Explaining Immigrant Naturalization

Explaining Immigrant Naturalization¹

Philip Q. Yang

University of California, Los Angeles

Prior research on immigrant naturalization has focused mainly on the effects of immigrants' adaptation experiences and demographic characteristics on their propensity to naturalize. This article proposes a broader analytical framework which incorporates immigrants' individual characteristics and larger social contexts in the country of origin and the country of destination to explain the likelihood of citizenship acquisition. The framework is tested for a cohort of recent immigrants, using the PUMS data from the 1980 U.S. census. The results show that economic, political, social, cultural and geographical conditions in the country of origin, and immigrants ethnic communities and urban concentration in the country of destination, to a large extent influence immigrants' propensity for naturalization and that, net of the contextual factors, many of the immigrants' adaptation and demographic characteristics are also significant predictors of citizenship acquisition. The costs, benefits and meaning of naturalization and their intervening roles in the naturalization process are also discussed.

With the advent of another surge of immigration to the United States, immigrant naturalization or citizenship acquisition is becoming a subject of growing importance. Citizenship acquisition is a direct force which affects immigrants' potential political influence and their mobility in the host country (Portes and Curtis, 1987). Citizenship acquisition signifies the shift of immigrants' allegiance and commitment to the receiving country, and therefore it also measures the extent to which the American system can absorb immigrants or, from the perspective of immigrants themselves, the extent to which immigrants are willing to become an integral part of American society. From a policy point of view, an understanding of the naturalization process could have significant implications for the improvement of immigrant resocialization and naturalization services.

It is a common perception of students of naturalization studies that research on immigrant naturalization is not fully developed (Alvarez, 1987; DeSipio, 1987; Garcia, 1981; Portes and Curtis, 1987). As DeSipio (1987:402) stated, "Naturalization, as a focus of social science research, has not received the attention it merits." Previous studies have centered on the influence of immigrants' adaptation experiences and demographic characteristics on their propensity for naturalization. Within this framework, there are two research traditions. The "socioeconomic-determinants" model, which prevailed in ear-

^JAn earlier version of this paper was presented at the annual meeting of the American Sociological Association, August 1992, Pittsburgh, PA. I thank Donald Treiman for his valuable suggestions on earlier drafts. Two anonymous IMR reviewers provided insightful and constructive comments. Thanks are also due to Ivan Light, Kenneth Bailey, Valerie Oppenheimer, David McFarland, Edward E. Telles, Warren TenHouten, and Jianling Li for their useful comments on earlier drafts.

IMR Vol. xxviii, No. 3 449

450 International Migration Review

Her studies, stresses the role of socioeconomic status in the naturalization process (Beijbom, 1971; Bernard, 1936; Krassowski, 1963). In his pioneering analysis, Bernard (1936) found that immigrants with higher levels of formal education, occupational status, and family income were more likely to become naturalized citizens.² The evidence provided by Barken and Khokhlov (1980) countered the Bernard hypothesis and gave rise to the "nonsocioeconomic determinants" model. This second tradition emphasizes the influences of immigrants' cultural, economic and structural assimilation into the host society as well as immigrants' demographic characteristics. The foregoing two types of traditions are essentially noncontradictory; rather, they complement each other. The conflicting empirical evidence may well be associated with the nature of data (individual vs. aggregate data). Despite their differences, both traditions use immigrants' individual characteristics as predictors of their naturalization, and both view naturalization as an outcome of immigrants' successful integration or assimilation into the host society either culturally, economically, or socially. In fact, immigrants' socioeconomic status can be viewed as an indicator of their economic and social integration. More recent studies using multivariate techniques have shown that immigrants' individual assimilation and demographic characteristics are useful for explaining naturalization, but the empirical evidence is largely mixed (e.g., Evans, 1988; Garcia, 1981; Guest, 1980; Portes and Curtis, 1987; Portes and Mozo, 1985). Especially in some of these studies, several commonly perceived determinants of naturalization, such as education, occupational status, income, English competence, etc., were found to have no association with it. Hence, stable multivariate statistical results based on large samples are needed to determine whether many of the common-sense explanations of the naturalization process are erroneous or plausible propositions.

Moreover, past research has given less attention to the influence of larger social contexts in the country of origin and the country of destination in the naturalization process. Specifically, most previous studies have tended to overlook the effects of country-of-origin characteristics on immigrants' propensity to naturalize. There are some exceptions. For instance, Portes and Mozo (1985) provided weak evidence for die effects of proximity to the United States and emigration from a refugee-sending country (significant at the .10 level). Jasso and Rosenzweig (1986) showed that some country-of-origin characteristics, such as GNP per capita, literacy rate and emigration from countries with centrally planned economies, were consistently reliable predictors of naturalization both for females and for males. Nonetheless, the generalizability of Portes and Mozo's findings is limited because they studied only Mexican and Cuban immigrants using aggregate data. Jasso and Rosenzweig's study (1986) did not include all important country-of-origin characteristics such as refugee-sending country, recognition of dual citizenship, etc., and it did not estimate the effects of

^Bernard (1936) labeled the variables used in his analysis as "cultural determinants of naturalization." But the term socioeconomic determinants of naturalization is probably a more accurate description of his model.

Explaining Immigrant Naturalization 451

immigrant assimilation experiences and many demographic characteristics due to the limitations of their data. Destination characteristics have also largely been neglected in prior studies. We know little about the conditions in the host country which may be relevant to naturalization. The implied assumption that destination conditions have no impact on naturalization could hardly be true because immigrants' ethnic communities and the environment in the host country certainly influence their decisions and behavior, including their propensity to naturalize. The exclusion of these kinds of important social contexts could result in biased or inaccurate estimates of the effects of individual characteristics on naturalization. More importantly, without understanding these contextual determinants of naturalization, our knowledge of the naturalization process would not be complete. Systematic evidence regarding the effects of country-of-origin and destination conditions is required.

Another shortcoming of prior studies lies in the neglect of the direct roles of incentives or disincentives for naturalization. It is not obvious whether, and what kinds of, benefits and costs involve in immigrants' naturalization decisions. Furthermore, we understand little about how immigrants' individual characteristics and the social contexts influence their naturalization decisions by shaping their perceptions of the costs, benefits and meaning of naturalization, which are probably the most direct determinants of naturalization decisions. Admittedly, available data do not allow us to operationalize directly the intervening roles of incentives or disincentives in the naturalization process, but some theoretical exploration is necessary.

There are other limitations in past research as well. For instance, most prior studies have concentrated on Hispanic immigrants (e.g., Alvarez, 1987; Garcia, 1981; Grebler, 1966; Portes and Curtis, 1987); therefore, it is not clear whether the findings are generalizable to other immigrant groups and to all the immigrants in the United States. In addition, microlevel analysis of naturalization is relatively scarce. As a consequence, empirical results based on aggregate data sometimes are used as if they were derived from individual data, leading to the well-known ecological fallacy.

This article analyzes the determinants of naturalization in the United States for a cohort of recent immigrants, using a broader analytical framework and the 1980 census data. The general questions guiding this analysis are: What conditions influence immigrants' propensity to become naturalized U.S. citizens? How do immigrants' individual characteristics and larger social contexts affect their propensity to naturalize through the perceived costs, benefits and meaning of naturalization?

COSTS, BENEFITS AND MEANING OF NATURALIZATION

Costs, benefits and meaning of naturalization are the most immediate considerations in immigrants' decisions to naturalize. Citizenship grants immigrants certain political, civic and social rights and privileges to which permanent

452 International Migration Review

residents are not entitled. One of the most important privileges of citizens, which inspires immigrants to seek citizenship, is political rights, namely, the rights to vote and run for public office. Only U.S. citizens are able to vote in all elections and therefore to influence political decisions and outcomes at the national, state and local levels which may have bearings on their lives. As naturalized citizens, immigrants can run for any public office, except that of president and vice president.

It is a widely held belief that many immigrants primarily pursue naturalization not for political rights, but for immigration benefits accompanying U.S. citizenship - easier and faster immigration of their relatives to the United States. Under the U.S. immigration system operating from 1965 to 1991, permanent residents can only sponsor their spouses and unmarried children as immigrants within a numerically limited category (the second preference), which only accounted for about 26 percent of the total quota of the numerically limited categories. On the other hand, citizens can sponsor not only their immediate relatives with no numerical restriction, but also their unmarried or married adult children and their siblings, who made up about 54 percent of all the numerically limited categories. The earlier the immigrants become citizens, the shorter the waiting time for their relatives to immigrate.

A third significant advantage of acquiring U.S. citizenship is the qualification for a U.S. passport. An American passport is often considered the most desirable to possess. It confers U.S. citizens worldwide protection of the United States; and it possibly makes travel abroad easier. For immigrants who want to travel overseas, U.S. citizenship is certainly attractive.

Access to broader life opportunities is another naturalization benefit. Although citizens and permanent residents now enjoy similar opportunities in employment, welfare assistance and education, certain restrictions do apply to noncitizens. For instance, certain government jobs (e.g., some jobs in federal government agencies, bureaus, think tanks, and many public safety positions with state and local governments) and certain private sector industries, such as defense contractors, require U.S. citizenship.

Noncitizens who retire abroad are denied full Social Security benefits. Some educational opportunities, loans, and fellowships (e.g., Rhodes Scholarships, Fulbright Graduate Study Abroad Program and Fulbright Teacher Exchange Program, White House Fellowship Program, all National Science Foundation Fellowships, all Ford Foundation Fellowships) also require citizenship. Naturalization qualifies immigrants for those jobs, programs, and fellowships for which citizenship is a prerequisite. It offers immigrants important avenues of economic and social mobility.

There are other kinds of convenience provided by citizenship. For example, immigrants will no longer have to carry their alien registration cards or notify the Immigration and Naturalization Service (INS) of their address.

It should be noted that while the value of citizenship in political rights, immigration benefits, and international travel has remained stable over the past

Explaining Immigrant Naturalization 453

three decades, the specific marginal value of citizenship in the sphere of civic and social rights has diminished over time as additional privileges attached to permanent residence have increased. For instance, up to the late 1960s most states still had laws which required citizenship in a total of 81 licensed occupations, such as accountant, architect, attorney, dentist, physician, private detective, funeral director, liquor dealer, etc. (Sanders, 1968). However, my own research on the current state codes or statutes reveals that almost all of these occupational restrictions have been repealed over the years, except for those on some higher-level executive, legislative and judicial positions.

Citizenship also entails costs. These costs comprise at least four types: 1) increased citizen obligations in the host country, which, in addition to universal obligations of all residents such as paying taxes and obeying laws, include the responsibilities to participate in the political process, to uphold the Constitution, to serve in the military if necessary, and to remain loyal to their adopted country during wartime; 2) lost or reduced political, civic and social rights previously enjoyed as nationals in the country of origin; 3) voluntary renunciation of their former nationality and of allegiance to their native land through oath, which, in some sense, can be considered as a psychological cost; 4) the actual input in a long, complicated application process. To obtain citizenship, a minimum of five years of continuous residence in the United States is required. Applicants must also have resided for six months in the state where they file their petition. Once the residence requirements are met, applicants have to file their applications, take a preliminary examination in order to meet the language, education, and other requirements, and attend a final hearing. Applicants are also charged a fee. Some immigrants do not bother to apply for citizenship because they either find the naturalization procedures too complex or are afraid of the preliminary examination. In fact, a significant proportion of applicants (30% in fiscal year 1986) failed to pass the examination (North, 1987).

Citizenship indicates the change of one's national identity, a change from a "foreigner" to an "American." The feeling about this identity change varies considerably from person to person. Some immigrants find it honorable to be an American while others feel ambivalent or indifferent or even dishonorable, depending upon their sense of belonging and commitment to their native countries. The attitudes of their relatives, friends, and significant others toward the change of national identity may also be very different, either positive or negative. This factor enters the immigrants' naturalization decisionmaking process.

How much immigrants understand the costs, benefits and meaning of naturalization, how much they need the benefits conferred by citizenship, how much they perceive the loss, and how much they appreciate being an American certainly vary across individuals. These variations in the perceptions of the costs, benefits and meaning of naturalization most directly explain individual differences in the propensity to naturalize. A previous study using attitudinal variables

454 International Migration Review

has shown that the perceived benefits or costs and attitudes towards being an American or a foreigner influenced immigrants' naturalization decisions (Garcia, 1981). Unfortunately, since the 1980 U.S. census data do not contain information on the perceived costs, benefits and meaning of naturalization, it is impossible in this study to test the effects of these factors. Thus, the remainder of this study focuses on the effects of structural factors on naturalization for which I have data. On the other hand, because of the direct and important role of the perceived costs, benefits and meaning of naturalization, any analysis of the naturalization process cannot ignore them. Meanwhile, the perceptions of incentives or disincentives for naturalization probably depend on individual characteristics and larger social contexts. For example, immigrants who are successfully integrated into American society are more likely to value the political rights of citizens and appreciate being an American; immigrants from countries with poor economic and/or political conditions probably feel the urgency to help their relatives to immigrate. It is these mechanisms that allow us to link structural factors and naturalization together without direct information on the perceived costs, benefits and meaning of naturalization. Therefore, when appropriate, I will discuss how contextual factors and individual characteristics determine the naturalization propensity through their effects on immigrants' perceived incentives or disincentives for naturalization.

HYPOTHESES

The framework proposed here may be called a "contextual perspective," which views naturalization as individual behavior not only determined by individual adaptation experiences and demographic characteristics. However, it is also conditioned by the social contexts in both the country of origin and the country of destination through the intervening roles of the perceived benefits, costs and meaning of naturalization. This approach permits more accurate estimates of the influences of individual characteristics; moreover, it allows the examination of the effects of contextual factors. Within this framework, several hypotheses are subject to empirical testing.

Individual Characteristics

Some previous studies have found that successful cultural and economic integration into American society increases immigrants' propensity to naturalize (e.g., Beijbom, 1971; Bernard, 1936; Garcia, 1981; Krassowski, 1963; Portes and Curtis, 1987). But what mechanisms produce this positive association is the question. Cultural assimilation has at least two functions: 1) it fosters immigrants' sense of belonging to the host society by promoting their familiarity with its language, norms, values, history, government, and social systems; it therefore nurtures positive attitudes towards being an American; 2) by so doing, cultural assimilation also enhances immigrants' ability to meet the language and literacy

Explaining Immigrant Naturalization 455

requirements of naturalization. Successful economic adaptation to the recipient country (e.g., acquisition of a higher occupational status, higher income, home ownership, and entrepreneurship) increases immigrants' commitment to the host society. As economic status improves, the cost to give it up rises, and therefore the demand for political rights and participation in order to influence decisions which could affect their achieved economic status also increases. In other words, immigrants' perceived benefits of naturalization will increase. Consequently, as immigrants' positive attitude toward being an American and their perceived value of American citizenship increase, so does their propensity to become American citizens. Hence, I predict that the better the immigrants' cultural and economic adaptations to the United States, the more likely the immigrants are to naturalize, other things being equal.

Several demographic characteristics are expected to affect immigrant propensity to naturalize through incentives or disincentives for naturalization. One such characteristic is age at immigration. The relationship between age at immigration and naturalization is probably curvilinear. When immigrants arrive at early ages, they do not understand the value of citizenship and therefore have low propensity to naturalize. Immigrants who are more mature at the time of immigration, especially those entering the labor force, realize the benefits citizenship can bring for their education and work opportunities and therefore have higher propensity for naturalization. However, for immigrants who are growing older at the time of entry, enthusiasm for citizenship probably fades because the length they can possibly enjoy citizenship benefits is shorter and the difficulties in meeting the language and education requirements and in overcoming commitment toward the native land are greater. Thus, I expect that as age at immigration rises, immigrants' propensity for naturalization first increases and then declines after a certain point in their life cycle.

Sex may make a difference in the naturalization decision since some prior studies pointed to significant gender differences in naturalization (e.g., Grebler, 1966). I expect that men have a higher inclination to seek citizenship than do women because they are more likely to work at jobs where citizenship may be an advantage. A competing hypothesis would predict the opposite, arguing that females are more vulnerable than their male counterparts and therefore have stronger motivations to acquire citizenship in order to be rid of disadvantageous situations such as repressive marriages or jobs (Alvarez, 1987).

Another demographic characteristic is marital status. Previous studies provide mixed evidence. While Guest (1980) reported no significant effect of marital status on naturalization, Beijbom (1971) did find that married Swedes were more likely to naturalize. Despite the conflicting evidence, I hypothesize that married immigrants are more likely to seek citizenship than single persons, other things being equal. There are two reasons for this: 1) since married persons have a relatively stable life and it costs more for them to return to their homelands (Grebler, 1966), they are more likely to stay in the United States than are single persons; 2) because families have more contacts with American

456 International Migration Review

institutions, it is easier to become assimilated into American society and therefore to accept the identity of being an American.

Whether immigrants have minor children may contribute to their naturalization decisions. Presence of minor children encourages naturalization in two ways. First, when parents become citizens, their minor children can derive citizenship from them. Children do not have to go through the complicated process again. This is certainly cost effective. Second, with the presence of children, immigrants are more likely to "take root" in America (Portes and Curtis, 1987). Thus, I predict that immigrants with minor children have a higher propensity to naturalize than those without minor children, all else being equal.

According to the naturalization laws, aliens serving in the U.S. Armed Forces during World War I, World War II, the Korean conflicts, and the Vietnam conflicts are exempt from the residence requirement and some other requirements; aliens serving in the military during peacetime are required to reside in the United States for only three years (Krichefsky, 1963; Smith, 1973). This means that the cost involved in the application process is much less and the benefits of citizenship can be derived earlier. This condition certainly promotes naturalization. In fact, some immigrants served in the Armed Forces specifically to become U.S. citizens (Fujiwara, 1969). This action also indicates their appreciation of the meaning of being an American. Hence, I hypothesize that immigrants who served in the U.S. Armed Forces are more likely to naturalize than those who did not, all else being equal.

Social Contexts in the Countries of Origin and Destination

Immigrants' perceptions of the costs, benefits and meaning of naturalization and their naturalization decisions are conditioned by the large social environment around them. This kind of influence comes from both the country of destination and the country of origin. Research on international migration has pointed to the important role of immigrants' ethnic networks in the receiving country in extracting new immigrants, ensuring their early survival, findingjobs, circulating information, etc. (e.g., Anderson, 1974; MacDonald and MacDonald, 1974; Tilly, 1978). The role of immigrants' ethnic networks could be further extended to the naturalization process. There are two competing hypotheses about the effect of the size of immigrant community on naturalization. One predicts that the size of an immigrant community in the host country is inversely related to the probability of naturalization. The argument is that a large immigrant community increases the likelihood of forming a self-contained ethnic network, in which new immigrants can function well without citizenship. As the size of one's own group increases, the perceived citizenship benefits, especially protection by citizenship, decreases and immigrants' sense of belonging to the whole society attenuates. Thus, immigrants are less likely to use citizenship acquisition as a strategy for the purpose of self-protection and are less likely to identify themselves as Americans. A contesting hypothesis predicts

Explaining Immigrant Naturalization 457

a positive association between the size of the immigrant community and the likelihood of its members' naturalization, because a large immigrant community facilitates naturalization by helping its members' assimilation into American society andby providing its members with information concerning the benefits, procedures and experiences of naturalization.

The literature suggests that immigrants tend to concentrate in large urban areas (Portes and Rumbaut, 1990; Waldinger, 1989). Urban concentration facilitates the circulation of information about naturalization and therefore increases immigrants' perceived value of citizenship. Urban environments also offer convenience for naturalization services and thereby reduce the cost of the application process. Hence, I hypothesize that the degree of urban concentration of an immigrant group is positively associated with the probability of its members' naturalization, other things being equal.

In the naturalization process, a very important decision immigrants have to make is whether or not they are eventually going to return to their homelands. Such decisions are largely shaped by economic, political, social and geographical conditions in their countries of origin. As predicted by the "reversibility" hypothesis labeled by Portes and Rumbaut (1990), less favorable conditions in the country of origin reduce the reversibility of migration flow and therefore increase the probability of naturalization; conversely, more favorable conditions lead to exactly the opposite. More importantly, in addition to deterring return migration, conditions in the home country may affect immigrants' naturalization decisions by boosting the demand for emigration and by pressing immigrants in the United States to maximize the citizenship benefits - especially the benefits to sponsor relatives' immigration - through naturalization.

Several specific conditions in the country of origin may be important in promoting or discouraging immigrants' propensity to naturalize. Poor economic conditions and low standards of living in the country of origin may deter immigrants' desires to return to their homelands and therefore inspire them to stay in the United States forever. More significantly, not only do immigrants from poor countries gain more by staying in the United States, but their relatives in the home country also benefit if they acquire U.S. citizenship because citizenship can accelerate the immigration of those still remaining in poor countries. Thus, the level of economic development in the country of origin is negatively associated with the propensity for naturalization, other things being equal.

The system that provides less economic freedom and imposes tighter political control over individuals, such as the Stalinist style of socialism, may deter emigrants from returning to their homelands and encourage them to finally settle in the host country. Furthermore, the demand for immigration to the U.S. from these countries is higher than that from nonsocialist countries, and therefore immigrants from socialist countries in the United States will feel more pressures from their relatives in the home country to acquire citizenship. Hence, emigration from socialist countries increases the likelihood of naturalization.

458 International Migration Review

Refugee immigrants who exile from countries experiencing wars/revolutions/religious persecutions, or other political chaos are more likely to seek citizenship because their countries of origin do not allow them to return without suffering. Immigrants from these countries suffer less by settling in the United States. Through citizenship acquisition, they can further benefit by helping their relatives to immigrate to the United States.

Prior studies have found that the physical proximity of the home country to the host country increases the reversibility of migration and therefore suppresses immigrant propensity for citizenship acquisition (Portes and Curtis, 1987) because the nearness to the homeland allows lower cost return journeys and retards the pace of immigrant integration through frequent, continuous ties with the nearby homeland (Barken and Khokhlov, 1980). Conversely, a greater distance between the home country and the receiving country will reduce the likelihood of return migration and encourage naturalization. In addition, the necessity to bring in remote relatives is more justifiable than to sponsor the immigration of relatives in nearby countries. Accordingly, I expect that the distance between the country of origin and the United States is positively associated with the likelihood of naturalization, all else being equal.

Cultural similarity (e.g., use of the same language, common history, and similar traditions) between the country of origin and the country of destination reduces the difficulty of immigrants' integration and fosters their cultural adaptation to the host society. Consequently, immigrants may be more likely to accept the identity of being an American. Hence, the "cultural similarity" hypothesis predicts that cultural similarity between the home country and the host country increases immigrants' propensity for naturalization, other things being equal.

Another important condition that influences immigrants' propensity for naturalization is the legal or actual acceptance of dual citizenship by the country of origin. Israel, Canada, Thailand and Taiwan are examples of such countries. The recognition of dual citizenship allows immigrants to take advantage of citizenship in the host country without giving up full privileges and identity in the country of origin. In other words, it costs less to gain citizenship benefits of the host country. Hence, the "dual citizenship" hypothesis predicts that the recognition of dual citizenship increases the propensity for naturalization.

DATA AND METHODS

Sample

Data from the 5 percent Public Use Microdata Sample (PUMS) of the 1980 U.S. census are used to test the proposed hypotheses. The 1980 PUMS is a nationally representative sample of the U.S. population. It contains both household records and the records of individuals residing in the household. Variables at both the individual level and the household level are used in the analysis. In

Explaining Immigrant Naturalization 459

addition, aggregated data on destination and country-of-origin characteristics are attached to individual records.

This analysis is restricted to the immigrants who were aged 18 and over in 1980 and who immigrated to the United States during 1970-1974. Since the U.S. naturalization laws generally require an immigrant to be 18 years of age or older and to have resided in the United States for at least five years in order to be eligible for naturalization, the restriction by age and the exclusion of those arriving after 1974 are justified. In addition, because immigrants who arrived before 1970 may have naturalized much earlier, it is not appropriate to use their characteristics circa 1975-1980 to predict an event which might have occurred a long time ago. Hence, I analyze only the 1970-1974 cohort of immigrants, who started to be eligible for naturalization in the period of 1975-1979 and who were most likely to begin their acquisition of citizenship around 1980.³ The restricted sample contains 83,437 cases. After listwise deletion of missing values, a minimum number of 82,241 cases remains for the logistic regression analyses.

Compared to samples in prior studies, the data set used in this analysis has at least two advantages: 1) the sample represents the 1970-1974 cohort of all immigrants who met basic eligibility requirements for naturalization and 2) it contains a large number of cases, which permits the use of a large number of explanatory variables without compromising the reliability of estimates.

Although the PUMS data have merits in studying the naturalization process, some limitations of their use deserve to be noted. First, since not all the foreign born enumerated in the census are eligible for naturalization, some ineligible foreign born (e.g., illegal immigrants, nonimmigrant aliens) might be included. However, the problem of illegal immigrants may have been alleviated to a great extent through the restriction of the analysis by age and year of immigration discussed above.⁴ The likelihood of including nonimmigrants in the sample may be small because nonimmigrant aliens usually cannot stay in the United States longer than five years.

Another limitation in the use of the PUMS data is that since census data do not include information on the date of naturalization, the causal order between naturalization and such explanatory variables as occupational status, income, home ownership, self-employment, English competence, education, and mari-

³According to the INS (1989), the median number of years of residence for the total naturalized population has remained fairly constant since I960, at about eight years. This suggests that many of the people who immigrated in 1970 would not start to naturalize until 1978, and the naturalization date of those immigrating later (i.e., 1971-1974) should be even closer to 1980.

⁴According to Passel and Woodrow (1984), among all illegal immigrants present and counted in the 1980 census, nearly half of them entered die U.S. between 1975-1980, and 18 percent entered under the age of 15. These undocumented have been eliminated from the sample by the restriction procedure. In addition, researchers believed that illegal immigrants were undercounted by one-third to one-half in the 1980 census (Keely, 1983:17; Bean a al, 1987). This further reduces the likelihood of including illegal immigrants in the sample. To further test how well the restriction procedure has worked to minimize the impact of illegal immigrants on die results of this analysis, I excluded all Mexicans, who would have accounted for more than half of the illegal immigrants to the United States in the sample had no restriction on the sample been imposed. I tested the same logistic regression models presented in Table 3, which produced virtually the same results.

460 International Migration Review

tal status is ambiguous. The measures for these predictors may be best viewed as proxies at the time of naturalization. Nevertheless, by restricting the analysis to the 1970-1974 cohort of immigrants, the timing of naturalization and that of those explanatory variables should be very close to each other. Moreover, the very large sample size of this study over any previous analyses of naturalization in the United States yields very stable statistical results that will lay a solid foundation for further examination of the effects of these variables using more sophisticated research designs.

The unavailability of certain variables in the PUMS data set which imposes some constraints on the model specification constitutes the third limitation. The present data set does not allow for testing the effects of types of entry visa and attitudinal variables. Despite the limitations, the inclusion of individual and contextual characteristics and the large representative sample provide a unique opportunity to study the naturalization process in the United States.

Measures

Table 1 summarizes the measurements, means, standard deviations, and predictions for the variables used in the analysis. To conserve space, only a brief description is presented. The dependent variable in this analysis is a dichoto-mous measurement indicating whether the respondent was a naturalized citizen or a noncitizen immigrant at the time of the 1980 census.

Immigrants' cultural integration is measured by English competence and years of schooling completed. I use a set of dummies for English competence and a squared term for education to allow for possible curvilinear effects of these two variables on naturalization. The effects of immigrants' economic integration are tested using four measures: Duncan's Occupational Socioeconomic Index updated for application to the 1980 census (Stevens and Cho, 1985), occupational classification, annual income from all sources in 1979, a dummy variable indicating home ownership, and a dummy for self-employment, that is usually used as an indicator of entrepreneurship. Despite the imperfection of these measures, as mentioned earlier, their inclusion can ensure that important variables are not left out and can provide heuristic evidence for settling the previous controversy about the effects of some of these variables.

Five measures are used to represent demographic characteristics. Sex is a dummy and is used primarily as a control. Age at immigration is a ratio measure computed as actual age minus the estimated years of U.S. residence (1980 - the midpoint of the year of immigration). I include a squared term for age at immigration to examine its curvilinear effect. Marital status, presence of children aged 12 or below in 1975,⁵ and serve in the U.S. Armed Forces in 1975 are all dummies.

⁵This variable was constructed by assigning 0 to those without children in 1980 and those with children under 6 years only in 1980 (they did not have children in 1975) and 1 to those with children aged 6-17 in 1980 (the children were under 12 in 1975).

Explaining Immigrant Naturalization

461

TABLE 1

Description of Variables and Predictions

Variable	Measurement	Mean	Standard Predicted Deviation Direction
Dependent Variable
Citizenship	l=citizen, 0=else	.27	.44
Predictors
English Competence
Not at all	(Reference category)
Not very well	l=not very well, 0=else	.21	.40 +
Well	l=well, 0=else	.26	.44 +
Very well	l=very well, 0=else	.27	.45 +
English only	l=English only, 0=else	.14	.35 +
Education	Years of schooling	12.84	5.03 с^а
Occupational status	Duncan's SEI score updated, 1980	29.69	17.10 +
Annual income	Dollars, 1979	8,657	10,557 +
Home ownership	l=homeowner, 0=else	.45	.50 +
Self-employment	l=self-employed, 0=else	.04	.19 +
Sex	l=male, 0=female	.48	.50 +
Age at immigration	Years	27.46	12.84 c^a
Marital status	l=married, 0=else	.80	.40 +
Presence of children age 12 or under in 1975	l=present, 0=else	.42	.49 +
Service in the Armed Forces in 1975	l=in U.S. Armed Forces, 0=else	.01	.08 +
Size of immigrant ethnic community in 1975	Number of immigrants from same origin	472,687	499,073 +
Urban concentration	% urban population	53.84	11.01 +
GNP per capita	Dollars, 1976	1,508	1,623
Socialist country	l=socialist, 0=else	.14	.35 +
Refugee-sending country	l=refugee country, 0=else	.08	.27 +
Distance to nearest U.S. port	Miles	2,964	2,771 +
English an official language	l=English, 0=else	.23	.42 +
О О Dual citizenship	l=dual, 0=else	.08	.27 +

^a Curvilinear relationship.

462 International Migration Review

Two destination variables are included in the analysis. The effect of immigrants' ethnic networks on naturalization is tested by using total number of immigrants who immigrated before 1975 from the same country of origin and were 18 years or older (converted to 100% using the 5% sample). This is a crude measure, but it is important to test the possible impact of ethnic communities. Urban concentration is measured by percentage of urban population (i.e., those living in central cities) of an immigrant group.

Six measurements of country-of-origin characteristics were constructed by attaching certain generally recognized features of the origin country to individuals of that country through their birth place. Economic condition of the sending country is measured by per capita GNP in 1976 (World Bank, 1980). Two dummies - socialist country⁶ and refugee-sending country⁷ - measure sociopolitical conditions of sending countries. Geographic condition of the sending country is measured by distance in miles between the sending country (usually its capital city) and its nearest U.S. entry port. The data for this measure were drawn from Fitzpatrick and Modlin (1986). Dual citizenship is measured by a dummy indicating whether the country of origin recognizes dual citizenship (seeBar-Yaacov, 1961; United Nations, 1954, for classification). Finally, cultural similarity is measured by a dummy indicating whether English is an official language of the sending country (see Banks, 1988, for classification).

Methods

Logistic regression is used to test the effects of the explanatory variables on the probability of citizenship acquisition. The logistic regression model is appropriate since the dependent variable is dichotomous with a skewed distribution (Aldrich and Nelson, 1984; Hanushek andjackson, 1977 ). The model takes the form:

where 1п[р/(1-р^] is the logged odds of becoming a naturalized U.S. citizen, a is the intercept, X_t are the predictive variables, with coefficients b_it and e is the error term. One of the major advantages of logistic regression is that predictive variables may be discrete or continuous since no assumptions are made concerning the distributions of predictive variables (Afifi and Clark, 1990).

^Socialist countries include the countries of Eastern Europe, Albania, Yugoslavia, the USSR, China, Vietnam, Laos, Cambodia, North Korea, Mongolia, Cuba, and Afghanistan (Directorate of Intelligence, 1985).

'Countries with more than 50% of their immigrants admitted into the United States as refugees were coded as refugee-sending countries. These include Cuba, Vietnam, Laos, Cambodia, Romania, the Soviet Union, Latvia, and Lithuania (see INS, 1989:xxiii-xxvii).

Explaining Immigrant Naturalization 463

RESULTS

Descriptive and Bivariate Analyses

Recall that Table 1 also gives the central tendencies and standard deviations of the variables used in the analysis. Interpretation of central tendencies is straightforward for interval-ratio measures. Means for dummy variables can be interpreted as percentages by moving the decimal point two places to the right. Table 1 shows that in 1980, 27 percent of the 1970-1974 cohort of immigrants eligible for naturalization had become naturalized U.S. citizens and 73 percent either resisted or delayed citizenship change. On average, these new immigrants had completed high school (almost 13 years of schooling), and 67 percent (= 26% + 27% + 14%) could speak English well or better. They earned an annual income of $8,657 in 1979 and had a Duncan Occupational SEI score of nearly 30. Forty-five percent of these immigrants owned their own homes, and 4 percent were self-employed. Females slightly outnumbered males. They typically immigrated to the United States at the age of about 27. Most (80%) were married, and slightly more than two-thirds had minor children in 1975. Less than 1 percent served in the U.S. Armed Forces in 1975.

Table 1 also reveals that more than half of these immigrants lived in central cities, and the average size of their country-of-origin groups in 1975 was 472,687. The average level of economic development and living standards in their origin countries in 1976 stood at the middle level as reflected by the per capita GNP of $1,508. The average distance between the origin country and the United States was nearly 3,000 miles. Approximately 14 percent of these immigrants came from socialist countries, eight percent from refugee-sending countries, eight percent from countries recognizing dual citizenship, and nearly one-quarter from countries using English as an official language.

To further understand the data and determine what needs to be taken into consideration in the multivariate analysis, I cross-tabulated the dependent variable by all the independent variables. During this procedure, the interval-ratio measures were recoded into ordinal measures to reduce the number of categories. The results are presented in Table 2. A number of preliminary observations can be made from Table 2. First, consistent with my predictions is the linear relationship between naturalization and several independent variables, including Duncan's Occupational SEI score, home-ownership, self-employment, service in the U.S. Armed Forces, marital status, presence of minor children, and all country-of-origin characteristics except for refugee-sending country. Second, gender, percentage of urban population, and refugee-sending country seem to have little impact on naturalization. However, these bivariate relationships may be misleading due to the associations between the independent variables. Thus, a multivariate analysis is required to determine the independent effect of each variable. Third, several variables reveal possible nonlinear effects on naturalization. The effects of English competence and age at immigration are apparently curvilinear, and therefore the use of a set of

464

International Migration Review

TABLE 2

Percentage Distribution of Naturalized Citizens by Independent Variables,
the 1970-1974 Cohort of Immigrants, Untied States, 1980

Variable Categories

i Naturalized Citizens

i Base

English Competence
Not at all	11.2
Not very well	19.3
Well	31.3
Very well	34.5
Speak only English	27.3
Years of Schooling
0-6	17.5
7-12	21.2
13-16	29.4
17 and above	33.9
Duncan's Occupational SEI Score
Below 30	24.0
30-49	31.3
50 and above	33.6
Annual Income
Below 5,000	24.2
5,000-9,999	24.9
10,000-19,999	30.3
20,000-29,999	34.7
30,000-49,999	34.2
50,000 and above	33.2
Homeownership
Homeowner	31.7
Nonhomeowner	22.5
Self-employment
Self-employed	32.2
Employee	26.6
Sex
Male	26.5
Female	27.1
Age at Immigration
Under 18	24.9
18-34	28.3
35-64	26.1
65 and above	16.7
Marital Status
Married	27.4
Never married	24.4

(9,766) (17,132) (21,963) (22,945) (11,361)

(9,142) (24,014) (30,704) (19,577)

(54,189)

(20,253)

(8,995)

(34,357)

(21,218)

(20,089)

(5,111)

(1,710)

(952)

(37,028) (45,213)

(3,182) (80,255)

(40,076) (43,361)

(20,793)

(43,783)

(17,365)

(1,496)

(66,876) (16,561)

Explaining Immigrant Naturalization

465

TABLE 2 (Continued)

Percentage Distribution of Naturalized Citizens by Independent Variables, the 1970-1974 Cohort of Immigrants, Untted States, 1980

Variable Categories % Naturalized Citizens		% Base
Presence of Children Age 12 or Below in 1	1975
Yes	28.1	(34,887)
No	25.9	(48,550)
Service in U.S. Armed Forces in 1975
Yes	57.6	(552)
No	26.6	(82,885)
% of Urban Population
Below 45%	25.8	(17,862)
45-54%	27.4	(45,137)
55% and above	26.4	(20,438)
Number of Immigrants from Same Country of Origin in 1975
Less than 10,000	29.8	(4,051)
10,000-99,999	29.2	(23,232)
100,000-549,000	32.3	(28,666)
550,000 and above	18.6	(27,488)
GNP Per Capita
$250 or less	31.1	(6,599)
$251-$l,500	26.3	(55,161)
More than $1,500	26.7	(21,677)
Socialist Country
Socialist	31.8	(11,939)
Nonsocialist	26.0	(71,848)
Refugee-Sending Country
Refugee	26.8	(6,589)
Nonrefugee	26.8	(76,848)
Distance to Nearest U.S. Entry Port
Below 1,000 miles	19.1	(33,192)
1,000-4,999	25.7	(25,544)
5,000 miles and above	38.3	(24,701)
English an Official Language
English an official language	31.5	(19,605)
English not official language	25.3	(63,832)
Dual Citizenship
Dual citizenship	30.7	(6,385)
Nondual citizenship	26.5	(77,052)

466 International Migration Review

dummies for the former and a quadratic term for the latter is justified. The effect of education seems to be linear, but it shows a large jump after high school and then a smaller increment after college. Hence, to determine the necessity of a quadratic term for education, it may be useful to compare a model with the quadratic term and a model without it. Income and size of immigrant group also show possible nonlinear effects. In consideration of this and their extreme skewness, log transformations for both variables are appropriate.

Multivariate Analysis

To determine whether the proposed analytical framework better captures the naturalization process than other alternative formulations, I tested three logistic regression models. Model 1 uses only the individual characteristics as independent variables to predict the likelihood of naturalization; Model 2 uses only the contextual factors as predictors; and Model 3 (the full model) incorporates both the individual characteristics and the contextual factors. The results of these three models are given in Table 3.

As seen in Table 3, all three models predict naturalization to varying degrees, and despite some differences, the overall patterns of the logistic regression coefficients are basically the same across models. Hence, we need to decide which of the three models is the best. This involves testing the significance of the differences in model Chi-squares (X²) being compared. If the addition of an independent variable or independent variables produces a significant increment in the model Chi-square, the model with more variables fits the data better since the comparison of interest for logistic regression is not with a saturated model but with a baseline model.

The goodness of fit statistics associated with the three models are shown at the bottom of Table 3. The explanatory power of each model is represented by a" model X²" statistic, with degree of freedom equal to the number of covariates in the model. When models are nested, the differences between the -2 log likelihood X² for the baseline model with only the intercept and the -2 log likelihood X² for the model being evaluated with covariates are asymptotically distributed as X², and then the goodness of fit of the models can be compared. Since Model 1 and Model 2 are not nested, they are not comparable. Our major concern is whether Model 3 fits the data better than Models 1 and 2. By comparing Model 3 with Model 1, we see that Model 3 has a higher explanatory power than Model 1 because the difference in the model X² is very large (1,654 = 5,672 -4,018) and is statistically significant at beyond .0001, with a difference of 8 (= 24-16) degrees of freedom. Likewise, the superiority of Model 3 to Model 2 is evident since the difference in the model X² (=2,492) is also sizable and significant, with a difference of 16 degrees of freedom.

Alternatively, the goodness of fit of a model can be assessed by using the BIC* coefficient (Treiman and Yamaguchi, 1992), a modification of the BIC

Explaining Immigrant Naturalization

467

TABLE 3

Logistic Regression Estimates for Determinants of Citizenship Acquisition for the 1970-1974 Cohort of Immigrants, United States, 1980

		Model 1			Model 2			Model 3
	Logged	Standard	Odds	Logged	Standard	Odds	Logged	Standard	Odds
Predictors	Odds	Error	Ratio	Odds	Error	Ratio	Odds	Error	Ratio
Individual Characteristics
Ability to speak English
Not at all							-
Notwell	.598	(.039)	1.818				.544	(.039)	1.723
Well	1.190	(.039)	3.289				1.080	(.040)	2.943
Very well	1.304	(.040)	3.683				1.219	(.042)	3.384
English only	.947	(.042)	2.578				1.287	(.047)	3.621
Years of schooling	.038	(.007)	1.038				.051	(.007)	1.053
Years of schooling squared	-.001	(.0003)	.999				-.002	(.0003)	.998
Duncan's Occupational SEI score	.001 "	(.0006)	1.001				.001	(.0006)	1.001
(In) Income	.022	(.003)	1.022				.019	(.003)	1.020
Homeownership	.299	(.017)	1.348				.258	(.017)	1.294
Self-employment	.077 "	(.040)	1.080				.045 "	(.041)	1.046
Sex (male=l)	-.149	(.018)	.862				-.110	(.018)	.896
Age at immigration	.028	(.003)	1.029				.021	(.003)	1.021
Age at immigration squared	-.0003	(4E-5)	.9997				-.0003	(4E-5)	.9997
Marital status (married=l)	.140	(.024)	1.150				.166	(.025)	1.181
Children age 12 and under, 1975	.100	(.017)	1.106				.128	(.017)	1.136
In the U.S. Aimed Forces, 1975	1.270	(.092)	3.562				1.172	(.093)	3.228
Characteristics of Countries of Origin and Destination
(In) Size of immigrant group, 1975				.033	(.006)	1.034	.074	(.007)	1.077
% Urban population of immigrant group			.011	(.001)	1.011	.011	(.001)	1.011
GNP per capita (in $1,000)				-.022	(.006)	.978	-.079	(.006)	.924
Socialist country				.262	(.036)	1.299	.249	(.037)	1.282
Refugee-sending country				.105	(.048)	1.111	.136	(.050)	1.146
Distance to nearest U.S. entry port (in 1,000 miles)		.179	(.005)	1.196	.150	(.005)	1.162
English an official language				-.036 "	(.021)	.965	-.321	(.026)	.725
Dvial citizenship				-.254	(.034)	.775	-.224	(.035)	.799
Intercept	-3.199	(.070)	.041	-2.562	(.118)	.077	4.808	(.146)	.008
-2 log likelihood	91,392			93,799			89,736
Model Chi-square	4,018			3,178			5,672
BIC*	3,837			3,087			5,400
Degree of freedom	16			8			24
Number of cases	82,241			83,434			82,238

" Not significant at the .05 level.

Note: The odds ratio is the antilog of the logged odds and can be interpreted as the change in the odds of becoming a naturalized U.S. citizen associated with a one-unit change in the independent variable. Most coefficients are statistically significant at the .001 level or better.

468 International Migration Review

(Bayesian Information Criterion)⁸ coefficient (Raftery, 1986), which takes into account the number of cases in the model. The larger the BIC*, the greater the probability that the model is true given the observed data. Inspecting the BIC* coefficients in Table 3, we see that the BIC* coefficient of Model 3 (=5,400) is significantly larger than that of Model 1 or of Model 2. Thus, Model 3 is the most likely, given the data, and the naturalization process can be best explained by immigrants' individual characteristics and destination and country-of-origin characteristics.

Having settled upon Model 3 in preference to Models 1 and 2,1 now turn to the interpretation of its specific coefficients. To facilitate the interpretation, the antilogs of the logged odds are given in the Odds Ratio column for each of Models 1, 2, and 3. The odds ratio can be interpreted as the change in the odds of becoming a naturalized U.S. citizen associated with a one-unit change in the independent variable, net of other variables in the model. An odds ratio of 1 indicates no difference in the odds between the categories of the independent variable, a ratio above 1 indicates increased odds, and a ratio below 1 indicates decreased odds. It can be seen from Table 3 that, except for self-employment, all the coefficients are statistically significant. Since the sample is very large, tests of statistical significance are not very informative. Hence, in the remainder of this paper the interpretation of the results will generally focus on the pattern and effect size of the odds ratios.

Model 3 reports the results of the full logistic regression model. In general, controlling for the contextual factors, the immigrants' adaptation and demographic variables very well predict the likelihood of naturalization. As anticipated, the variables measuring cultural integration show strong effects on naturalization, net of other variables in the equation. As immigrants' level of English proficiency increases, so does their likelihood of becoming naturalized citizens. Notice that in Model 1 the effect of English competence seems to be curvilinear because those who speak English only are less likely to naturalize (odds ratio = 2.578) than those who speak English very well (3 .683) and well (3.289). However, in Model 3 the inclusion of the contextual factors yields a linear relationship between English competence and naturalization. This also illustrates the importance of the contextual factors in generating the accurate estimates of the effects of independent variables. This positive linear relationship indicates that knowledge of English facilitates assimilation and therefore naturalization.

The relationship between education and naturalization is curvilinear.⁹ Be-

⁸The BIC* is defined as

BIC* = Model X²- (model d.f.)* ln(N) where Nis the number of cases in the analysis.

⁹To examine the possibility of linear relationship between these two variables, an exact duplicate model of Model 3 was tested excluding years of schooling squared. Comparison of the goodness-of-fit statistics of Model 3 and this alternative model suggests that Model 3 fits the data significantly better than its alternative; notably, in the alternative model, the coefficient for years of schooling is not statistically significant, indicating distortion of the relationship between education and naturalization and the necessity of including the quadratic term for education.

1.4O

1.38-

1.06-

1.00

"hip	1.30-
.zene
-ri	1.26-
о
Lring	1.20-
Acqui	1.16-
41 0
Odds	1.10-

3 4 S 6 7 8 9 10 11 12 13 14 16 16 17 18 19 20 21 22 Years of Schooling

FIGURE I

The Effect of Education on Naturalization, United States, 1980

м x

Z |

470 International Migration Review

cause the coefficient of years of schooling squared is significant, the coefficients of the linear term and the squared term must be interpreted together. Using the coefficients of both terms, I plotted the effect of education on the odds of acquiring citizenship (Figure I). It is visible from Figure I that as the level of education increases, the odds of acquiring citizenship also increases up to the completion of about high school; after that point, an additional year of schooling begins to reduce the odds of naturalization. How do we explain the negative effect of postsecondary education on naturalization? According to the ethnic resilience perspective, as immigrants obtain higher levels of education, they have more exposure to and gain more awareness of the realities in American society, especially the position of one's group in the ethnic stratification system and the various forms of discrimination, and therefore their awareness of their own ethnic identity heightens (Portes and Bach, 1985). As a result, highly educated immigrants may resist further assimilation and therefore are less likely to identify themselves as Americans and to naturalize. This idea is supported by the data.

Immigrants' economic adaptation has positive impact on their naturalization decisions, but with a lesser degree of importance. Each point on Duncan's Occupational SEI score slightly increases the odds of naturalization by a factor of 1.001, a weak effect. Therefore, the odds of acquiring citizenship for a professional with an average Duncan Occupational SEI score of 68.5 in 1980 (see Stevens and Cho, 1985:151) were nearly 7 percent larger than the odds for an operative or a laborer with an average Duncan SEI score of 19.0 (.068 = _e.ooi3(68.5) __e.ooi3(i9.o)) Similarly, each unit increment in the logged income increases the odds of naturalization by a factor of 1.02. Immigrants who are homeowners are about 1.3 times as likely to naturalize as those who are not. Although self-employment displays an expected positive sign, the difference in the odds of naturalization between those self-employed and employees is not significant. These results generally imply that economic advancement in the host country makes immigrants feel some degree of necessity to have U.S. citizenship in order to protect their achieved success.

The effects of the demographic variables contain few surprises. For instance, immigrants who were ever married were 18 percent (= 1.181 -1) more likely to become citizens than those who were never married. This suggests that family life facilitates immigrants' assimilation and therefore their naturalization. Having minor children at the time of naturalization further encourages naturalization. Immigrants who had minor children in 1975 were about 14 percent (= 1.136-1) more likely to naturalize than those without such children, indicating that the reduced cost in the application process and the increased sense of taking roots associated with the presence of children may stimulate naturalization. Serving in the U.S. Armed Forces shows a substantial effect. The odds of citizenship acquisition for those who served in the Armed Forces in 1975 were 3.2 times the odds for those who did not. This also suggests that reduced cost

1.66-

I 1 I I I 1 I I t 1 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I t I I I I I I I t I I I I 1 I I I I I I f

10 16 20 26 30 36 40 46 60 66 60 66 70

и x

s о

Age at Immigration

FIGURE II

The Effect of Age at Migration on Naturalization, United States, 1980

472 International Migration Review

in the application process and early acquisition of citizenship benefits largely stimulate naturalization.

The expected nonlinear effect of age at immigration is confirmed by the data. Figure II shows this convex function. The odds of acquiring citizenship ascend as age at immigration rises, reaching its peak at age 39, and then descend. This evidence suggests that as age at immigration increases, the perceived benefits of naturalization first increase and then decline beginning in the early 40s, and therefore the likelihood of naturalization follows the same trajectory.

Unexpectedly, sex shows a negative sign. The odds ratio indicates that men are about 10 percent less likely to naturalize than women. This evidence seems to provide mild support for the hypothesis that women have stronger motivations than men to acquire citizenship in order to free themselves from repressive marriages or occupations.

In general, this analysis yields very consistent estimates of the effects of most individual characteristics, including many so-called "common-sense" determinants. The above results lend large support to the assimilation argument that successful cultural and economic integration increases immigrants' sense of belonging and commitment to American society and therefore promotes naturalization. The findings also indicate that cultural integration plays a more important role than economic integration in the naturalization process. Meanwhile, all demographic variables are significant predictors of naturalization. These demographic factors influence naturalization decisions by changing immigrants' perceptions of potential gains or losses from citizenship and of being an American. Overall, these neat results shed some light on the true relationships between these individual characteristics and naturalization.

In Model 3, the variables representing the social contexts are included in part as controls to take into account the differences in the conditions in the country of origin and the country of destination. They per se are important independent forces driving the naturalization process. Let us first examine the effects of the destination variables. Percentage of urban population of an immigrant group has the anticipated positive effect. Each percent increment in urban population increases the odds of becoming citizens by a factor of 1.011, net of other variables, so that, for instance, the odds of naturalization for members of an immigrant group with 70 percent of its members living in central cities are about 77 percent larger than the odds for members of a group with 30 percent of its members living in central cities (0.769 = _е-⁰¹¹<то) - e-⁰¹¹<³⁰>). This result clearly supports the urban concentration hypothesis.

The argument that sizeable immigrant ethnic communities facilitate members' assimilation and naturalization finds large support in the effect of size of immigrant group from the same country of origin. As expected, the larger the group size, the higher the probability of naturalization. The odds ratio indicates that, net of other factors, each unit increment in the logged group size increases the odds of its members' naturalization by a factor of 1.077.

Explaining Immigrant Naturalization 473

The effects of country-of-origin characteristics on citizenship acquisition are perhaps more interesting. As expected, per capita GNP of the sending country in 1976 is negatively related to naturalization. Each thousand dollar increment in per capita GNP reduces the odds of naturalization by about 8 percent (-.076 = .924 -1), holding other variables constant. Hence, immigrants from a country with a per capita GNP of $10,000 will be about 53 percent less likely to naturalize than those from a country with a per capita GNP of $250 (-.527 = e-⁰⁷⁹<¹⁰> . _e-.o79(.25O)) This indicates that a higher level of economic development and living standards in the country of origin discourages naturalization.

On the other hand, emigration from socialist countries and from refugee-sending countries noticeably encourages naturalization. The odds of acquiring citizenship for immigrants originating from socialist countries are about 1.3 times the odds for those from nonsocialist countries. Similarly, immigrants from refugee-sending countries are nearly 15 percent (= 1.146 - 1) more likely to naturalize than those not from such countries.

Also as anticipated, the distance between the country of origin and the United States is positively associated with the probability of naturalization. An additional 1,000 miles in the distance increases the odds of naturalization by 1.162. Thus, the odds of naturalization for an immigrant from a country 7,000 miles away from the United States (the approximate average distance from Asian countries to the United States) are nearly 1.7 times (1.696 = e-¹⁵⁰<⁷> - e¹⁵⁰^)) larger than the odds for someone from a country 1,000 miles away from the United States (the approximate average distance from the Caribbean countries to the United States). Obviously, the effect of distance is considerable.

The foregoing results provide the most compelling evidence to support the "reversibility" hypothesis that unfavorable conditions in the home country deter return migration and therefore encourage permanent settlement and naturalization. They indicate that "push" and "pull" factors are at play in the naturalization decisionmaking process. Furthermore, the results suggest that unfavorable conditions in the home country influence immigrants' naturalization decisions through another channel - increasing the demand for emigration in the home country and therefore pushing immigrants already in the United States to use citizenship acquisition as a strategy to help the immigration and economic mobility of their relatives in the home country.

In contrast, the effect of English as an official language runs against the " cultural similarity" hypothesis. Coming from an English-speaking country does not promote naturalization; rather, it reduces the odds of naturalization by about 27 percent (-.265 = .725 - 1). This unexpected result is not due to the possibility that English-speaking countries tend to be more developed because the level of economic development is already controlled. What causes this negative association is not clear at this point. Further examination is needed.

In spite of the significance of its coefficient, the negative effect of dual citizenship also contradicts the dual citizenship hypothesis that dual citizenship encourages naturalization. The odds of naturalization for immigrants from the

474 International Migration Review

countries which recognize dual citizenship are about 20 percent (-.201 = .799 -1) smaller than the odds for those from the countries which do not. Perhaps, immigrants may perceive dual citizenship as more responsibilities rather than more benefits, and/or they may have confusion about and difficulty in maintaining dual allegiance to both the country of origin and the host country. Thus, immigrants may be reluctant to identify themselves with Americans and therefore are less likely to naturalize.

CONCLUSION

Prior research on immigrant naturalization has focused mainly on the effects of immigrants' adaptation experiences and demographic characteristics on their propensity to naturalize. This article proposes a broader analytical framework which incorporates immigrants' individual characteristics and larger social contexts in the country of origin and the country of destination to explain the likelihood of citizenship acquisition. The framework is tested for a cohort of recent immigrants, using the PUMS data from the 1980 U.S. census. The results show that economic, political, social, cultural and geographical conditions in the country of origin, and immigrants' ethnic communities and urban concentration in the country of destination, to a large extent influence immigrants' propensity for naturalization as well, net of the contextual factors, many of the immigrants' adaptation and demographic characteristics are also significant predictors of citizenship acquisition. The costs, benefits and meaning of naturalization and their intervening roles in the naturalization process are also considered.

Because of the limitations of the current data set, the analytical framework developed in this article by no means exhausts all possible determinants of citizenship acquisition. Moreover, naturalization experiences probably differ considerably among various unique ethnic groups. The unique experience of each ethnic group will probably have significant impact on its naturalization rate and the behavior of its members in the naturalization process. Until we have explored this ethnic diversity, we will not be able to understand completely the forces leading to immigrant naturalization and its long-term consequences in this country.

REFERENCES

Afifi, A. A. and V. Clark

1990 Computer-Aided Multivariate Analysis, Second Edition. New York: Van Nostrand Reinhold.

Aldrich, J. H. and F. D. Nelson

1984 Linear Probability, Logit, and Probit Models. Newbury Park, CA: Sage Publications.

Alvarez, R.

1987 "A Profile of the Citizenship Process among Hispanics in the United States," International Migration Review, 21:327-351.

Explaining Immigrant Naturalization 475

Anderson, G. M.

1974 Networks of Contact: The Portuguese and Toronto. Waterloo, Ontario: Wilfrid Laurier University Press.

Banks, A. S.

1988 Political Handbook of the World: 1988. Binghamton: CSA publication, State University of New York.

Barken, E. R. and N. Khokhlov

1980 " Socioeconomic Data as Indices of Naturalization Patterns in the United States: A Theory
Revisited," Ethnicity, 7:159-190.

Bar-Yaacov, N.

1961 Dual Nationality. London: Stevens & Sons Limited.

Bean, Frank D., et al.

1987 "Undocumented Migration to the United States: Perceptions and Evidence," Population and Development Review, 13:671-690.

Beijbom, U.

1971 Swedes in Chicago: A Demographic and Social Study of the 1846-1880 Immigration. Chicago: Chicago Historical Society.

Bernard, W. S.

1936 "Cultural Determinants of Naturalization," American Sociological Review, 1:943--953.

DeSipio, L.

1987 " Social Science Literature and the Naturalization Process," International Migration Review,
21:390-405.

Directorate of Intelligence

1985 Handbook of Economic Statistics. Washington, DC: National Foreign Assessment Center.

Evans, M. D. R.

1988 "Choosing to Be a Citizen: The Time-path of Citizenship in Australia," International
Migration Review, 22:243-264.

Fitzpatrick, G. L and M. J. Modlin

1986 Direct-line Distances: United States Edition. Metuchen, NJ: Scarecrow Press.

Fujiwara, G. Y.

1969 " New Naturalization Laws," / &N Reporter, 18:18-23.

Garcia, J. A.

1981 "Political Integration of Mexican Immigrants: Exploration into the Naturalization Proc
ess," International Migration Review, 15:608-625.

Grebler, L.

1966 "The Naturalization of the Mexican Immigrants in the United States," International

Migration Review, 1:17-32.

Guest, A. M.

1980 "The Old--New Distinction and Naturalization: 1900," International Migration Review, 14:492-510.

Hanushek, E. A. andj. E.Jackson

1977 Statistical Methods for Social Sciences. New York: Academic Press.

INS (Immigration and Naturalization Services)

1989 1988 Statistical Yearbook of the INS. Washington, DC: Government Printing Office.

476 International Migration Review

Jasso, G. and M. R. Rosenzweig

1986 "Family Reunification and the Immigration Multiplier: U.S. Immigration Law, Origin-
country Conditions, and the Reproduction of Immigrants," Demography, 23:291--311.

Keely, С

1983 Affidavit submitted for plaintiffs in Cuomo et al. v. Baldrige et. al, U.S. District Court,
Southern District of New York, 80 Div. 4550 (JES).

Krassowski, W.

1963 Naturalization and Assimilation-proneness of California Immigrant Populations: A Statistical

Study. Unpublished dissertation. Los Angeles: Department of Sociology, University of

California, Los Angeles.

Krichefsky, G.

1963 "Naturalization Rates of Immigrants," I& NReporter, 12:43-50.

Liang, Z.

1991 "The Naturalization Process among Six Ethnic Groups: A Comprehensive Perspective."

Paper presented at 1991 American Sociological Association Annual Meeting, Cincinnati,

Ohio.

MacDonald,J. and L. MacDonald

1974 "Chain Migration, Ethnic Neighborhood Formation, and Social Networks." In An Urban World. Ed. С Tilly. Boston: Little, Brown.

North, D. S.

1987 "The Long Grey Welcome: A Study of the American Naturalization Program," Interna
tional Migration Review, 21:311-326.

Passel, J. S. and K. A. Woodrow

1984 "Geographic Distribution of Undocumented Immigrants: Estimates of Undocumented
Aliens Counted in the 1990 Census by State," International Migration Review, 18:642--671.

Portes, A. and R. L. Bach

1985 Latin Journey. Berkeley and Los Angeles: University of California Press.

Portes, A. and J. Curtis

1987 "Changing Flags: Naturalization and Its Determinants Among Mexican Immigrants,"

International Migration Review, 21:352-371.

Portes, A. and R. Mozo

1985 "This Political Adaptation Process of Cubans and Other Ethnic Minorities in the United
States: A Preliminary Analysis," International Migration Review, 19:35--63.

Portes, A. and R. G. Rumbaut

1990 Immigrant America. Berkeley and Los Angeles: University of California Press.

Raftery, A. E.

1986 "Choosing Models for Cross-classifications," American Sociological Review, 51:145-146.

Sanders, G. H.

1968 "Aliens in Professions and Occupations -- State Laws Restricting Participation," I & N

Reporter, 16:37-40.

Smith, L. E.

1973 "Naturalization Rates of Immigrants," I &N Reporter, 21:29-35.

Stevens, G. andj. H. Cho

1985 " Socioeconomic Indexes and the New 1980 Census Occupational Classification Scheme," Social Science Research, 14:142--168.

Explaining Immigrant Naturalization 477

Tilly, С

1978 "Migration in Modern European History." In Human Migration Patterns and Policies. Ed. W. McNeill and R. Adams. Bloomington: Indiana University Press.

Treiman, D. J. and K. Yamaguchi

1992 "Trends in Educational Attainment in Japan." In P ersisting Barriers: A Comparative Study of

Education Inequality in 13 Countries. Ed. Y. Shavit and H.P. Blossfeld. Boulder, CO: Westview

Press.

United Nations

1954 Survey of the Problem of Multiple Nationality. U.N. DOC.A/CN.4/84.

Waldinger, R.

1989 "Immigration and Urban Change," Annual Review of Sociology, 15:211-232.

World Bank

1980 World Development Report 1978. New York: Oxford University Press.

Оставить комментарий © Copyright Yang P. (kazgugnk@yahoo.com) Обновлено: 19/04/2011. 207k. Статистика. Статья: США Иллюстрации: 2 штук.		Ваша оценка:

Yang P.: другие произведения.

Explaining Immigrant Naturalization