeconomic factors, demographic

factors specific to the immigrant population such as

nativity, age, length of stay, and gender may have a

substantial impact on participation.3 These factors are

particularly important for studying the participation of new

immigrant groups, for these are the factors that may have
direct or indirect bearing on the extent of social learning

and integration. As one ages or stays longer in the adopted

land, one may acquire language skills, know more about the
host community, produce a second or more generation, become

eligible for citizenship, and develop political interests in
the adopted society because of the many stakes involved.

Sheer time or the passage of years in chronological

age, commented Converse (1969), only serves as a proxy on
how much exposure an individual has in the political


environment. For foreign-born immigrants, he noted that the

length of stay may be a better indicator of exposure to the
U.S. politics than age. Lee (1980) observed that being

foreign born may imply a lack of experience with the

American political system, the preoccupation with economical

survival, interference from the home government, an

unawareness of the need to assimilate, weak organization

membership, cultural ambivalence, and a fear of social

rejection by the mainstream. However, when length of stay

(or age for the native-born) or percentage of life time

spent in the U.S. (length/age) is used along with nativity,

past studies have found them to be less useful than age to

predict participation (Uhlaner et al. 1989; Uhlaner 1991;
Lien 1994). Gitelman's (1982) study on the resocialization

of Russian immigrants in Israel also noted a lack of linear

relationship between length of stay and political adaptation

except in the early years of arrival.
Part of the reason for the counterintuitive findings

may be that, like the political learning of young adults
(Jennings 1989), there is a certain formative period in the

immigrant's life history when loyalty to a new political

system can be imprinted. This threshold effect appears to
be at work in Black, Neimi, and Powell's (1987) finding of a

positive relationship between age and the acquisition of

partisanship and other indicators of political involvement

among new immigrants to Canada.4 More basically, much of


the relationship between time and participation is made

under the assumption that there is an upward social mobility

over time. Although Dahl's (1961) assimilation theory that

the importance of ethnicity will succumb to class concerns

has been seriously attacked for its lack of validity, his

assertion of a positive association between social mobility

and time in the host nation- -particularly over immigration

generations--has not been seriously challenged. As a matter
of fact, demographic factors, like the socioeconomic

factors, have received substantial support from studying

Latinos because of the relatively larger presence of the

young, the new, and the poor among the present generation

(e.g., de la Garza 1987; Calvo and Rosenstone 1989; Fuchs

1990; Hero 1992; Garcia, Garcia, de la Garza, and Falcon

1992).s Many Asians are similarly young and new but not
necessarily poor or under-educated. It is therefore

hypothesized that demographic variables will have less

impact on the participation of Asians than for other grou~ps.
Further, among the demographic variables for Asians, age may

be more useful than other indicators of time to predict

participation. Yet, among the foreign-born generation,

length of stay may be an even more important determinant of

Gender. The role of gender in ethnic group political

participation has been controversial because of the implied
conflicts between feminism and cultural nationalism.


Critics in both the Chicano and Asian American Movements

have worried that the emphasis on women's identity and

rights would divide the ethnic community in its struggle

against racism ((2xow 1987; Garcia 1989). Yet, research

using either an individual or a community approach has not

provided much support for the hypothesis that feminism hurts
ethnic minority group empowerment. Being minority and

female is often not a liability in political participation.

Instead, particularly for African American women, it may

help to motivate participation in politics.
Perhaps boosted by being conscious of the "dual

oppression" of both sexism and racism, black women typically

participate at higher levels than their male counterparts
(Verba and Nie 1972; Shingles 1981; Baxter and Lansing 1983;

Jennings 1993). Although the gender advantage of Black

female participants disappeared in the turnout of 1984 when

factors such as income, home ownership, and political

interest were controlled, Tate (1993) reported that black

women were more likely to be partisan, interested in

political campaigns, and registered to vote than black men.
The observation on Latina women is more complex. In the

first national survey of Latinos, being a woman of Mexican

or Cuban origin was insignificant to predict turnout but

being a woman of Puerto Rican origin depressed the turnout
in 1988 (Garcia et al. 1992). For Californian Asians in the

election of 1984, gender was insignificant to predict voting


or other types of electoral participation when

sociodemographic and attitudinal factors are controlled

(Lien 1994).

Support for the role of gender in motivating ethnic

political participation is more evident from recent
community studies. In a study of the Latino community in
Boston where Puerto Ricans constitute a substantial portion

of the population, Latina women are found to make up the

majority of the participants and activists at all types of

political events (Hardy-Fanta 1993). Similarly, Saito
(1992) observed that Asian women are involved in politics at

all levels and their success as candidates and community

activists have been crucial to constructing the Asian

political base in Monterey Park, CA. This is so despite the
consistent findings from census studies that more Asian

American women are compelled to involve themselves full-time
in the labor force and receive lower returns to their

education than do white women (Matthaei and Amott 1990;

Chan, 1991; Bennett 1992; Kim and Lewis 1994). A rationale

behind the phenomenon is that, like the Latinas in Pardo's

(1990) study, the dual obligation for the family and work

place may facilitate the developing of social networks and
transform the concern over education and other family issues

to more involvement with community issues and political

action. This, in turn, may compensate for the socialization

bias that encourages submission and passivity. The net


effect in terms of mass politics may be the insignificance

of gender in participation.6
The role of gender in mobilizing Asian participation

can take a downward turn when it intersects with nativity.

For many--especially non-English speaking-- foreign-born

Asian women, their potential for political participation can

at least be triply depressed by their prior socialization

which does not treat women as equal partners, by their new

obligation to join the immigrant labor market to meet the
needs for survival, and by their necessity to adjust to

being members of an American ethnic minority. It is
therefore hypothesized that, though gender may not be

significant in and by itself, being a foreign-born Asian
woman may depress the likelihood to participate in the U.S.

political system.
Socio-Psychological ]Model

The role of socio-psychological factors such as

partisan attachment, sense of civic duty, political
efficacy, interest in politics, trust in government, and
concern over election outcome has asserted its importance on

participation ever since the seminal study by Campbell and
his associates (1960). In terms of comparing participation

across ethnic groups, Verba and Nie (1972) introduced the

concept of group consciousness which was found to compensate

for the disadvantaged SES of blacks in the late 1960s.

Later scholars found that black group consciousness matters


because of its linkage to political efficacy and trust

(Shingles 1981). Extending the examination to cover members

of both subordinate and dominant groups, Miller and his

associates (1981) emphasized that it was the interaction of

both group identification and politicized consciousness

which mobilized participation. Adopting the concept of

Miller et al., Uhlaner and her associates (1989) found that

group consciousness can help explain away the participation

disparity between Anglo whites and Latinos but not between

Angles and Asians. Lien's follow-u~p study (1994) found that
for both Asian and Mexican Americans in California, the

concept of group consciousness is multidimensional, with

each dimension having different ramifications for voting and

other types of participation for each of the two immigrant

group. She found that the addition of racial alienation and

deprivation indicators, along with other indicators of the
ethnicization process such as acculturation and ethnic ties,

improved our understanding of participation for both ethnic

In light of the strong empirical support for socio-

psychological factors, it is likely that a similar multi-
faceted nature of ethnicity can be found in this study and

that measures associated with the shaping of ethnic group

consciousness and identity in America such as personal

experiences of discrimination, concern over group status and
interests, and attitudes or efforts supporting


integrat ion/acculturation may have a pos it ive impact on

participation, whereas indicators of ethnic resilience will
not have a negative impact.

Legal Constraints Model
Unlike conventional legal models that deal with the

differences in registration requirements across states

(Wolfinger and Rosenstone 1980; Caldeira, Patterson, and
Markko 1985; Calvert and Gilchrist 1993; Rosenstone and

Hansen 1993), this discussion of immigrant group

participation examines the effect of two legal prerequisites
for enfranchisement, citizenship status and voter

registration.' In the past, at least before the 1950s, the
lack of political participation of Asians can mostly be

attributed to the discriminatory immigration and

naturalization laws and other legal restraints at federal,

state, and local levels. Among the more infamous were the

Chinese Exclusion Act of 1882, the Alien Land Law of 1913,

and the internment of Japanese Americans in WWII. Between

1879 and 1952, the California constitution prohibited

employment of Chinese by any government entity or

corporation in the state. Until 1943, Chinese were

prohibited from immigration to the U.S. and citizenship was
denied to those already in the country. Naturalization for

immigrants from all countries in Asia was possible only
after the passing of the McCarran-Walter Act of 1952. But

the number eligible for this process was not significant


until after the implementation of the 1965 Immigration and

Naturalization Act when U.S. finally opened its arms to

embrace immigrants from the "Thnird World".

Today, citizenship status is granted automatically to
all those who were born in this country and it is one of the

prerequisites for voting. For those who were born outside
of the country but want to be politically empowered by being

able to vote, they need to go through first a lengthy and

complicated (mad often costly) process of naturalization

that requires legal residence of 5 or more years with at

least 30 months of continuous stay during this period, a

record of good moral character, and successful completion of

exams on the.English language and U.S. history and

government .

If speedy naturalization is an indicator of one's
desire to become integrated, Asians at the aggregate have

the strongest commitment to stay and be incorporated into

the American society. Barkan (1983) found that between 1958

and 1978, despite great variations among nationality groups,

eight out ten foreign-born Asians were naturalized by the

8th year of residence; comparatively, the ratio for non-
Asians was five and a half out of ten. From 1976 to 1986,

Portes and Rambaut (1990) reported that the naturalization

rate of Asians was six times that of Mexicans. But they

also noted great variations among the Asian nationality


groups, particularly that between Vietnamese and Japanese


Studying the pattern of naturalization for 23 countries
in the 1970s, Portes and M~ozo (1985) made a similar

observation and they explained that the high rate of Asian

naturalization may be related to the lack of geographic

proximity to the homeland, the more political than
economical motive of emigration, and the higher educational

or occupational background. In contrast, Pachon (1991)

noted that Latino immigrants had the highest incidence of

noncitizenship because of a general confusion over the real

benefits of citizenship and the process to attain it.

Although political participation other than voting does not

require citizenship status, it is likely that those who
either took pains to become eligible to vote or were born

eligible to participate in basic politics will also be more
attuned to participate in politics other than voting. Yet,

judging from the non-political motivation of recent non-

refugee immigrants from Asia, a more likely scenario is that
citizenship status will not have much impact on

participation (Erie and Brackman 1993).
A second (mo, for almost all of the U.S. born adults,

the only) prerequisite for participation in voting is voter

registration. For most citizens, registration can be
discouraged by the residency requirement, the need to re-

register after moving to a different jurisdiction, or to


file for a change of address if moving within same

city/county. Colored citizens (residing especially in the
South) before 1965 could be prevented from registering

because of discriminatory attitudes and procedures such as

literacy test, all-white primary, poll tax, harassment, and

violence (Conway 1991a). For citizens with a non-English

background, their likelihood of registering can be further
hindered by cultural passivity,g ignorance about the proper

procedures, the lack of foreign-language forms and other
election materials,g and the lack of perceived benefits.

The registration requirement, therefore, can be a major

barrier to voting for immigrant groups. However, past

studies have shown that, like voting, registration has a

class bias--for instance, among those who moved,

registration is less likely for those who have lower levels
of education or political interests (Squire, Wolfinger, and

Glass 1987). Because registration costs are part of the

costs for voting, it is hypothesized that registration will

be determined by similar factors that predict voting.

Moreover, because registration signifies more than a basic

interest in politics, being registered to vote may have a

spill-over impact on the probability and extent of

participation in other kinds of election-related activities.

In this chapter, two key concepts of the study--

ethnicity and political participation--are conceptually


defined. The five sets of factors previously found to be

important to influence ethnic political participation:
ethnic culture, socioeconomic status, demographic

background, socio-psychological attitudes, and legal
constraints are discussed. These five models by no means

exhaust the possible explanation of political participation.

Some studies, for instance, emphasize the importance of

strategic mobilization by political elites (Patterson and

Caldeira 1983; Cox and Munger 1989; Uh~laner 1989; Rosenstone

and Hansen 1993). Some found prior socialization or

behavior to be crucial (Marquette, Green, and Wattier 1991).

Only these five sets of theories are proposed because they
are theoretically important and available for empirical
tests in the datasets used--which are to be discussed in the

next chapter. Also, because each model of participation can

only be thought of as being one small chip in a gigantic
machine, the efficacy of each may be best tested when others

are controlled. Multiple regression analysis, a statistical

method designed for such a purpose, as well as other

methodological issues will be dealt with in the next


1. A potential problem with the author's past (and the
current) investigation on the inter-diversity within Asians
in America is that the sample size for each nationality
group is relatively small--ranging from n=50 for Filipinos
to n=89 for Koreans in the 1984 survey and n=23 for Koreans
to n=57 for Japanese in the 1993 survey. The number of
survey questions suitable to study the issue is also
wanting. Results on the impact of different Asian

nationality are therefore only indicative. However, our
understanding of the impact of one specific Asian ethnicity
(the Korean) may be greatly enhanced by results from a
relatively large survey on Koreans in Los Angeles (n=750)
which are reported in chapter 5.

2. In the census publication on foreign-born population in
1990 (1993), about half of the Asian immigrants reported
that they could not speak English well (49.9%) and about 1/3
reported living in a linguistically isolated household
(30.3%) .

3. Citizenship status is also a very important factor.
However, for the purpose of analysis, we delay the
discussion until in the legal model.

4. Alternatively, threshold effect or period effect may be
involved. The effect of time may also be curvilinear, but
we do not have a longitudinal data set to test these
hypotheses .
5. Support for the influence of demographic variables was in
fact mixed in the Latino National Political Survey (Garcia
et al. 1992). Whereas immigration generation had positive
effect on the Mexican and Puerto Rican American turnout in
1988, the impact of age was insignificant when other factors
were held constant. And neither immigration nor age could
predict turnout for the Cuban Americans.
6. Part of the reason for the lack of effect may be that
many individual level studies did not control for working
women outside the home, who may be more likely to
participate than those who work at home.
7. This legal model is excluded in the analyses using the
survey on Koreans because citizenship status and
registration are the dependent variables.
8. The primacy of the cultural factor has been seriously
challenged by some recent scholars of Asian American and
Latino politics.

9. Although Asians are a designated minority that is
protected by the foreign-language provision of the 1975
Voting Rights amendment which requires the printing of
foreign language ballots and related materials if 5% of the
population uses the language, the law is more discriminatory
to Asians because of the multi-lingual and residentially-
dispersed characteristics of the population. A recent
change of criterion to 10,000 people in a jurisdiction is
expected to alleviate some of the problem.


Studying Asian American Politics
As indicated in the first chapter, the majority of

studies on Asian Americans are at the aggregate level and

provide little quantitative data. In reviewing research
dealing with Asian American politics, Nakanishi (1985-1986,

2) lamented that "there is an extreme paucity of analysis"
and that whatt exists is largely confined to descriptive

and historical inquiries" which can be broken down into four

areas: 1) discriminatory political actors and policies in

the American society; 2) responses by Asian immigrants to
both discrimination in the U.S. and events in their

homelands; 3) case studies of the actions of Asian

Americans; and 4) community and organizational studies

exploring the internal dynamics among competing groups. An

important reason for this may be that the history of Asian

participation in electoral politics that emphasizes numbers
is very short. Although Asians in America have a long,

though mostly subdued and unrecognized, tradition of

engaging in political action, Chan (1991) noted that they
never did so on a publicly visible scale until the 1960s.

With a rapidly growing population and an increase in

visibility and influence in American politics, a constant


flow of journalistic articles on Asian American political

participation started to appear in a certain ethnic media

(e.g., Asian Week, KoreAm Journal, and many English or
vernacular language newspapers) and, to a lesser extent,

mainstream media (e.g., Los Angeles Times and New York

Times). Academic works on Asians in U.S. electoral politics

also started to accumulate, especially in the last few

years. Beginning with the publication of an anthology based
mainly on papers delivered in three conventions in 1978 (CRD

1980), there have been at least two dissertations (Din 1984;

Saito 19 92 ), a few specially- commissioned reports (Cain and

Kiewiet 1986; Nakanishi 1986; Eckrich, Lew, and Treisman

1987; MJuratsuchi 1991; Erie and Brackman 1993), a number of

single-authored or edited book chapters (e.g., chapter 9 of
Chan 1991; chapter 2 of Jackson and Preston 1991; chapter 3

of Espiritu 1992; chapter 5 of Hing 1993; chapter 8 of Wei

1993; chapter 7 of Fong 1994; chapter 8 of Ong, Bonacich,
and Cheng 1994), and finally, a journal Asian American

Public Policy Review (1990- ) to complement the lone and

interdisciplinary Amerasia Journal (1971- ).
The Need for Survey Data

However, studies adopting the micro approach to Asian

participation are rare and have generally been conducted at
the local or state level in California. The main obstacle

is the lack of data. To this student's knowledge, during

the entire decade of the 1980s, only one set of survey data


(thlaner et al. 1989) was available for systematic
examination of differences in various aspects of

participation across Asian and other ethnic/minority groups.

But perhaps because of the emphasis on the comparative

perspective on political participation, the highly valuable
data set omitted many factors specific to the immigrant

group community that might influence participation (Lien
1992; 1994). The reliance on the regional telephone

directory of one nationality group (1984 Korean Telephone

Directory of Southern California) to yield more than one-

fourth of the state's Asian sample was a source of potential

bias. The information revealed through the survey may also

be dated because it was collected about a decade ago.

The problem with the lack of data, though acute, is

probably understandable, given the pan-ethnic group's

comparatively small size, racial\1ingual diversity, the
extremely dispersed residential pattern, and the many

constraints to design a cost-efficient and representative

survey even within a small geographic region. By this, it

also means that Asian respondents, in a few occasions where

their ethnicity could be determined by their self-

identification as belonging to the Asian race, were often

systematically underrepresented in multi-group surveys. One

example of this was in the California statewide exit polls
conducted by the Field Institute in 1992--of the 8,170

questioned, less than 300 were Asians. Even in a huge


national level survey such as the ABC/CNN exit polls of the

1992 general election (N=15,490), only 156 Asians were

In designing a survey that focuses on Asian Americans,

scholars will need to make tough choices to balance the

demands for efficiency and coverage; and either can be very

costly, if the size of the sample is also a concern. Yet,

the cost of not having more individual level data on

political participation may be even higher. We will not be
able to know, for instance:
What is the likelihood that an Asian American will
register and actually vote in elections? What factors--be
they social class, level of educational attainment,
ethnicity, generation, sex, occupation, or religion--have
the greatest influence on an individual's likelihood to
register, affiliate with a specific party, or become
involved in other activities, such as contributing campaign
funds, or seeking public office? Which issues are most
likely to gain Asian American voters? (Nakanishi 1985-1986,

These are explanatory questions about electoral reality to

which "[s]urveys, with all of their limitations, constitute

the most direct, and thus most valid, way of finding

answers (]Dennis 1991, 52).

Limitations of Secondary Survey Analysis

A compromise between the need for collecting survey

data and the lofty costs involved is secondary analysis.

Although the research method is increasingly popular,

particularly in times of economic fluctuations, secondary
survey analysis sometimes is impossible because of the

unavailability of data, the mismatch between primary and


secondary objectives, the unduly long delivery time, poor

data quality, errors made in the original survey, and--as

illustrated above--the number of cases in a specific

sulbpopulation is too small to conduct the desired
statistical analysis (Kiecolt and Nathan 1985). In

addition, whatever problems intrinsic to the survey method

are applicable to secondary analysis, too. According to

Babble (1989), one of the complaints of the survey approach

is that it can fragment the complexities of life into

discrete, clean-cut, and unitary variables. The other is

that it is superficial and can seldom deal with the context

of social life. Neither can surveys measure social action--

only self-reports of recalled, hypothetical, or prospective
action. Surveys may also suffer from the lack of validity

because of the need for standardization. Fortunately, some

of the problems with surveys can be partly offset through

sophisticated analyses, while most of the problems dealing
with secondary analysis were overcome because of locating

two Southern California surveys conducted by the Los Angeles

Times Poll in 1992 and 1993.


Dne of the data sets on which this study is based was

drawn from an August 1993 Los Angeles Times (:UAT) survey of

adults residing in one of the following six counties in

southern California: Los Angeles, Orange, San Diego, San

Bernadino, Riverside, and Ventura. Random-digit dialing


techniques were used to produce the base sample. Asians and

African Americans were selected by oversampling telephone

numbers of persons identified as Asians or blacks in LAT

polls conducted within the previous two years. The margin
of sampling error for the total sample is plus or minus 3%.

The margin of error for Asians is plus or minus 8%.

Telephone interviews using English or Spanish language were

completed with 221 Asians, 199 Latinos, 144 blacks, and 646

(Anglo) whites.2 About half (55%) of the entire sample and
three-fourths (72%) of the Asian sample were from the Los

Angeles city and county area.

Compared to the 1990 Census, the proportion of Los

Angeles County residents in the 6-county area was about the
same as in the sample (52%), but the proportion of Asians in

the sample was 10% higher than that in the census (62%). As

is expected with any survey data, respondents' levels of

education, income, and the extent of political participation

(citizenship, registration, and voting) were higher and the

share of foreign-borns was lower than in the census

\p~pendix B). However, the share of high family income in
the Southern California sample was about the same as in the

census of Los Angeles County; and the rate of voting among

citizens was only slightly higher in the Southern California

sample than that in the national sample reported by census.
A second data set used by the study was drawn from a

Los Angeles Times survey of adult Korean residents of Los


Angeles County between February 26 and March 27 of 1992.
Random-digit dialing techniques were used to produce about

one-third of the sample, the rest were drawn from lists of

Korean-surnamed households countywide. Telephone interviews

using Korean (93%) and English were completed with 750
Koreans. Translation and interviewing were conducted by

Interviewing Services of America. The margin of error for

the sample is plus or minus 5%.

Compared to the 1990 census, the percentage of foreign-
borns (99%) was overrepresented by 17% whereas the

percentage of citizens (36%) was underrepresented by 9% in
the Los Angeles County (A~ppendix C). The percentage of

Koreans with a bachelor's degree or higher (49%) was much

higher than what was reported (34%) for the census; but it
was comparable to the figure reported for the sample of
Asians in Southern California. Moreover, in a pattern

consistent with the Asian data mentioned above, the share of

high income families in the Korean sample was about the same
as in the census. Yet, the percentage of Koreans registered

either among all members or among citizens was much lower in

this sample than in the samples for Asians shown in Appendix

Although the two Los Angeles Times polls are area

surveys, they are preferable to previously collected data of

a similar nature in that a popular variant of probability

sampling technique--random-digit dialing--was incorporated,


enabling the presentation of more representative results and

the calculation of sampling errors for this extremely under-

researched group.' Another advantage of the polls is that

they contain a fairly large amount of information on
race/ethnic relations and other political attitudes. The

major drawback of using the two polls, as is true with any
secondary analysis, is the lack of a perfect match between

proposed research topics and surveyed items. For example,
omitted in the polls are a direct measure of panethnicity as

well as indicators of the extent of political party

affiliation and pre-immigration political socialization.

The limit to examining only the California communities also

impairs the generalizability of the study to other parts of
the nation. The extent of bias for Asians resulting from

using these two sets of data has been discussed earlier.

The major purpose of the study is to explore the

relationship between political participation and various

socio-political factors of Asians as compared to other

ethnic groups as well as within the pan-ethnic group. The

goal is to test hypotheses derived from the five alternative
models discussed in Chapter (2) and find out whether Asian

ethnicity matters, how it matters, and what accounts for the

political participation of the immigrant group. In the

process, three levels of analysis will be involved. At the
univariate level, the operational definitions of dependent


variables and independent variables will be introduced,

followed by a discussion of their frequency distributions.

At the bi-variate level, the relationships of participation

with basic sociodemographic and socio-psychological factors

will be discussed. At the multi-variate level, a number of

explanations for ethnic political participation will compete
with each other among respondents of all groups and of the

Asian group alone. For the Southern California data,

logistic regression, a statistical method for handling
dichotomous dependent variables, is used to analyze the

likelihood of citizenship/naturalization, registration,

voting, and whether one participated in any one of the other

political activities. Multiple regression is used to

analyze the extent of participation in activities other than
voting. For the Korean data, because of the absence of

voting and other participation indicators and the presence

of a fuller range of ethnicity indicators, attention will be

paid to the development of ethnicity and its causal
influence on citizenship, citizenship intent, and voter

registration. Possibilities of using structural equation
models that describe causal relations among latent variables

and include coefficients for endogenous variables will be

Operational Definitions

Political participation. As indicated in Chapter 2,

the definition of political participation in this study


refers to the conventional, election-related activities

that private persons use to show their allegiance to the

political system. Examples of such activities in both data
sets include naturalization for the foreign-borns and voter

registration for citizens. For the Southern California

sample, participation activities also include citizen

voting, and, for all members of the society, making campaign

contributions, contacting officials, attending political

meetings or fund-raisers, and volunteering for a political

cause. For the Korean sample, the intent to become
naturalized is conceived to be another indicator of

participation. Question wording for political participation
and other variables is reported in Appendix D.
Because the Southern California data include an

extensive list of participation items used by previous

scholars to simulate the structure of participation (e.g.,

Verba and Nie 1972; Milbrath and Goel 1977; Bobo and Gilliam

1990), several principal component analyses using varimax
rotation were conducted to determine the dimensionality of

participation for this sample. The results for all the four

groups of respondents and the Asian respondents indicated
that those who registered and voted were not necessarily

more likely to participate in other types of activities. On
the other hand, those who contacted officials were also more

likely to contribute money, attend political functions, or

volunteer for a political cause.4 A summed index of


participation other than voting was created by taking the
summed score of the four items of participation mentioned

above. In addition, a dummy indicator of participation was

also calculated by assigning a value of 1 to those who

indicated their participation in any of the four activities.

For the Korean immigrants in Los Angeles, their level

of participation into the American political system may be
conceived as comprised in an ascending order of citizenship

intent, citizenship, and voter registration, with each

succeeding act demanding more of the newcomers than the

previous one. Because of the non-conflictive nature of
these participation items, the term "political integration"

is sometimes used to refer to the participation of Koreans

in later chapters. Principal component results using
varimax rotation indicated that the three acts could all be

loaded into one dimension." A four-point scale of

participation was created by assigning a value of 3 to those
who registered to vote, a value of 2 to those who were

naturalized but not registered, a value of 1 to those who

were not naturalized but expected to become one in the next

few years, and a value of 0 to those who had done nothing.
(Pan)ethnicity--Function of Objective background. For
the Southern California data, a respondent's ethnic

background is measured by his/her answer to the question
about racial origin. Dummay variables for black, Latino, and

Asian ethnicity are created to allow comparisons with the


impact of being white. For the Asians, a respondent's
nationality background is determined by his/her response to

the question about Asian country of origin. Examination of
the effect of nationality within Asians is conducted by

first limiting the respondents to those belonging to the

five major nationality group in the sample and then creating

durmmy variables for each of five groups except Chinese.

(Pan)ethnicity--Function of Subjective factors. In
concluding her widely-acclaimed book on the institutional

aspect of Asian American panethnicity, Espiritu (1992, 168)
remarked that "[a]n important next step would be to quantify

this consciousness by studying interpersonal pan-Asian

ethnicity--most important, its marriage patterns." Although

intermarriage is the most studied form of

assimilation/ada~ptation, Williams and Ortega (1990)

summarized six other types of indicators basing on Gordon's

theory (1964; 1978): culture (language, cultural practices),

primary-group structure (neighborhood, friendship,

organization membership), ethnic group identification

(importance of ethnic background), attitude reception

(perceived prejudice), behavior reception (experience of
discrimination), and issue opinion (conformity to the

maj ority opinion) For studying Asian participation, it
seems that what of urgent need then is not only a formal

measure on the state of (pan)ethnic group consciousness, but

the possible sources of fusion (English proficiency, common


experiences of prejudice and discrimination, organizational
influence--exposure to ethnic/nonethnic media, affiliation

with political parties and other social organizations) and

fission (attachment to ethnic group culture, concern for

homeland politics, etc.) .

The two data sets do not contain a direct measure of

(pan)ethnicity, but they do hold quite a number of socio-

psychological factors relating to the construction of

(pan)ethnicity. For the Southern California sample,

possible indicators of ethnicization include those

approximate the extent of group consciousness such as in
domains of attitudinal reception (i.e., the perceptions of

Dwn pan-ethnic group being the most deprived, of own pan-
ethnic group as having fewer opportunities to get adequate

housing, education, or jobs, and of racial discrimination as

an important problem in the community) and behavior

reception (i.e., the personal experiences of being
discriminated against and of being verbally or physically

abused because of one's panethnic background).

Ethnicization can also be measured by indicators of

structural (i.e., having cross-racial friendship) and

marital integration (i.e., not opposing to inter-

racial/ethnic marriage).
For Asians in Southern California, there are a few more

possible indicators of ethnicity formation: 1) the frequency
of hearing racial slurs about the pan-ethnic group, 2)


naming of at least one of the following--i.e., M~ichael Woo,

March Fong Eu, Jay Kim, and Daniel Inouye--as a prominent

group political leader, 3) knowing the length of pan-group

immigration history, 4) knowing the unfair deprivation of
one sub-ethnic group in the past (internment of the Japanese

in World War II) and support for remedial actions (awarding

reparation). Dummy variables were created for these eleven
indicators except those measuring the degrees of concern

with racial discrimination in community, personal experience

of discrimination, and hearing of racial slurs against
Asians. These ordinal measures are scored from 1 to 4 with

the highest score assigned to the strongest degree of

concern or the most frequent experiences of discrimination.

Similarly, the formation of ethnicity for Koreans in

Los Angeles is conceived to consist of a number of

adaptation stages, which do not necessarily occur in any
chronological order. Group consciousness attributable to

behavior reception is measured by one's personal experiences

of discrimination and hate crimes; that associated with

attitudinal reception is indicated by the perceptions of own

group condition being worse off than other minorities and of
racism being the primary barrier holding Koreans back.

Marital integration for Koreans is assessed by one's degree

of approving inter-ethnic/-racial marriage" and structural

integration by the incidence of having non-Korean friends.


However, a number of items exist exclusively in the

1992 survey of Koreans that can serve as additional

indicators of the complex process of immigrant

ethnicization. Besides cross-ethnic friendship, possible

measures of structural integration include a dummy variable

indicating one's membership in a church organization as well

as two ordinal measures--the conducting of business with

Koreans and/or non-Koreans and the frequency of speaking

with white persons in a week. Korean American acculturation

can be estimated by the extent of using English and/or

Korean languages) and media in everyday life and the level

of proficiency in English. The strength of identifying with
one's ethnic background can be gauged by the perceived

importance of Koreatown and the preservation of Korean
culture for future generations in Los Angeles as well as by

one's residing in Korean neighborhood and the expectation to

return to Korea sometime in the future. In a range of

scores (i.e., 1 to 4 or 1 to 5), higher scores are assigned

to those responses indicating greater support of

intermarriage and greater ability and likelihood to use

English, to speak with white persons, to do business with
non-Koreans, and to value Korean culture and Koreatown.

The Structures of Panethnicity

Previous studies have found that the concept of

(pan)ethnicity or (pma)ethnic identity is far from
monolithic. Although the precise configuration of the


process is still under dispute, the few recent studies that

empirically examine the concept tend to differentiate among
acculturation, ethnic attachment, and group consciousness or

the levels of perceived prejudice and discrimination (Keefe

and Padilla 1987; Williams and Ortega 1990; Lien 1994). To

find out if a multidimensional structure of :panethnicity

exists in the two data sets, several principal component

factor analyses were run using an oblique rotation.'

Results for the Southern California sample are reported in
Tables 3-1 to 3-3. To avoid confusion with the somewhat

different composition of ethnicity for a specific subgroup,

results for the Korean sample are reported in chapter 5.

For both citizens and all respondents in the Southern

California sample, the concept of panethnicity has three

dimensions: personal experience of racial discrimination,

concern over a troubled group status, and acculturation

through interracial friendship or marriage. For the Asian

subsample, four items specific to the group are added to the

analysis--know group history, know group leaders, secure

group interest, and hear racial slurs against Asians. The
concept of Asian American panethnicity is found to consist

of one additional dimension: group deprivation, which is

part of the group concern factor for the four panethnic

groups and it is negatively related to other dimensions for
Asians. Another part of the group concern factor in the

entire sample, racial problem, is now closely related to the

Table 3-1
Principal Component Analysis of Ethnicity Among Citizens
and All Respondents in the Southern California Survey

(A) Among Citizens

Oblique Rotate Componentsa 99
Variable I II III Communalityb

Personal Discrimination
Victim of Hate Crime .772 .637
Being Discriminated .751 .317 .600
Concern over Group Status
Perceive Group Condition Bad .710 .574
Racial Discrim. a Problem .376 .621 .462
Own Group Most Deprived .385 .518 -.363 .505

Acculturation/Racial Integration
Cross-racial Friendship .775 .642
Interracial Marriage .607 .440

Eigenvalue 1.74 1.00 1.12
Variance (%) 24.9 14.3 16.0O

(B) Among All Respondents

Obique Rotate Componentsa N113
Variable I II III Communalityb

Personal Discrimination
Victim of Hate Crime -.821 .683
Being Discriminated -.700 .387 .586

Concern over Group Status
Own Group M~ost Deprived .657 .525
Racial Discrim. a Problem .640 .475
Group Condition Bad -.322 .580 .415
Accul turation
Cross-racial Friendship .740 .644
Interracial Marriage .688 .540

Eigenvalue 1.01 1.70 1.16
Variance (%) 14.5 24 .2 16.6

-.30 are reported. b A communality in factor analysis shows
how much variance of an observed variable is accounted for
by the common factor. It is calculated by summing the
squared factor loadings of a variable.


Table 3-2
Principal Component Analysis of Ethnicity Among Asian
American Citizens in the Southern California Survey

Oblique Rotated components" (=31
Variable I II III rV Comrmunalityb

Personal Discrimination
Victim of Hate Crime .651 .445
Being Discriminated .618 -.354 .482
Hear Racial Slurs .607 .503

Racial Discrim a Prob. .599 .448 .550

Concern over G'roup Status
Know Group Leaders .643 .501
Perceive Group Condition Bad .617 .541
Secure Group Interest .612 .397

Group Deprivation
Perceive Own Group Most Deprived -.782 .645

Know Group History -.691 .577

Accul turation
Interracial Marriage .768 .623
Cross-racial Friendship .333 .742 .647

Eigenvalue 2.06 1.47 1.26 1.11
Variance (%) 18.8 13.3 11.5 10.1

Note: (see Table 3-1)

Table 3-3
Principal Component Analysis of Ethnicity Among All
Asian Respondents in the Southern California Survey

Oblique Rotated Componentsa (N=180)
Variable I II III IV Communalityb

Personal Discrimination
Victim of Hate Crime .721 .525
Hear Racial Slurs .707 .614
Racial Discrim a Prob. .576 -.427 .544

C'ancern over Group Status
Know Group Leaders .666 .462
Cross-racial Friendship .632 .518
Secure Group Interest .595 -.357 .470
Perceive Group Condition Bad .518 -.388 .434

Group Depriva tion
Know Group History -.721 .581

Perceive Own Group M~ost Deprived -.681 .574
Being Discriminated .478 -.482 .460
Accul turation
Interracial Marriage .824 .710

Eigenvalue 2.05 1.58 1.16 1.09
Variance (%) 18.6 14 .4 10 .6 9.9

Note : (see Table 3-1)


personal experiences of discrimination for Asians.
There are also some differences between the results

excluding and including noncitizens in terms of the order or

the composition of items in a certain factors. Between

Table 3-2 and Table 3-3, for instance, while the experience

of discrimination loads with the personal factor for Asian

citizens, it has a closer relationship to the group

deprivation factor for all Asian respondents. Similarly,

Asian citizens' having close friends of other races is an

indicator of acculturation, but it loads with the concern

over group status when noncitizens are included. Between

parts A and B of Table 3-1, the sign of the personal
discrimination factor is reversed--indicating that,

different from the case with citizens, what accounts for the

experience of discrimination for all respondents is opposite

to the underlying reasons shaping acculturation and group

concern. Because the reliability coefficient testing

additivity of items in each dimension of panethnicity is no

higher than .55, only those variables that have the highest
factor score on each dimension were entered in the

regression analyses for the Southern California sample.s


This chapter opens with a brief review on the general

approaches to the studying of Asian American politics which
often fall into the descriptive or historical category,

followed by the justifications for the need for a survey

approach. The two datasets used in this micro-level study

are introduced, along with the operational definitions of

key concepts. The structure of political participation and

ethnicity for respondents in the Southern California sample

is examined through several principal component analyses.

In the following chapter, indicators of primordial and

constructed panethnicity joins socioeconomic, demographic,

and legal factors to explain the extent and incidence of

Asian American political participation within the pan-ethnic

group and as compared to other groups in this microcosm of
Southern California.


1. The problem of the small sample size in surveys is not
limited to the Asians. In the same exit poll, the
percentage of Latino respondents is 2.3%--a percentage much
lower than their 9.5% share of the nation's population.

2. The use of "white" in the sample is short for "non-
Hispanic white", which is used interchangeably with "Anglo"
in this study. This differs from census reports where,
unless otherwise specified, "white" includes those with a
Hispanic origin.

3. Because the use of weights will significantly reduce the
size of the Asian subsample in one survey, weighted results
will only be reported for the survey on Koreans.

4. The respective factor score for each activity in the
Southern California sample is as follows: Among all
respondents, contact officials =.649, attend functions
=.641, donate money =.638, political volunteer =.598,
register =.725, and vote =.631. The Eigenvalue for
participation other than voting is 2.37 with 39.5% of the
variance explained and a standardized 4-item alpha of .70.
Among Asians, contact officials =.767, attend functions
=.727, donate money =.706, political volunteer =.631,
register =.833, and vote =.824. The Eigenvalue for
participation other than voting is 2.35 with 39.2% of the
variance explained and a standardized 4-item alpha of .71.

5. The factor score for each of the indicators is as
follows: citizenship intent =.616, naturalization =.819, and
registration =.644. The Eigenvalue is 2.08 with 69.3 of the
variance explained.

6. Another measure of marriage assimilation is whether a
respondent ever married a person of another ethnicity or
race. This is not possible here because in this Korean
sample almost every married respondent (96% among all) had a
Korean spouse. Besides, lacking information on the place
and time of marriage, it is unrealistic to estimate
outmarriage for a new immigrant group if most of the
marriages took place in the homeland.

7. Oblique rotation is a conservative test of
dimensiona lity. Principal components results using varimax
rotation which assumes independence among variables,
however, produce very similar structures.

8. The indicators of ethnicity used in each set of
regression analyses and the dimension of ethnicity each
represents are as follows: For citizen respondents in the
entire sample, victim of hate crime (personal
discrimination), perceive group condition bad (concern over
group status), cross-racial friendship
(acculturation/integration). For all respondents in the
entire sample, interracial marriage replaces cross-racial
friendship on the acculturation dimension; the rest are the
same as for citizens. For citizen respondents in the Asian
subsample, victim of hate crime (personal discrimination),
know group leaders (concern over group status), perceive own
group most deprived (cpaxqp deprivation), and interracial
marriage (acculturation). For all respondents in the Asian
subsample, knowledge of group history replaces perception of
own group being most deprived on the group deprivation
dimension; the rest are the same as for Asian citizens.


Comp~aring Asian American to Other Ethnic Grou~ps
in the Aqqregate

The extent of participation for the Southern California

sample is reported in Section I of Table 4-1. At the

aggregate, the frequency results generally support the

impression that Asians (mul Latinos) participate at a rate

much lower than whites or blacks--except that the

naturalization rate of foreign-born Asians (42%) is much

higher than that for Latinos (23%). However, when

citizenship is controlled, the participatory disparity is

significantly reduced--though it does not disappear--for

registration and voting. Among the registered, Latinc

voting rate is as high as the rate for blacks and whites,

but Asians still vote less. The rates of Asian

participation in other activities (including campaign

donations) are also low when compared to whites' (:but higher

than Latinos'), and citizenship status does not account for

the difference.

A description of the sociodemographic background of the

respondents can be found in Section II of Table 4-1. Not

surprisingly, Asians are overrepresented in higher

educational and income categories, exceeding even the levels


Table 4-1
Percentage Distributions of Political Participation and Its
Possible Causes Across Ethnic Groups in Southern California,






43 (64)





Voter Regristration
Registered 47(68)*
-among registered-
--Democrats 39
--Republicans 38
--Other Party 7
--No Party 16
N= 150

Vot ing
Voted in 1992 37(53)
(among registered) 78
--Clinton (D) 46
--Bush (R) 44
--Perot 10
N= 100








Participation Other Than Voting
Participated 24(27) 16(18)
--Contact Official 11(11) 9(13)
-- Cont ribut e$ 12 (14 ) 5 (6)
--Attend Meeting 8(10) 5(5)
--Volunteer 9(10) 5(5)
N= (varies)


0-7th grade
8th grade
9-11 grade 6(4)
12th grade 10 (8)
Technical training 3(3)
Some college 30(33)
College degree 31(31)
Some graduate work 8(9)
Graduate degree 12 (12)
N= 218(153)

23 (23)
3 (2)



Table 4-1--continued




14 (17)

Lat ino

24 (16)

4 (4)
2 (2)

18 (21)



Family Income
Less than $10K
$60K or more

Work full-time
Work part-time
Look for work
Not look for work










in Southern California
5(6) 2(0)
15(7) 11(5)
20(15) 12 (8)
16(17) 11(8)
11(12) 9(5)
14(20) 19(24)
19(26) 37(51)
20(154) 199(133)

Length of Residence
Fewer than 2yrs
2-5 yrs
6-10 yrs
11-15 yrs
16-20 yrs
Whole life


Immigration Generation
First generation 52(42)
Second generation 34 (38)
Third or more gen. 14 (20)
N= 220(154)



Table 4-1--continued

Asian Latino Black Anglo
Female 48(54) 48(49) 62 53
N= 220(154) 199(133) 144 646

A. Group Consciousness

Dwn Grou~p Most Deprived
Asian 18(15) 8(5) 3 8
Lat ino 37 (40) 63 (54) 56 33
Black 35(38) 42(46) 78 25
Anglo 2(3) 6(6) 7 16
N= 221(154) 199(133) 144 646

Group Condition Bad
--percentages of those who think conditions are bad or very
Asian 11(10) 16(19) 10 16
Lat ino 49 (48) 43 (41) 50 43
Black 45(47) 46(45) 67 47
Anglo 4(4) 8(10) 8 18
N= (various)

Racial Discrimination A Problem
Maj or 9 (7) 12 (9) 26 9
Moderate 25(27) 34 (28) 25 28
Minor 43(41) 32(37) 27 37
No 24 (25) 22 (27) 22 26
N= 213(149) 190(127) 141 606

Personal Exp~erience of Discrimination
Great deal 3(5) 5(3) 13 4
Fair 10(11) 9(9) 16 8
Some 51(51) 36(35) 47 31
None 37(33) 50 (52) 25 58
N= 221(154) 196(130) 143 639

Ways Discriminated (amonlg those being discriminated)
Jobs 20(20) 27(21) 42 20
Education 8(6) 7(8) 8 6
Housing 3(3) 6(6) 19 2
Government 10 (8) 5(5) 3 7
Business 18 (22) 8(8) 29 8
Ne ighbor 7 (7) 4 (4) 2 3
Stranger 30(34) 14(14) 15 11
N= (various)

Know Internment of Japanese Americans
Know 89(90) 85(86) 91 94
N= 187(136) 160(110) 118 609

Support Rep~aration

B. Acculturation/Integration

Cross-Racial Friendship
Asian 31(38) 34 46
Lat ino 57 (64) 50 67
Black 54(62) 52(60) 61
Anglo 68(71) 35(36) 58
Any Group 85 (88) 72 (78) 68 85
N= (various)

Table 4-1--continued

Victim of Hate Crime
Victimized 18(17)
N= 217(150)

Lat ino






Hear Racial Slurs About Asians
Very often 12 (11) 13 (14) 6 5
Fairly oft 14 (13) 21(19) 12 10
Fairly infreq 26(28) 23(22) 26 28
Very infreq 48(48) 44 (46) 56 57
N= 216(151) 195(133) 139 643

Know Asian American Political Leaders
Name one+ 38(42) 22 (29) 24 26
Mike Woo 25(25) 21(26) 25 16
D. Inouye 9(13) 4(5) 2 11
Connie Chung 9 (11) 9(7) 7 6
N= 221(154) 199(133) 144 646

Know Asian American History
100+ years 58(63) 43(52)
N= 202(142) 162(122)



194 (141)




Favor strg
Favor somwt
Oppo sorrst
Oppo strg

Table 4-1--continued

Asian Latino Black Anglo
Interracial Marriage
Appr strongly 14 (14) 12 (13) 16 12
Appr somewhat 7(7) 7(6) 4 8
Not care 70(69) 70 (73) 75 59
Disapp somewhat 6(6) 5(4) 2 12
Disapp strongly 3(3) 6(5) 3 9
N= 207(147) 187(126) 140 611

Source: The Los Angeles Times Poll #318, August 7-10, 1993,
released through the Roper Center for Public Opinion
*Entries in parentheses refer to the percentages among


for non-Hispanic whites. However, about one out of ten

Asians--the highest ratio of the four ethnic groups--also

reported being unemployed at the time of interviewing.
Both the Asian and Latino groups were clearly younger than

blacks or whites. About two-thirds of the respondents in

the two recent immigrant groups reported as younger than 40;

in contrast, those over 65 occupy a relatively small

proportion in each of the two subsamples. In terms of the

length of stay in the Southern California community, a

higher percentage of Asians arrived within the last fifteen

years than any other group, while over half of the

respondents in the three non-Asian groups had lived in the
area for at least twenty years. A similar trend is evident

in the distribution of immigration generation across the

groups. About half of all Asians were born outside of the
United States, whereas at least two-thirds of blacks and

whites had U.S.-born parents. The distribution of Latinos

across generations is more even, with slightly over one-

third of the subsample in each of the first two generations.

As to those socio-psychological indicators of

ethnicity, Asians, like their white counterparts, were quite

positive about their experiences in Southern California.
Few Asians thought their own group suffers the most

discrimination (18%) or had fewer opportunity to get

adequate housing, education, or jobs (11%). Most of them
did not think racial discrimination was a serious problem in

the community (67%), though about two-thirds of them had


experienced at least some degree of personal discrimination,

primarily from dealing with strangers (30%). Less than one-
fifth of Asians (18%) reported having been victimized

because of one's ethnicity, but a higher percentage of

Asians (26%) heard racial slurs about Asians in

interpersonal communications.

The development of ethnic group identity can also be

indicated by one's knowledge about group immigration history

and its prominent political leaders. About four out of ten

Asians could name at least one :prominent group political

leader. Michael Woo was the most recognized personality;

others included U.S. Senator Daniel Inouye, March Fong Eu,

and U.S. Representative Jay Kim. About six out of ten

Asians knew the correct length of Asian immigration (58%),

but a greater percentage of Anglos knew the Asian history

(66%). One exemplary instance of the deprived Asian group

status is the internment of Japanese Americans during the

World War II. About 9 out of 10 persons of Asian origin

knew this historical event and almost an equal percentage of

Asians favored awarding reparation payments to those who
were interned.

Another factor influencing the development of an ethnic

group ident ity in America i s acculturat ion/adaptat ion. Thi s
is when members of an ethnic group develop their sense of

belongingness to the American community. ?Two of the stages

in the adaptation of immigrant groups to the American

society are structural and marital integration. Based on


these two accounts, Asians have a very high rate of social

integration. The percentage of those persons reporting
having a close friend of another race is as high as the

Angles' (85%); and only one out of ten Asians opposes the
idea of having someone in the family marrying a person of a

different racial or ethnic background.

Exp~laining Asian Participation in a Comparative Perspective
Our first question is whether Asian ethnicity matters

for political participation. The logistic regression
estimations of citizens' voting turnout and all respondents'

participation other than voting across four ethnic groups
are reported in Tables 4-2 and 4-3. As shown in column I,

both Latinos and Asians can be predicted to turnout or

participate less than whites simply by their self-identified
ethnic group identity. Although blacks voted at a rate

equal to whites, they tended to participate less than
whites, too. Similar impact of the ethnic culture factor on

the extent of other participation is observed in the

multiple regression results in Table 4-4. Thus, when ethnic

group culture is considered alone, Asian (pan)ethnicity, as
well as Latino or even African American (pan)ethnicity, does

matter for (the lack of) political participation.

To account for the participatory disparity between the

white majority and the ethnic minorities, we first turn to
socioeconomic factors. Results in column II of Tables 4-2,

4-3, and 4-4 indicate that, although education and income
differences are crucial in determining turnout as well as

Table 4-2
Logistic Regression Estimations of Citizens' Registration
and Voting Participation in Southern California, 1993

Models I II III IV Regi strat ion

Note: The dependent variable in models I-IV is scored 1 if
the respondent reported voting for the president in 1992.
Numerical entries are logistic coefficients except where
noted. Standard errors are in parentheses.
* p<.05 ** p<.01 *** p<.001

(1( Ethnic C'ulture
Constant 1.13*** -1.61***
(.10) (.33)
Black -.22 .10
(.22) (.23)
Latino -.85*** -.54*
(.21) (.22)
Asian -1.02*** -1.33***
(.19) (.21)
(.2) Socioeconomic Status
Education .36***












Unemployed -.L
(3) Demographic



(4) Socio-psychological
Personal Discrimination

Concern over Group Status

Ac cul turat ion/ Int egrat ion





Initial -2 Log Likelihood
At Converg. 1146 1057
% Correct 69.17 72.41



Table 4-3
Logistic Regression Estimations of Participation Other Than
Voting in Southern California, 1993 (N=1094)

Models I II III rV v

Note: (s~ee Table 4-2). The dependent variable is scored 1
if the respondent engages in any political activity other
than voting and 0 otherwise.
a The partial coefficient for citizenship is taken from a
separate model. Compared to other coefficients in the model
using registration (column V), the significance of Asian
ethnicity increases to p=.0018 and age has a borderline
impact (p=.0523), the rest are no different in either model.

(1() Ethnic Culture
Constant -.42*** -2.79***
(.08) (.32)
Black -.42* -.15
(.21) (.22)
Latino -1.28*** -.88***
(.22) (.23)
Asian -.66*** -.81***
(.18) (.19)
(.2) Socioeconomic Status
Education .26***
Income .15***
Unemployed .13
(3) Demographic



(4) Socio-psychological
Personal Discrimination

Concern over Group Status

Rac ial Integrat ion/Acculturat ion

(5) Legal
Naturalization/Citizenship a
Regist ration

Initial -2 Log Likelihood=1370
At Converg. 1325 1254
% Correct 68.10 70.02
















Table 4-4
Multiple Regression Estimations of Participation Other Than
Voting in Southern California, 1993 (N=1108)

Models I II III IV V

Note: Numerical entries are regression coefficients except
where noted. Standard errors are in parentheses.
* p<.05 ** p<.01 *** p<.001
The dependent variable is scored by the summed value of
participation in any of the activities other than voting and
has a range between 0 and 4.
a The coefficient for citizenship is taken from a separate
model where other coefficients are similar to those in
column V, except that Latino (p=.04) is also significant.

(1) Ethnic Culture
Constant .58*** -.30**
(.03) (.10)
Black -.21** -.10
(.08) (.07)
Lat ino -.38*** .19**
(.07) (.07)
Asian -.25*** -.29***
(.06) (.06)
(.2) Socioeconomic Status
Education .100***
Income .059***
Unemployed -.062
(3) Demographic



(4) Socio -psychologi cal
Personal Discrimination

Concern over Group Status

Ac cultural ion/ Int egrat ion

(5) Legal
Natural ization/Cit izenship a













.12 .13
17.18*** 15.11***


Adj -R2 .03
F 13.62***



influencing the incidence and extent of participation in

other types of activities, they are not sufficient to

account for the turnout or participation deficit for Asians

or Latinos. However, the reduction of the significance of

the Latino ethnic label in two of the three cases indicates

that the SES model is more useful for Latinos than for

Asians. The disappearance of the significance of the black

ethnic label for other participation demonstrates that the

SES model works best for the black group. In contrast,

after controlling for SES factors, the Asian turnout

and participation deficit is even deeper than indicated by
ethnic label alone.

The addition of demographic factors does not seem to

help much in accounting for the under-participation of
Asians, either. As shown in column III of Tables 4-2 to 4-

4, the slope coefficients for Asian ethnicity are smaller

than in the SES models but not smaller than the ethnicity

models and they remain significant at p=.001 level. By

comparison, the Latino citizens' disadvantages in length of

stay and age, in addition to education and income, can help

explain the differences between their turnout level and that
of Anglo whites. None of the demographic factors has any

impact on the incidence of Asian or Latino participation,
but age seems to help account for the lesser degree of

participation for Latinos.'
Past research has indicated that the participation

deficit of the Asians cannot be explained away by


sociodemographic and group consciousness variables. Judging
from results shown in column IV of Tables 4-2 to 4-4, we

agree with this assessment. The addition of socio-

psychological factors changes little the slope coefficients
of the Asian ethnicity and they remain highly significant,

depressing turnout or participation in other activities.2
Although none of the socio-psychological factors matter for

voting turnout, controlling for personal experience of

discrimination (being victimized) appears to help reduce the

Latino deficit in the probability of participation in

activities other than voting.

The last column in Tables 4-3 and 4-4 compares the

effects of legal factors such as registration and

citizenship status when differences in SES, demography, and

socio-psychological factors are controlled. As

hypothesized, being registered increases the likelihood and
the rate of participation in activities other than voting.

It also helps to reduce the participation deficit between

Asians and whites beyond the ethnic group culture variable.

Being or becoming a U.S. citizen does not, however, increase

participation in activities other than voting. For
citizens' voting turnout, we would like to estimate the

effect of registration but are concerned about the possible

violation of the statistical assumption of independence. A

separate run using registration as a dependent variable
(shown in the last column of Table 4-2) indicates that the

determinants of registration are similar to those of voting,


but citizens who make more cross-racial friends are also

more likely to register. Among those who are registered,

another logistic regression result (not shown) indicates

that their likelihood of voting can be negatively influenced

by being an Asian and positively influenced by having more
education, family income, and concern over group status.

In the end, these comparisons among ethnic groups

indicate the degree of distinction of the Asian political

participation. Despite controls over four different sets of
factors that are often found to determine American political

participation, Asians simply do not turnout or participate
as much as the majority members of the society, let alone

participate more because of their higher overall
socioeconomical achievement. What then influences the

probability and extent of Asian political participation?
Can we conclude from results comparing Asians to other

ethnic groups that SES and other factors do not have

independent impacts on Asilan turnout /participation?
Sorting Out Sources of Influence: Comparatively and

Part of the answer can be visualized by charting out

the relationships between education/income and

voting/part icipat ion across ethnic groups (Figures 1 to 4) :

For Asians, the educational and income payoffs in voting

turnout are the lowest among the four groups. However,

within each of the four groups, more education generally has

the predicted effect of facilitating turnout. A similar but








E ~e '~




Levels of Education

Figure 1
Ethnic Groups by Education

Voting of

~- White

- --- Black

--- -- -- Latino


a) ~IP)
dig) a)CJ)
O' Em o,
o a,
I 'V Ece o a~
CI1 or ynO E-~
o, cnoh o
o, og

Levels of Education

Figure 2
Participation of Ethnic Groups by Education



30 -

I \


------ ~~ Latino

----- Asian

0 1.


30k- 40k-
40k 50k





Levels of Income

Figure 3
Voting of Ethnic Groups by Income

I /

--- -
I -

50 -







- ------ Black






40k- 50k-
50k 60k



Levels of Income

Figure 4
Participation of Ethnic Groups by Income


weaker case can be made for the income effect. Here for

both Asians and Blacks, turnout peaks at the family income

level of $40K to $50K but resurges at the $60K or more

level. In terms of the probability of participation in any

of the four activities other than voting, education appears

to have a very limited impact for Asians. Whereas those

Asians who did not attend college seem to participate at a

rate higher than their counterparts in other ethnic groups,

those achieve higher educational status do not participate

more. Relatedly, the effect of income for Asians is almost

concave, with those having the lowest and highest level of

income participating more, but generally at a rate lower

than other ethnic groups when income level reaches $40K. In

sum, socioeconomic factors are less useful to explain Asian

turnout or participation. Compared to other groups, the

payoffs of educational and income increments are either low

or none. Although more education is correlated with higher

turnout rate for Asian citizens, the effect of education on

other participation and the overall effect of income diverge

greatly from the patterns for whites.
This observation on the usefulness of the socioeconomic

factors to explain turnout and participation within the

Asian group is supported by analyses using logistic and

multiple regressions. As shown in column II of Tables 4-5,

4-6, and 4-7, when SES is considered along with Asian

nationality, education is the only significant determinant

and it only applies to turnout among citizens. However,

Table 4-5
Logistic Regression Estimations of Asian American
Citizens' Registration and Voting Participation in Southern
California, 1993 (N=131)

Models I II III rV Regi s trat ion

Note: The dependent variable
the respondent reported voting
Numerical entries are logistic
noted. Standard errors are in
* p<.05 ** p<.01 *** p<.001

(1) Asian Nationality



















.2) Socioeconomic Status
Education .37*
Income .14
Unemployed .75
(3) Demographic


Immigration Generation


(4) Socio -Psychologi cal
Personal Discrimination

Concern over Group Status

Group Deprivation

Accul turat ion/ Int egrat ion

Initial -2 Log Likelihood


At Converg. 175
% Correct 60.31





in models I-rV is scored 1 if
for the president in 1992.
coefficients except where

Table 4-6
Logistic Regression Estimations of Asian American
Participation Other Than Voting in Southern California, 1993

Models I II III rV v

-(1) Asian Nationality




















.2) Socioeconomic Status
Education .04
Income .07
Unemployed .72
3) Demographic


Immigration Generation


(4) Socio -Psychol ogi cal
Personal Discrimination

Concern over Group Status

Group Deprivation


(5) Legal
Naturalization/Citizenship a

Initial -2 Log Likelihood=
At Converg. 189 186
% Correct 74.57 74.57




Models I II III IV V

Note: (see Table 4-5) The dependent variable is scored 1 if
the respondent engages in any political activity other than
voting and 0 otherwise.
a The partial coefficient for citizenship is taken from a
separate model which uses citizenship as a legal factor.
The coefficients for the other variables are virtually the
same as in the legal model using registration.


Table 4-6--continued

Table 4-7
Multiple Regression Estimations of Asian American
Participation Other Than Voting in Southern California, 1993

Models I II III ry v

(1) Asian Nationality
Constant .26** .10 -.14 -.359 -.269
(.09) (.26) (.32) (.332) (.345)
Japanese .07 .09 .01 .049 .050
(.13) (.14) (.15) (.147) (.147)
Korean .36* .40* .42* .324 .327
(.18) (.19) (.19) (.186) (.186)
Vietnamese .16 .16 .22 .241 .217
(.16) (.16) (.17) (.169) (.170)
Filipino -.05 -.04 -.08 .006 -.005
(.14) (.15) (.15) (.150) (.150)
(2) Socioeconomic Sta tus
Education .016 .011 .005 .003
(.038) (.040) (.040) (.040)
Income .006 .000 .005 .000
(.029) (.030) (.030) (.030)
Unemployed .172 .210 .184 .179
(.172) (.174) (.171) (.171)
(3) Demographic
Length .013 .006 -.001
(.030) (.029) (.030)
Age .049 .055 .047
(.039) (.038) (.039)
Immigration Generation .084 .066 .061
(.076) (.075) (.075)
Male -.084 -.112 -.101
(.107) (.106) (.107)
(4) Socio -psychological
Personal Discrimination .459** .449**
(.141) (.142)
Concern over Group Status .054 .040
(.108) (.109)
Group deprivation .132 .117
(.105) (.107)
Acculturat ion .130 .098
(.146) (.150)
(5) Legal
Naturalization/Citizenship a g40
Registration .111
Adj -R2 .010 .000 .004 .062 .061
F 1.45 1.00 1.06 1.77* 1.71*

Note: Numerical entries are regression coefficients except
where noted. Standard errors are in parentheses.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - Version 2.9.7 - mvs