ethnic enclosure or ethnic competition: ethnic identification among hispanics in texas

8
ETHNIC ENCLOSURE OR ETHNIC COMPETITION : Ethnic Identification Among Hispanics in Texas Sean-Shong Hwang* University of Alabama-Birmingham Steve H. Murdock Texas A&M University Recent theoretical debates have focused on the utility of two alternative perspectives on ethnic identification. However, empirical support for either is neither consistent nor conclusive and previous analyses seldom conceptually or empirically examine the aggregate macrostructural determinants of ethnic identification. This study reevaluates the relative merits of the ethnic enclosure and ethnic competition perspectives among Texan Hispanics using aggregate 1980 Census of Population data for 139 Texas cities with large Hispanic populations. The rejection of the ethnic enclosure thesis in some recent research is unwarranted, as the empirical evidence provides it substantial sup- port and little for the ethnic competition perspective. Further analyses of structural factors in ethnic identification are merited. This study empirically evaluates two competing models for ethnic identification. The first, that of “ethnic enclosure,” derives primarily from Milton Gordon’s (1964) assimilation theory, subsequently expanded by human ecologists (Marston and Van Valey 1979; Mas- sey and Denton 1985; Massey and Mullan 1984) incorporating spatial assimilation as an intermediate variable. This approach sees ethnic identity as immigrants’ sense of common nationality, a residue of shared cultural heritage brought to the host society and maintained as long as they are physically and socially isolated from members of groups outside their ethnic enclave. Newness in the host society and cultural and structural barriers that hinder full integration create isolation. Ethnic identity erodes over time, however, as minorities learn the host’s language, adopt its cultural patterns, improve their socioeconomic standing, and are exposed to and interact more frequently with majority members and other minorities on a primary basis. In sum, the ethnic enclosure thesis maintains that lack of opportunities for intergroup contacts and socioeconomic advancement sustain ethnic identity. As cultural and struc- tural barriers fall, most minority members relinquish their ethnic identity and seek full assimilation. *Direct all correspondence to: Sean-Shong Hwang, Department of Sociology, University of Alabama-Birmingham, Bir- mingham, AL 35294. The Sociological Quarterly, Volume 32, Number 3, pages 469-476. Copyright Q 1991 by JAI Press, Inc. AU rights of reproduction in any form reserved. ISSN 00384253.

Upload: steve-h

Post on 07-Apr-2017

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: ETHNIC ENCLOSURE OR ETHNIC COMPETITION: Ethnic Identification Among Hispanics in Texas

ETHNIC ENCLOSURE OR ETHNIC COMPETITION :

Ethnic Identification Among Hispanics in Texas

Sean-Shong Hwang* University of Alabama-Birmingham

Steve H. Murdock Texas A&M University

Recent theoretical debates have focused on the utility of two alternative perspectives on ethnic identification. However, empirical support for either is neither consistent nor conclusive and previous analyses seldom conceptually or empirically examine the aggregate macrostructural determinants of ethnic identification. This study reevaluates the relative merits of the ethnic enclosure and ethnic competition perspectives among Texan Hispanics using aggregate 1980 Census of Population data for 139 Texas cities with large Hispanic populations. The rejection of the ethnic enclosure thesis in some recent research is unwarranted, as the empirical evidence provides it substantial sup- port and little for the ethnic competition perspective. Further analyses of structural factors in ethnic identification are merited.

This study empirically evaluates two competing models for ethnic identification. The first, that of “ethnic enclosure,” derives primarily from Milton Gordon’s (1964) assimilation theory, subsequently expanded by human ecologists (Marston and Van Valey 1979; Mas- sey and Denton 1985; Massey and Mullan 1984) incorporating spatial assimilation as an intermediate variable. This approach sees ethnic identity as immigrants’ sense of common nationality, a residue of shared cultural heritage brought to the host society and maintained as long as they are physically and socially isolated from members of groups outside their ethnic enclave. Newness in the host society and cultural and structural barriers that hinder full integration create isolation.

Ethnic identity erodes over time, however, as minorities learn the host’s language, adopt its cultural patterns, improve their socioeconomic standing, and are exposed to and interact more frequently with majority members and other minorities on a primary basis. In sum, the ethnic enclosure thesis maintains that lack of opportunities for intergroup contacts and socioeconomic advancement sustain ethnic identity. As cultural and struc- tural barriers fall, most minority members relinquish their ethnic identity and seek full assimilation.

*Direct all correspondence to: Sean-Shong Hwang, Department of Sociology, University of Alabama-Birmingham, Bir- mingham, AL 35294.

The Sociological Quarterly, Volume 32, Number 3, pages 469-476. Copyright Q 1991 by JAI Press, Inc. AU rights of reproduction in any form reserved. ISSN 00384253.

Page 2: ETHNIC ENCLOSURE OR ETHNIC COMPETITION: Ethnic Identification Among Hispanics in Texas

470 THE SOCIOLOGICAL QUARTERLY Vol. 32/No. 3/1991

The second view, “ethnic competition,” counters that above. It derives from Glazer and Moynihan’s (1970) and Greeley’s (1971) work, and has been expounded recently by Portes and his associates (Portes 1984; Portes and Bach 1985; Portes, Parker, and Cobas 1980). Ethnic competitionists see ethnic identity as a dormant political consciousness aroused among minorities as they confront majority prejudice and discrimination. Accord- ing to Portes and Bach (1985, p. 285), “Ethnicity is not . . . a mere consequence of the persistence of traditional cultural traits, but . . . created by the host society.” Confronta- tion with host society “realities” makes ethnics of “immigrants who shared only the most tenuous linkages in the old country” ( p. 25). Such realities, intrinsically discriminatory and prejudicial, exist primarily outside of ethnic enclaves and hence are encountered only when minority members leave their ethnic enclaves and interact with majority members. Those confined in ethnic enclaves are unlikely to experience such outside hostility. This perspective expects ethnic identity to heighten as minorities increase their knowledge about the host society, enter mainstream occupations, and interact more frequently with majority members. In sum, the ethnic competition thesis asserts that greater intergroup contact intensifies minority ethnic identity.

The preceding literature review suggests two opposing views of ethnic identity. Where- as the “ethnic enclosure” thesis anticipates higher ethnic identity among minorities little experienced with the host language and culture and residentially segregated from, and socioeconomically inferior to, the majority group, the “ethnic competition” thesis sug- gests the opposite. It emphasizes that acculturation and socioeconomic status (SES) ad- vancement expose minorities to majority hostility, in turn heightening ethnic identity. Thus, the two views predict reverse effects of the same factors.

Despite extensive discussion of these perspectives’ conceptual merits, few empirical studies assess their relative value (Aguirre, Saenz, and Hwang 1989; Portes 1984; Portes, Parker, and Cobas 1980). Neither view consistently receives empirical support. For exam- ple, Portes and colleagues’ (Portes, Parker, and Cobas 1980; Portes 1984) research pro- vides only limited support for the ethnic competition thesis, despite its much acclaimed theoretical significance. Although Aguirre, Saenz, and Hwang’s (1989) recent study using the National Chicano Survey little supports the ethnic competition view, its support for the ethnic enclosure view is also limited. Given the theoretical significance of these perspec- tives and empirical equivocation, additional research, as provided here, is clearly necessary.

Not only the relative merit of the two perspectives, but the underlying structural factors they imply initiate and maintain ethnic identity need to be examined. Although adherents of both views argue the influence of intergroup contacts and that ecological and structural variables-that is, communities’ aggregate characteristics-markedly affect them, most previous analyses examine only individual-level, social psychological dimensions of eth- nic identity. We directly examine the macrostructural and ecological factors that thesis proponents claim set the environmental, ecological, and structural context for ethnic identity. In addressing such ecological and structural factors we offer a structural indicator of ethnic identity, important since data-sets that effectively combine both structural and individual indicators are lacking. As noted below, language usage is our indicator, one accessing issues related to identity in diverse contexts for which census data are widely available.

Our analysis is no substitute for study of the social psychological aspects of ethnic identity, for structural determinants alone cannot fully explain all its dimensions. Rather,

Page 3: ETHNIC ENCLOSURE OR ETHNIC COMPETITION: Ethnic Identification Among Hispanics in Texas

Ethnic Enclosure or Ethnic Competition 471

we acknowledge, with numerous other scholars (see, e.g., Denton and Massey 1989; Stevens and Swicegood 1987), both structural and social psychological determinants and manifestations. Analyses at multiple conceptual levels are necessary.

In sum, then, this study provides needed additional empirical assessment of the relative utility of the ethnic enclosure and ethnic competition perspectives on ethnic identity. In addition, however, it expands such research by concentrating on ecological and aggregate factors, extending the conceptual bases underlying ethnic identity analysis. Below we develop structural-level variables and hypotheses based on the perspectives and test them with Hispanics in Texas.

HYPOTHESES

Although the two perspectives suggest opposite effects of intergroup contacts on ethnic identity, they agree they are essential determinants. To make the views’ competing nature explicit, Table 1 presents different hypotheses about the impact of contextual factors conducing or hindering contact.

The ethnic enclosure perspective expects SES differentials and residential segregation to increase ethnic identity. Ethnic identity brought to this country will be maintained by minorities isolated from interaction with the host majority. Contrarily, the ethnic competi- tion hypothesis claims entry into the mainstream or breakdown of SES and spatial barriers exposes minority members to host discrimination, in turn heightening ethnic identity. Because minority/majority group interaction presupposes shared residential and SES settings, significant minority/majority differences in such settings and hence in social circles (Blau 1977) renders minority/majority interaction improbable.2

The two theses also differ regarding other factors’ effects on ethnic identity. According to the ethnic enclosure thesis, ethnic minorities less exposed to mainstream society are more likely to orient themselves to their own group. Thus, areas constantly replenished by foreign-born persons and recent immigrants from the same cultural origin are likely to maintain higher levels of ethnic identity than other areas. By extension, Hispanics in

Table I Juxtaposition of Hypotheses Based on Two Theses

of Ethnic Identity

Ethnic Ethnic Independent Variables Enclosure Competition

Education Differential Occupation Differential Income Differential Residential Segregation % Foreign Born % New Immigrants Population Size % Anglo % Black

Notes: +Indicates factors expected to increase ethnic identity. -Indicates factors expected to decrease ethnic identity.

Page 4: ETHNIC ENCLOSURE OR ETHNIC COMPETITION: Ethnic Identification Among Hispanics in Texas

472 THE SOCIOLOGICAL QUARTERLY Vol. 32/No. 3/1991

places with large populations and high percentages of Anglos and blacks are less likely to maintain ethnic identity. Such conditions, ceteris paribus, expose Hispanics more fre- quently to other groups, decreasing ethnic identity (Blau 1977; Hwang and Murdock 1988a).

Conversely, although the ethnic competition thesis also expects large foreign-bornlnew immigrant proportions in ethnic enclaves to reduce outgroup contacts, it holds this should lessen ethnic consciousness as intergroup conflict is avoided. Large population size and greater racial/ethnic heterogeneity, on the other hand, increase minorities’ exposure to hostility and raise ethnic consciousness.

DATA AND METHOD

Analysis draws on 1980 Census of Population summary data for 139 Texas cities with a population of 10,OOO or more and a sufficiently large number of Spanish-origin persons. Smaller cities are excluded residential segregation measure, a key study variable, requires subcity (e.g., block) statistics unavailable for them. Since many analysis items are 1980 Census sample data estimates, we treat only cities with at least 400 Hispanics, thus to heighten data reliability. These selection criteria produce a sample including all but 2 of Texas’ 38 central cities, 68 suburban cities, and 35 nonmetropolitan places of 10,OOO or more population in 1980.

Table 1 lists the variables used, operationalized as indicated below. Residential Segre- gation between Hispanics and Anglos (i.e., whites not of Spanish origin) is measured by the index of dissimilarity. Although alternative measures of residential segregation are available (Massey and Denton 1988), the index is appropriate in this context. As an indicator of evenness in two groups’ residential distributions, it may reveal structural as well as psychological (i.e., fairness) dimensions essential for understanding ethnic identity.

SES differentials can affect intergroup contacts either directly (Alba and Chamlin 1983; Portes 1984) or indirectly through their effects on residential segregation (Hwang et al. 1985; Marshall and Jiobu 1975; Massey and Mullan 1984). Three indices of dissimilarity capture three separate dimensions of group SES differentials. Educational Differential compares Anglo and Hispanic distributions for those 25 or older; and Occupational Diferential, for employed persons 16 years and above. Detailed occupation categories on Summary Tape File 4B (STF4B) from the 1980 Census (U.S. Bureau of the Census 1983) are combined into 13 groups similar to those of the printed reports before occupational differential was calculated. Finally, the Income DifSerential measures the groups’ dis- similarity in family income. STF4B is the data-source employed to calculate these and other variables that require ethnicity-specific data.

Contextual factors related to minority group composition and expected to influence intergroup contact are included as exogenous variables. These include the percent of Hispanics (1) Foreign Born and ( 2 ) New Immigrants (immigrating after 1970). Ecological variables with known effects on group interaction (Hwang and Murdock 1988a, 1988b) such as city (3) Population Size and city racial/ethnic composition as measured by (4) Percent Anglo and (5) Percent Black are also included.

Two census items that classify Hispanics by language spoken at home and the ability to speak English assess the dependent variable, Ethnic Identity. Of all Hispanics 5 and older, we exclude those who speak English only poorly or not at all and then determine the

Page 5: ETHNIC ENCLOSURE OR ETHNIC COMPETITION: Ethnic Identification Among Hispanics in Texas

Ethnic Enclosure or Ethnic Competition 473

proportion who speak only English at home using the limited population. We thus contrast two groups of English-speaking Hispanics, those who speak a language other than English (presumably Spanish) at home and those who speak only English at home. Since both groups can speak English but differ in language spoken at home, possibly, ceterisparibus, those who speak only English at home have lower ethnic identity than their counterpart^.^

Specifically, ethnic identity is operationalized by a logit transformation of the odds that contrasts the proportion of Hispanics who choose to speak Spanish (P,) with the remain- ing who choose not to speak Spanish (1 - P,) at home. That is,

Ethnic Identity = In [P,l(l - P 3 ]

As this study’s major purpose is to compare the two perspectives’ merits by testing hypotheses derived from each, regression analysis is appropriate. However, because our dependent variable is a logit transformation of a proportional variable with variances equal to 1 INP( 1 - P), weighted regression provides more efficient parameter estimates. Hanu- shek and Jackson (1977, pp. 193-194) point out that OLS models using logit dependent variables are heteroscedastic. “Because of these different variances, generalized least squares estimation that weights each [sample] by the reciprocal of these variances [i.e., NP(1 - P)] is appropriate. Thus the logit model is a straightforward application of the generalized least squares model” (Hanushek and Jackson 1977, p. 194).

FINDINGS

Table 2 presents the results of the logit analysis of ethnic identity. Metric regression coefficients (b), as well as standardized regression coefficients (B) that measure the rela- tive strength of each effect, are reported. Standard errors (s.e.) associated with the b’s needed for the significance testing of individual hypotheses are also reported. Individual

Table 2 Weighted Regression of Ethnic Identity on Variables

Representing Two Theoretical Perspectives (N = 139)

Standard Independent Variables b Error Beta Tolerance

Education Differential Occupation Differential Income Differential Residential Segregation % Foreign Born % New Immigrants Ln (Population) % Anglo % Black Intercept R2

Adjusted R2

.025e

.016‘

.001 - .001 - .Ooo

.ow - .038‘ - .022e - .030e 1.967

.83

.81

.006 ,008 ,006 .006 ,004 ,003 ,020 ,002 ,004 ,440

.271

.153

.007 - .006 - ,004

.088 -.091 - .651 - SO7 -

.30

.24

.60

.35

.30

.20

.57

.29

.34 -

~

Notes. =Indicates supports Ethnic Enclosure thesis, p < .05, one-tailed &test. .‘Indicates supports Ethnic Competition thesis, p < .05, one-tailed t-test

Page 6: ETHNIC ENCLOSURE OR ETHNIC COMPETITION: Ethnic Identification Among Hispanics in Texas

474 THE SOCIOLOGICAL QUARTERLY Vol. 32/No. 311991

coefficients significant at the .05 level (one-tailed t-tests) are indicated by “c” or “e” as they support the ethnic competition thesis or ethnic enclosure thesis.

In addition to the standard items usually provided with regression results, tolerance values that indicate the extent of model multicollinearity are presented. These are calcu- lated by first regressing each independent variable on the remaining independent variables in the model. The resulting R2 indicates the extent to which an independent variable relates to all the others. Its tolerance value equals 1 minus its R2. A value of 0 indicates complete linear dependency, and of 1 no relationship. Given the conceptual relationships between several of the independent variables, a test for multicollinearity is necessary.

Analysis results strongly support ethnic enclosure but reject ethnic competition hypoth- eses: Of the 9 model independent variables, the 5 found significant affect ethnic identity as the former predict. These include 2 of the 3 SES differential measures (education and occupation) and 3 that assess the ecological potential for intergroup interaction. Although the income differential is not significant, its sign is as ethnic enclosure predicts.

negative but not significant. Such nonsignificance is unexpected given that segregation is perhaps one of the most direct indicators of oppor- tunity for intergroup contact. This result may, however, be due to segregation’s high correlations with other independent variables such as SES differentials between Anglos and Hispanics (Hwang and Murdock 1988b). That is, although the bivariate correlation coefficient between segregation and ethnic identity is .55 (p < .01, not shown), high correlations between segregation and other independent variables may suppress their unique contributions (Gordon 1968).5

In addition to SES and residence differential effects, findings show the significance of ecological factors such as population size and the percentages Anglo and black. This bolsters ethnic enclosure premises that large population size, combined with greater proportional representation of other racial and ethnic groups in an area, creates a structural condition that favors intergroup contacts and assimilation. Thus, ethnic identity, often viewed in terms of individual variation in experience and psychology, is also influenced by ecological factors that determine potential for intergroup contacts.

Residential segregation’s effect

CONCLUSIONS

We conclude that ethnic identity is affected by factors identified by urban ecologists and other structural theorists of ethnic relations. Measure of ethnic identity primarily as voluntary language use not only consistently supports the ethnic enclosure perspective, but also explains a relatively large proportion of variance. This suggests that, given sufficient time, Hispanics become familiar with the culture and language of the U.S., and as free interaction with the Anglo majority on an equal (SES) basis increases, identity assimilation occurs.

Because our conclusions derive largely from one analysis focused on Hispanics in Texas, only limited generalizations to other “colored” minorities such as blacks, Asians, and Puerto Ricans and other regions are warranted. Although Asians and Hispanics share some patterns of spatial and SES assimilation (Massey and Denton 1987), identity change will likely be more difficult for the former and other minorities with more distinctive physical traits.

Despite these limitations, our research clearly demonstrates that ethnic identity, often considered a social psychology topic, is fruitfully studied at the aggregate level, incorpo-

Page 7: ETHNIC ENCLOSURE OR ETHNIC COMPETITION: Ethnic Identification Among Hispanics in Texas

Ethnic Enclosure or Ethnic Competition 475

rating community ecological and structural characteristics that determine intergroup contact. Full understanding of how ethnic identity forms and changes clearly requires interpersonal-level analyses of intergroup interaction, but such interaction cannot occur without structural opportunities. In addition, household-level factors might affect ethnic identity. The presence of either an elderly or recent immigrant, for example, may require native language use at home by even the most assimilated household members. Assess- ment of such potential effects, however, awaits a subsequent analysis that accesses data not incorporated here.

NOTES

1. One reviewer questions our ethnic competition thesis rendition. A rereading of the represen- tative work (Portes 1984; Portes and Bach 1985; Portes, Parker, and Cobas 1980) confirms our initial interpretation. For example, the depiction of U.S. society as discriminatory and prejudicial reflects not our but Portes and associates’ view. We encourage interested readers to read the original work in this area.

We thank the reviewer who points out that SES differentials’ theoretical hindering of oppor- tunity for intergroup contact is only one of their many possible effects on ethnic factors. These differentials, for example, might indicate discrimination. The second conceptualization suggests that greater SES differentials mean higher levels of discrimination, which, according to the ethnic competition thesis, creates higher, not lower, ethnic identity levels. However, the first approach follows Portes and associates’ (Portes, Parker, and Cobas 1980; Portes 1984; Portes and Bach 1985) prediction of higher ethnic identity among higher SES minority members.

Some of these contrasts are “ideal typical” and involve some exaggeration for clear delinea- tion of the difference between the two perspectives.

Researchers interested specifically in ethnic language retention (e.g., Stevens and Swicegood 1987) consider language spoken at home a salient indicator of ethnic identity. Because language clearly marks group boundaries, a language shift indicates not only objective but subjective aban- donment of culture-based group identity. Given the shared indicator, findings presented here can address specific issues in ethnic language retention.

The lack of a segregation effect may also stem from our segregation measure. The exposure index (P*) (Massey and Denton 1988) might obtain different findings. We cannot compute this measure, however, and thus cannot test this possibility, left for future studies.

2.

3.

4.

5.

REFERENCES

Aguine, Benign0 E., Rogelio Saenz, and Sean-Shong Hwang. 1989. “Discrimination and the Assimilation and Ethnic Competition Perspectives.” Social Science Quarterly 70: 594-606.

Alba, Richard D., and Mitchell B. Chamlin. 1983, “A Preliminary Examination of Ethnic Identifi- cation Among Whites.” American Sociological Review 48: 240-247.

Blau, Peter M. 1977. Inequality and Heterogeneity. New York: Free Press. Denton, Nancy A., and Douglas S. Massey. 1989. “Racial Identity Among Caribbean Hispanics:

The Effect of Double Minority Status on Residential Segregation.” American Sociological Review 54: 790-808.

Glazer, Nathan, and Daniel P. Moynihan. 1970. Beyond the Melting Pot. 2nd ed. Cambridge: MIT Press.

Gordon, Milton M. 1964. Assimilation in American Life. New York Oxford University Press. Gordon, Robert A. 1968. “Issues in Multiple Regression.” American Journal of Sociology 73: 592-

616.

Page 8: ETHNIC ENCLOSURE OR ETHNIC COMPETITION: Ethnic Identification Among Hispanics in Texas

476 THE SOCIOLOGICAL QUARTERLY Vol. 32/No. 3/1991

Greeley, Andrew M. 1971. Why Can’t They Be Like Us? American White Ethnic Groups. New York: E.P. Dutton.

Hanushek, Eric, and John E. Jackson. 1977. Statistical Method for Social Scientists. New York: Academic Press.

Hwang, Sean-Shong, and Steve H. Murdock. 1988a. “Population Size and Residential Segregation: An Empirical Evaluation of Two Perspectives.” Social Science Quarterly 69: 813-834. . 1988b. “Residential Segregation and Ethnic Identification Among Hispanics in Texas.” Urban Affairs Quarterly 23: 329-345.

Hwang, Sean-Shong, Steve H. Murdock, Banoo Parpia, and Rita Hamm. 1985. “The Effects of Race and Socioeconomic Status on Residential Segregation in Texas, 1970-80.” Social Forces 63: 732-747.

Marshall, Harvey, and Robert Jiobu. 1975. “Residential Segregation in United States Cities: A Causal Analysis.” Social Forces 53: 449-460.

Marston, Wilfred G., and Thomas L. Van Valey. 1979. “The Role of Residential Segregation in the Assimilation Process.” Annals 441: 13-25.

Massey, Douglas S., and Nancy A. Denton. 1985. “Spatial Assimilation as a Socioeconomic Outcome.” American Sociological Review 50: 95- 105. . 1987. “Trends in the Residential Segregation of Hispanics, Blacks, and Asians: 1970- 1980.” American Sociological Review 52: 802-824. . 1988. “The Dimensions of Residential Segregation.” Social Forces 67: 281-315.

Massey, Douglas S., and Brandan P. Mullan. 1984. “Processes of Hispanic and Black Spatial Assimilation.” American Journal of Sociology 89: 836-873.

Portes, Alejandro. 1984. “The Rise of Ethnicity : Determinants of Ethnic Perceptions Among Cuban Exiles in Miami.” American Sociological Review 49: 383-397.

Portes, Alejandro, and Robert L. Bach. 1985. Latin Journey: Cuban and Mexican Immigrants in the United States. Berkeley: University of California Press.

Portes, Alejandro, Robert N. Parker, and Jose A. Cobas. 1980. “Assimilation or Consciousness: Perceptions of the U.S. Society Among Recent Latin American Immigrants to the United States.” Social Forces 59: 201-224.

Stevens, Gillian, and Gray Swicegood. 1987. “The Linguistic Context of Ethnic Endogamy.” American Sociological Review 52: 73-82.

U S . Bureau of the Census. 1983. Census of Population and Housing, 1980: Summary Tape File 4B (Texas). Machine readable data file. Washington, DC: US. Government Printing Office.