University of Utah

The Role of Group in Latino Public Opinion Author(s): Gabriel R. Sanchez Source: Political Research Quarterly, Vol. 59, No. 3 (Sep., 2006), pp. 435-446 Published by: Sage Publications, Inc. on behalf of the University of Utah Stable URL: http://www.jstor.org/stable/4148044 Accessed: 01/07/2010 10:59

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://dv1litvip.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=sage.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

University of Utah and Sage Publications, Inc. are collaborating with JSTOR to digitize, preserve and extend access to Political Research Quarterly.

http://dv1litvip.jstor.org The Role of Group Consciousness in Latino Public Opinion

GABRIELR. SANCHEZ,UNIVERSITY OFNEW MEXICO

The public opinion of the Latinocommunity is an understudiedarea within the political science literature. This analysiscontributes to this literatureby investigatingthe role of groupconsciousness across both Latino salientand generalpolicy areasutilizing the 1999 WashingtonPost/Henry J. KaiserFamily Foundation National Surveyof Latinos.By includingboth issue areasthat are salientto the Latinocommunity (immigration, bilin- gual )as well as those that are not directlytied to Latinos(abortion, death penalty),I test the pri- maryhypothesis that group consciousnesshas a greaterimpact on Latinopolitical attitudes across issues that are directlytied to ethnicitythan on those that are not. Resultsfrom this analysissupport the overalltheory, as perceiveddiscrimination motivates public opinion towardboth immigrationand bilingualeducation, and collectiveaction towardimmigration. Among other factors,nativity and the length of lived in the U.S. have the greatestinfluence on Latinopublic opinion.

LATINO PUBLIC OPINION about the political opinions of the Latinocommunity in the U.S. This is a result of a general disinterest in Our of Latino public opinion is much less the attitudesand opinions of Latinos prior to the late 1980s developed than that of Anglo public opinion. In , the (de la Garza 1987). The more recent interest in the political first survey focused on the political attitudes of Latinoswas attitudesand beliefs of Latinoshas been driven by the rapid not conducted until 1979, and that data only included Lati- of growth of this population, which has generated interest nos Mexican origin (Arce 1979). Despite this obvious of within both the political and marketing industries. This obstacle, scholars Latino politics have generated some analysis is an effort to contribute to the growing knowledge important insights in the area of Latino public opinion. of Latino public opinion by investigating the relationship Research has indicated that Latinos generally support an between group consciousness and the public opinion of activist government that protects minority civil and Latinosacross both generaland Latinosalient issue areas.By provides opportunities for individual citizens and minority including both general policy issues (death penalty, abor- groups (Anne Martinez2000). However, Latinos also tend tion) and those more salient to Latinos (immigration,bilin- to support the death penalty at higher rates than Whites or gual education), I test the primary hypothesis that group African Americans, and are also slightly more opposed to consciousness has a greaterimpact on Latino political atti- abortion than Whites (Uhlaner and Garcia 2002). This tudes when the issues are directly tied to ethnicity Group trend suggests that the foundations of Latinopublic opinion consciousness is a multi-dimensional developed are complex and diverse, motivating a desire to conduct in I when members of a group recognize their status as more research this area. intend to shed some light on the part of a deprived group (Milleret al. 1981). complex of Latino public opinion by investigating To investigatethe relationshipbetween group conscious- Latino policy attitudes across several issue areas. An issue often as ness and Latinopublic opinion this investigationutilizes four used of Latinoshaving a con- separateempirical models, one for each issue area;abortion, servative stance on social issues is abortion, the first non- death penalty,bilingual education, and immigration.The data salient issue area used in the analysis. Latinosare about ten used in this investigationoriginates from the 1999 survey of percentage points more likely than Anglos to oppose abor- 2,417 Latinosconducted by the WashingtonPost, The Henry tion (Leal 2004). However, multivariateanalysis has indi- that is J. Kaiser Foundation, and Harvard University A cated there no statistical difference between Whites reviewof the extant literatureon the role of group conscious- and Latinos regarding attitudes directed toward abortion ness in the formationof political attitudeswill follow a dis- (Leal 2004; Bolks et al. 2000). We also know from previous cussion of the currentliterature of Latino public opinion to work that religiosityincreases pro-life attitudes,as do higher provide a theoreticalfoundation for this investigationof the socioeconomic levels and being female (Bolks et al. 2000). impact of group consciousnesson Latinopolicy preferences. Further, opposition to abortion is strongest among the Spanish dominant and Puerto Ricans (Leal 2004). The second generalpolicy areaused in this analysisis the NOTE:I wouldlike to thankJohn Garcia, Barbara Norrander, Brad Jones, death penalty The only extant work focused on Latinoatti- andLaura Langer for their constructive comments on earlierdrafts tudes toward the death has found that both Latinos of thisarticle. penalty and African Americans express greater opposition to the PoliticalResearch Quarterly, Vol. 59, No. 3 (September2006): pp. 435-446 death penalty than do Whites, with opposition being greater

435 436 POLITICALRESEARCH QUARTERLY for women, the Spanish dominant, and non-citizens (Leal the Latino community that there are already too 2004). Thelack of extantresearch exploring the attitudesof many immigrantscoming to the U.S. annually.Actually, the Latinostoward abortion and the death penalty motivates the percentage of Mexicans and Puerto Ricans who believe there inclusion of these issues in the analysis, as well as the deci- are too many immigrantsin the U.S. is greaterthan the per- sion to treat the death penalty and abortion as non-salient centage of Whites (de la Garzaet al. 1990).1 These general policy areas. This analysis will provide some insight into trends have been reinforced by other surveys of Latinos Latino attitudes in these areas as well as the factors that (Uhlaner and Garcia 2002; Binder, Polinard, and Wrinkle influence those attitudes. 1997). For example, only 23 percent of Mexicans believe that more illegal immigrantsshould be allowed to enter the LATINOSALIENT ISSUES U.S., and 73 percent of Mexicans believe that immigration (BILINGUALEDUCATION AND IMMIGRATION) should be more strictly enforced (Binder,Polinard, and Wrinkle 1997). Further,Latinos are evenly divided on their The issue areas that generate the most divergence from opinions of the use of sanctions for employerswho hire ille- the general population are those with a cultural or ethnic gal immigrants(Cain and Kiewiet 1987). basis. Generally speaking, Latinos express a strong desire to Severalfactors have been found to influence the attitudes protect their cultural traits and , specifically the of Latinos toward immigration. For example, the native Spanish language. Although variationexists regardingform born are more likely to favor a more restrictivepolicy than of bilingual education program, overall bilingual education are the foreignborn (Binder,Polinard, Wrinkle 1997), as are helps Spanish-speakingindividuals retain their native lan- wealthier,more educated, and older Latinos(Hood, Morris, guage, thus contributing greatly to the maintenance of and Shirkey 1997; Binder, Polinard, and Wrinkle 1997). Latino (Houvouras 2001). This is reflected by lan- Further,non-citizens are more concerned with immigration guage policy being the issue area that has the highest level issues than citizens (Michelson 2001), and citizens and of consensus among Latinos and the issue that separates those who speak primarilyEnglish were more likely to sup- Latinos' political attitudes from that of other racial/ethnic port Proposition 187 in California(Newton 2000). Among groups (Uhlaner and Garcia2002). sub-groups, Cuban Americans are most concerned with Surveys of the Latino population consistently indicate illegal immigration,and MexicanAmericans are more likely that most Latinossupport bilingual education, in sharp con- to believe that the government is alreadydoing too much to trast to Whites and African Americans (Uhlaner 1991; stop illegal immigration(Michelson 2001). Hajnal and Baldassare 2001). For example, the Latino The most prominent for these trends is that National Political Survey (LNPS) (de la Garza et al. 1989- many Latinos are more concerned with the economic 1990) indicated that 80 percent of Mexican Americans, 87 impact of immigrationdue to their overall lower socioeco- percent of Puerto Ricans, and 89 percent of Cuban Ameri- nomic status, which may place them in direct competition cans support bilingual education, with majorities in each for jobs with immigrants (Rodriguez, and Nunez 1986; sub-group reportingthat they would be willing to pay more Gutierrez 1995; Polinard, Wrinkle, and de la Garza 1984; taxes to support the program.Although variationdoes exist de la Garza, 1998). In line with this notion, Chicanos from based on national origin and citizenship status, over two Texas who were higher educated were often more sympa- thirds of the least favorable group (U.S. born Mexican thetic to the plight of illegal immigrantsdue to a lack of job American citizens) supported bilingual education in the threat (Rodriguez, and Nunez 1986). Contraryto the eco- LNPS (Uhlaner and Garcia, 2002). Interestingly,despite nomic based explanation for immigration attitudes among similarities in language use patterns, Latinos display more Latinos, Newton (2000) found that ethnic meas- enthusiasm for bilingual education than Asians (Cain and ured by languageuse and citizenshipstatus accounted for Kiewiet 1987). differences in support for Proposition 187 in California. Despite clear support for bilingual education, over 90 Specifically,English speakers and citizens were more likely percent of Latinosin the LNPS agreed that all citizens and to support the anti-immigrationproposition. residentsof the U.S. should learn English. Further,less than It is importantto note however that despite concern over 10 percent of Latinos believe that the primary purpose of increased immigration, Latinos are generally supportive of bilingual education should be to maintainSpanish language recent immigrantsonce they have arrivedinto the U.S. (de la or culture (Schmidt 1997). Bilingualismappears to be the Garzaet al. 1991). For example, Latinosare more likely than desired goal for Latinos, as over 70 percent of each sub- Whites, AfricansAmericans, or Asians to support amnesty group indicated that learning two languages is the primary for illegal immigrants (Cain and Kiewiet 1987). Further, objective of bilingual education (ibid.). Mexicansin general are more likely to support policies that In addition to bilingual education, immigration is facilitate the political and social integration of Mexican included as a Latinosalient policy Despite the common per- immigrants, such as bilingual education and immigrant ception that Latinos support a very liberal immigration policy, analysis of Latino public opinion suggest that sup- 1 port for a relaxed immigration policy is not widespread The 73.8 percent of Anglos was lower than that of both Mexicans75 per- among the Latinopopulation. In fact, a sizable percentageof cent and PuertoRicans 79 percent in the LatinoNational Political Survey. THE ROLEOF GROUP CONSCIOUSNESSIN LATINOPUBLIC OPINION 437

access to services and citizenship (de la Garza1998). Clearly Hi: if groupconsciousness is relevantfor Latinos,it's effects Latino public opinion is a complex phenomenon that has will be morepronounced in the contextof policy areas been understudiedto this point in political science. It is my salientto the Latinoorigin population. intention to clarifythe role of group consciousness and other relevantfactors in the formationof Latinopolitical attitudes. H2: the dimensionsof groupconsciousness will be positively correlatedwith the pro-Latinostance on both Latino THE ROLEOF GROUPCONSCIOUSNESS IN salient issue areas:support for bilingualeducation and MINORITYPUBLIC OPINION increasedimmigration to the U.S..

Groupconsciousness is defined as instanceswhen a group Although this analysis is focused primarily on the rela- maintains a sense of affinity and group identificationwith tionship between group consciousness and Latino public other members of the group, which leads to a ori- opinion, there are other factorsthat are potentially relatedto entation to become more politically active (Garcia 2003). the political attitudes of Latinos. Literatureassociated with There is little work investigatingthe relationship between Latino political behavior suggests that cultural factors are group consciousness and public opinion among minority importantcontributing factors to Latinosalient policy areas. groups. Further, this small body of research has focused It is thereforeanticipated that almost exclusivelyon the AfricanAmerican population, moti- vating the necessity to draw from investigationsof the role of H3: theforeign born, non-citizens, and thosewho have spent group consciousness in AfricanAmerican public opinion to less timein the U.S. will supportincreased immigration develop a drivingtheory for this analysis.The most directtest to a greater extent than other Latinos due to recent of the relationshipbetween group consciousness and public arrivingimmigrants potentially having greater personal opinion is MichaelDawson's (1994) work exploring the role experienceswith the currentimmigration policies in the of group interests in African American policy preferences. U.S., and also being more likely to have friends and Dawson argues that both economic and racialpolicies have family who would benefitfrom a more relaxed U.S. been tied historically to African American group interests. immigrationpolicy. Group interestshave a significantimpact on AfricanAmeri- can views towardeconomic redistributionpolicies, as support Similarly,it is hypothesized that for redistributivepolicies are greatest among those whose perceived link to other AfricanAmericans is high. Further, H4: Latinoswho are morerecent arriving immigrants and are perceptionsof linked fate, an indicatorof group conscious- less proficientin Englishwill expressgreater support for ness played the greatestrole in predicting support for gov- bilingualeducation due to the actual and/or perceived ernment racialpolicies in Dawson's(1994) analysis. Overall, benefits this policy providesto Latinoswith this back- group interests help counterbalanceclass divisions among ground. AfricanAmericans in regardto policy preferences,particu- larlywhen the policy areais raciallydriven. Among the two non-salient issue areas of the death penalty Katherine Tate (1993) suggests that the and abortion, the contributing factor expected to be most among AfricanAmericans that there are inequalities based critical is Catholicism. Given the moral natureof both abor- on race in the U.S. motivates racial identity among African tion and the death penalty,I hypothesize that Americans,which leads to group interests dominating Black policy preferences.Tate finds that class influences this rela- H,: Catholicswill be morelikely than non-Catholics to have tionship, as AfricanAmericans who identify with the upper a pro-lifestance in the contextof bothabortion and the economic classes are less likely to have strong racial identi- deathpenalty. ties (Tate 1993). Finally,while there is little to draw from in this areaspecific to Latinos,cultural affinity has a significant TESTING STRATEGY/MODELSPECIFICATION influence on immigration policy, as the more "Mexican"a respondent is, the less likely they are to support restrictive As previously stated, the primarydata source utilized in immigrationpolicies (Binder, Polinard, and Wrinkle 1997; this analysis is the 1999 nationally representativesurvey of de la Garzaet al. 1991). 2,417 Latinosconducted by the WashingtonPost, the HenryJ. Kaiser Family Foundation, and Harvard University. This HYPOTHESES survey was conducted by telephone between June 30 and August 30, 1999, by the International Communications The primarytheory that drives this analysis is that group Research Firm. All interviews were conducted using the consciousness influences the political attitudes of Latinos, Computer Assisted Telephone Interviewing(CATI) system. particularly when the policy area is directly tied to the The CATIsystem ensuredthat questions followed logical skip Latino community Group consciousness is a resource that patterns and that the listed attributesautomatically rotated, generates political activity through an individual's attach- eliminating "question position" . Survey respondents ment to a group. Therefore,I theorize that were selectedat random,and the marginof samplingerror for 438 POLITICALRESEARCH QUARTERLY

respondentsis 2 percent for Latinorespondents. The survey tions (assimilation,political interest, ,partisanship), includes 818 Mexicans,318 PuertoRicans, 312 Cubans,and (4) political activities and experiences (Latino specific par- 593 Central or South Americans. The final results were ticipation, discrimination experience), (5) cultural factors weighted to the national Latino population to ensure that (nativity, citizenship status, English proficiency, length of nationalitywas representativeof actualproportions. time in U.S.), and (6) national origin-Cuban, Central/ South American, CaribbeanLatinos. DEPENDENTVARIABLE MEASUREMENT STRATEGY A discussion of the variableconstruction for the primary variablecluster of group consciousness will precede statisti- This investigation of Latino public opinion utilizes four cal analysis. The survey items utilized to construct the policy areas as dependent variables,two general issue areas remainingindependent variablesare included in the Appen- (abortion and the death penalty), and two Latino salient dix. A correlation matrix was created to test for multi- issues (bilingual education and immigration).This strategy collinearity among all explanatory variables. The two will allow me to determine whether or not group con- explanatory variables with the greatest correlation in the sciousness has a greater impact on policy areas that are analysis are citizenship status and nativity, with a Pearson directed primarilyat Latinos.The abortion variableis based coefficient of .61.2 on the following survey item: Do you thinkabortion should be legalin all cases,legal in mostcases, illegal in mostcases, or ille- GroupConsciousness Cluster gal in all cases?The categories of the abortion variablewere recoded to represent whether or not individuals believe The concept of group consciousness suggests that the abortion should be: legal in all or most cases 0, illegal in effects of group affinity and collective orientations are felt most or all cases 1. The death penalty variable is derived within Latino sub-groups (Puerto Rican, Columbian, Mexi- from the following survey question: Do you favor or oppose can etc.), as well as the broader pan ethnic grouping of thedeath penalty for personsconvicted of murder?This variable Latino. I agree with Milleret al. (1981) that proper concep- is dichotomous with 0 for favorand 1 for oppose as the two tualizationof group consciousness requiresthe categories of the variable. of multiple measures to tap into the main dimensions of The two Latino salient issues of bilingual education and group consciousness. Past literaturesuggests that there are immigration are included due to their close connection to three dimensions of group consciousness; general identifi- the Latino community and the fact that both issues are hot cation with a group, an awareness of that groups relative topics of debate generallyin the U.S.. The bilingual educa- position in society, and the desire to engage in collective tion variableutilizes the following survey item: Do you think activity that focuses on improving the situation of that all publicschool classes should be taughtin Englishor do you group (Gurin Miller 1980; Padilla 1985; Garcia2003). For- thinkchildren of immigrantsshould be able to takesome courses tunately,the Post/Kaisersurvey provides the opportunity to in theirnative language? There are two responses to this ques- capture all three aspects of group consciousness for Latinos. tion, taught in English 0, take courses in their native lan- This attempt to account for the multidimensional nature of guage 1. To measure the attitudes of Latinos toward immi- group consciousness advances previous researchinterested gration, the following survey question was used as an in the role of group consciousness in minority political indicator:Do you thinkthe numberof new immigrantsallowed behavior that has typically relied on only group identity to into the U.S. eachyear shouldbe increased,decreased, or kept measure group consciousness (Olsen 1970; Verba and Nie about the same?There are three categories for this variable; 1972; Padilla 1985; Uhlaner 1989). decreased 0, same 1, increased 2. Both Latino salient meas- To measure group identity in this analysis, a group com- ures were re-coded to make pro-Latinoresponses (courses monality index was createdusing a battery of questions that in native language, new immigrants increased) the high asked respondents how much he or she felt in common variablevalue. with other Latino sub-groups.3 Respondents were given a score based on their response to the of questions. For INDEPENDENTVARIABLE MEASUREMENT example, a response of a lot in common received +2 points, a fair amount in common +1, only a little in common -1, My general discussion of the public opinion literature and in common -2. These scores were used to con- has identified several factors that may impact the political struct an index that consists of seven values running from attitudes of Latinos. In addition to those factors identified by previous literature,there are several additional 2 that may have an impact on the public opinion of Latinos. These measureswere tested individuallyand collectively through the use This array of factors is grouped into six clusters represent- of a scaled measure with no change in statisticalsignificance or direction ing the differentperspectives attempting to explain partisan in any of the models. No other variable combinations approach .6. The behavior.These clusters are (1) group consciousness (Latino questions ask: Do youfeel (insertrespondents nationality group) have a lot, afair amount,or in commonwith Mexicans, commonality,shared political interests, perceived discrimi- nothing thefollowinggroups; Cubans,Puerto Ricans, and Central/SouthAmericans. Cronbach's Alpha sta- nation), (2) SES/Demographicfactors (income, education, tistic of .876 indicates with great confidence that these survey questions work status, gender, age, Catholicism),(3) political orienta- can be scaled to create the Latino Commonalityvariable. THE ROLEOF GROUP CONSCIOUSNESSIN LATINOPUBLIC OPINION 439 no sense of Latino commonality to a strong sense of com- Although neither income nor education had a significant monality with all Latinosub-groups.4 In addition to general impact here, both men and Democrats are more likely to commonality,a measure of Latino political commonality is oppose the death penalty. Catholicismis statisticallysignifi- also included in the group consciousness cluster. The polit- cant and as expected contributes to greater opposition to ical commonality measure was based on responses from the the death penalty The pro-life stance of the Catholic Church following survey question: Do you agreeor disagreewith the seems to contribute to this trend among Latinos who are following statement?Latinos in the Untied States share FEW Catholics. In addition, opposition to the death penalty is politicalinterests and goals?The two values for this variable greater among individuals who have participatedin multi- are: no commonality (agree) 0 and commonality (disagree) ple Latino specific activities. 1. The addition of the measure tapping into the extent of Cultural factors a meaningful role in Latino policy common political interest among the Latino population views, as nativity and English proficiencyboth have signifi- allows for the distinction to be made between social and cant relationshipswith opposition to the death penalty Lati- political commonality, an advantage over other studies of nos born in the U.S. oppose the death penalty with greater group consciousness. likelihood than the foreign-bornsegment of the Latinopop- Beyond notions of commonality, group consciousness ulation, as do those who have greaterproficiency with the requiresthat individuals recognize that their group shares a English language.Finally, attitudes towardthe death penalty disadvantaged position in society I employ a measure of vary across national origin groups, as opposition to the perceived discrimination to capture this component of death penalty is greateramong Cubans and CaribbeanLati- group consciousness based on responses from the following nos when compared to Mexican Americans and Central/ survey questions: Is discriminationagainst Latinos in our soci- South Americans. ety todaya problemor not?And, is it a big problemor not such The dependent variable for the second general policy a bigproblem? A three point scale is used as a measureof dis- area of the abortion model is also dichotomous, therefore criminationwith the following values 0 for those individu- logistic regression is utilized to estimate the model as well. als who believe discriminationis not a problem, 1 for those Of primary importance to this study, consistent with H1, who indicate that discrimination is a problem for Latinos none of the group consciousness dimensions have any but not a big problem, and 2 for those who believe that it is impact on Latinos' attitudes toward abortion. Therefore a big problem for Latinos. group consciousness has no impact on Latino attitudes that The final component of group consciousness is the desire are not salient to the Latino community Among SES and to improve the disadvantaged societal position of one's demographic variables, age and Catholicism are negatively group through collective action. I use the following survey correlated with abortion. Therefore, the odds of believing question as an indicator of one's that collective action that abortion should be legal in all or most cases decrease as can improve the groups position in society: Do you thinkthat Latinos become older. Consistent with H4, Catholics are if various Latino groups worked togetherpolitically Latinos significantly less likely to support the legality of abortion wouldbe betteroff, worseoff, or wouldn'tmake much difference? relativeto non-Catholics.Therefore the pro-life stance of the The values of this final component of group consciousness Catholic Church again has an impact on the attitudes of are 0 worse off, 1 no difference, 2 better off. The inclusion Latinos who identify themselves as Catholic, concerning of these four measureseffectively captures all dimensions of abortion. In addition, the negative coefficient on the ideol- group consciousness. ogy variable indicates that Latinos who identify themselves as ideologically conservative are more likely to believe that THE IMPACTOF CONTRIBUTINGFACTORS ON abortion should be illegal in all or most cases. GENERALPOLICY AREAS Among political activities and experiences, participating in political activities directly tied to the Latino community The first of this is to assess the influence of stage analysis increases that abortions should be illegal in all consciousness on areas in order to group general policy or most cases. Finally, the cultural factors of nativity and investigate any potential variation across issue areas. The length of time spent in the U.S. are significantin this model. firstmodel in this analysisutilizes a dichotomous variableto Latinos born outside of the U.S. are more likely to believe examine Latinos' attitudes toward the death penalty, and that abortion should be legal than Latinos born in the U.S. therefore logistic regression is used to estimate the model. This is potentially the result of the near universal illegality The categories for this variableare: favor the death penalty of abortion throughoutLatin America, a region where many for convicted murderers0, and oppose the death penalty for women are prosecuted for having abortions annually For- convicted murderers 1. Consistent with HI, none of the eign-born Latinosare potentiallymore likely to express sup- group consciousness variables are significantly related to port for the legality of abortion due to this hard-line this non-Latinosalient policy area. approachof most LatinAmerican countries. In addition, the odds of opposing the legality of abortion increasewith time lived in the U.S., which that to There is significantvariation in this measure,as just under 10 percent of implies greater exposure respondents are in the two highest commonality categories, while Americansociety increasesperceptions that abortionshould approximately30 percent are within the two lowest categories. be illegal. 440 POLITICALRESEARCH QUARTERLY

TABLE1 THEEFFECT OF THE FULL MODEL ON NON-SALIENTPOLICIES (LOGISTIC REGRESSION) Death Penalty Abortion B SE OR B SE OR GroupConsciousness Commonality -.027 .040 .972 -.041 .040 .958 Political Commonality .007 .129 1.00 -.003 .126 .996 PerceivedDiscrimination -.033 .090 .967 .097 .087 1.10 Collective Action -.121 .157 .885 -.215 .154 .806

SES/Demographics Income -.002 .004 .997 -.005 .004 .994 Education .027 .004 1.02 -.149 .039 .861 Work Status -.072 .293 .929 -.008 .279 .991 Gender .271** .125 1.31 .008 .123 1.00 Age .005 .007 1.00 -.014** .007 .986 Catholicism .594*** .151 1.81 .709*** .150 1.51 PoliticalOrientations Assimilation -.021 .066 .978 .061 .064 1.06 Political Interest -.038 .080 .962 -.039 .079 .961 Ideology .107 .082 1.11 -.359*** .081 .697 Partisanship .473*** .144 1.37 -.095 .139 .909

PoliticalActivities/Experiences LatinoSpecific Participation -.099 .077 .905 .030*** .076 1.03 DiscriminationExperience -.028 .081 .972 -.063 .079 .938 CulturalFactors Nativity .642** .219 1.90 -.838*** .215 .432 Citizenship Status -.039 .188 .961 -.261 .187 .769 English Proficiency .181** .096 1.19 .126 .095 .880 Length of Time in U.S. .009 .008 1.00 .018** .008 1.01

NationalOrigin Cuban .699*** .209 2.01 -.483** .200 .616 Central/SouthAmerican .144 .166 1.15 .014 .168 1.01 CaribbeanLatinos .393 * .202 1.48 -.419** .194 .657 N= 1193 N= 1230 Log Likelihood = -750.35834 Log Likelihood = -773.214 Pseudo R-square = .070 Pseudo R-Square = .093

Note: * p < .10 level; ** p < .05 level; *** p < .01 level in a two-tailed test.

Interestinglythere is variation across national origin, as consciousness and other factors on the Latino salient policy Cuban Americans and Central/SouthAmericans are more areas of immigrationand bilingual education. The immigra- likely to believe that abortions should be legal in all or most tion variable has three categories: immigration should be cases. This trend among Cubans is consistent with Leal's decreased 0, kept the same 1, increased 2. Given the (2004) findings regardingnational origin. Therefore,Latino ordered and categoricalnature of this variable,Generalized attitudes toward abortion are spilt across several factors, OrderedLogit (GOL)is used to estimate this model. GOLis including national origin, nativity, and religious preference. preferredhere due to its ability to handle the proportional odds assumption associatedwith the OrderedLogit Model.5 THE IMPACTOF CONTRIBUTINGFACTORS ON Much like Multinomial Logit, the GOL model provides LATINOSALIENT POLICY AREAS

With the two Latino models now the general specified, 5 Generalized Ordinal Logit was employed after Ordinal Logit models next step in this analysis is to determine the role of group failed Hausman tests for the proportionalodds assumption. THEROLE OF GROUPCONSCIOUSNESS IN LATINO PUBLIC OPINION 441

= TABLE2 THEEFFECT OF THE FULL MODEL ON IMMIGRATION(GENERALIZED ORDERED LOGIT) ImmigrationUnchanged IncreasedImmigration B SE OR B SE OR GroupConsciousness Commonality .076 .055 1.07 .044 .046 1.04 Political Commonality -.047 .168 .953 .044 .147 1.04 PerceivedDiscrimination .160 .112 1.17 .373*** .109 1.45 Collective Action .344** .186 1.41 -.042 .180 .958

SES/Demographics Income .016* .009 1.01 .003 .004 1.00 Education .008 .055 .991 -.023 .046 .977 Work Status .223 .338 1.25 .064 .324 1.06 Gender .063 .164 1.06 .000 .144 1.00 Age -.005 .011 .994 .019** .007 1.02 Catholicism .004 .191 1.00 -.163 .173 .849 PoliticalOrientations Assimilation -.119 .080 .887 -.043 .077 .957 PoliticalInterest -.064 .113 .937 -.009 .090 .990 Ideology -.039 .109 .960 -.033 .093 .966 Partisanship .161 .180 1.17 .254 .166 1.28

PoliticalActivities/Experiences LatinoSpecific Participation .402*** .110 1.49 .311*** .086 1.36 DiscriminationExperience -.127 .105 .880 .139 .091 1.14 CulturalFactors Nativity -.525* .307 .591 -.202 .257 .816 Citizenship Status .003 .302 1.00 -.593** .207 .552 English Proficiency -.256* .144 .773 -.051 .104 .950 Lengthof Time in U.S. -.023* .012 .977 -.025** .009 .975

NationalOrigin Cuban .246 .283 1.27 .478** .220 1.61 Central/SouthAmerican .336 .261 1.39 -.242 .188 .784 CaribbeanLatinos -.512** .220 .598 -.200 .243 .818

N = 1204 Log Likelihood = -1045.5321 Pseudo R-square = .100

Note: * p < .10 level; ** p < .05 level; *** p < .01 level in a two-tailed test results for each category in the dependent variable other Latino collective action support immigration numbers than the base category,in this case immigration should be remainingthe same versus being decreased.Among the SES decreased . However, with GOL results are relative to all cluster, Latinoswith higher incomes are more likely to sup- lower categories, not just the baseline. I will begin this sec- port immigration remaining the same, although the odds tion of the analysis with the variable category of immigra- ratio suggests that this relationshipis weak at best. Partici- tion being unchanged, with results being interpreted rela- pating in Latino specific activities increases the odds that tive to the baseline of immigrationbeing unchanged. Latinos will support increased immigration to the U.S.. In all, there are seven contributingfactors that have a sta- Working for a Latino organizationor candidate potentially tistically significant relationship with the belief that immi- strengthens the bond to other Latinos, many of whom may gration should remain unchanged. Most important to this be potential recent immigrants.In other words, it is - study, the group consciousness dimension of collective able to assume that if you are working for a Latino organi- action is significantly related to this category of the immi- zation you will be involved in activities with Latino immi- grationvariable. Therefore Latinos who perceive benefits to grants, as approximately 40 percent of the Latino 442 POLITICALRESEARCH QUARTERLY

TABLE3 for other immigrants. Finally, Latinos of Caribbean origin THEEFFECT OF THE FULL MODEL ON BILINGUAL EDUCATION are less likely to indicate that immigration should remain (LOGISTICREGRESSION) unchanged rather than being decreased. This is potentially B SE OR a result of Puerto Rican and Dominicans supporting greater efforts to curtail immigration as a result of viewing recent GroupConsciousness arrivingimmigrants as economic competitors. Commonality .023 .040 1.02 The second set of results report the impact of contribut- Political Commonality .015 .127 1.01 ing factorson Latinosindicating that immigrationshould be Perceived Discrimination .379*** .086 1.46 increased. These statistics are taken relativeto the two lower Collective Action .164 .147 1.17 categoriesof the immigrationvariable, unchanged immigra- tion to the U.S. and decreased immigration. Six contribut- SES/Demographics are in this context. Within the Income -.001 .004 .998 ing variables significant consciousness discrimination is Education -.005 .040 .994 group cluster, perceived correlatedwith increased Therefore Work Status -.296 .291 1.13 positively immigration. individuals who believe discriminationis a for Gender -.029 .124 .971 big problem Latinosin the U.S. are more to the notion that .012* .007 1.01 likely support Age more should be allowed to come to the Untied Catholicism .157 .147 1.17 immigrants States. This trend along with the significance of collective PoliticalOrientations action in the previous variablecategory provides support for Assimilation .017 .065 1.01 the primary hypothesis that group consciousness is a con- Political Interest -.052 .080 .949 tributing factor to Latino public opinion, particularlywhen the issues are salient to the Latino Ideology .091 .081 1.09 community. Partisanship .322** .139 1.38 Age also increases the odds that Latinos will support a more open immigration policy, as older Latinos are more PoliticalActivities/Experiences likely to believe that the number of immigrants allowed to Latino Specific Participation .248** .077 1.28 come to the U.S. should be increased. Further,participation DiscriminationExperience .032 .079 1.03 in Latino specific political activities is again positively cor- related with for increased cul- CulturalFactors support immigration.Among tural factors, although nativity and English proficiency are Nativity .480** .218 1.61 no longer significant, Citizenship status is negatively corre- Citizenship Status .185* .192 1.20 lated with support for increased immigration.Therefore, as -.348*** .097 .706 English Proficiency hypothesized non-citizens are more likely to believe that Length of Time in U.S. -.036*** .009 .964 more immigrants should be allowed to come to the U.S. Also Latinos who have lived in the U.S. for NationalOrigin supporting H2, Cuban -.052 .227 .949 a shorter period of time are again more likely to support Central/South American -.519*** .168 .594 increased immigration. These last two findings are consis- CaribbeanLatinos -.090 .195 .913 tent with previous research (Binder,Polinard, and Wrinkle 1997) and are interpreted as being a function of non-citi- N = 1235 zens and Latinos with limited residency in the U.S. being Log Likelihood = -770.88 immigrantsthemselves. Pseudo R-square = .090 Finally,consistent with Martinez(2000), Cuban Ameri- Note: * P < .10 level; ** P < .05 level; *** P < .01 level in a two tailed testt cans are more likely to support increasedimmigration to the U.S. Cubans, who have experienced a much different his- torical immigrationexperience as a result of being granted population is foreign born. In addition, given the tendency refugee status by the U.S. government, support a more of Latino organizationsand candidates to focus on making relaxed immigration policy to a greater extent than other immigrationmore open, Latinosworking in these areasmay Latinos. This is potentially due to Cubans having less per- be more likely to support policies aimed at making immi- ceived or actual fear of economic competition with immi- gratingto the U.S. easier. grants as a result of their general higher socioeconomic Culturalfactors are again important contributingfactors levels relative to other Latinos. to Latino public opinion. As hypothesized, Latinoswho are The next model utilized in this analysis focuses on the foreign-born,less proficientin English, and who have spent policy area of bilingual education. There are two categories less time in the U.S. are more likely to believe that immi- in the bilingual education measure, students should be gration policies should remain the same relative to immi- taught in English 0 or taught in their native language 1. gration being decreased. The role of cultural factors here Given the dichotomous nature of this measure, logistic suggests that Latinoswho have immigratedto the U.S. more regressionis used to estimate this model. Among group con- recently oppose the limitation of immigrationopportunities sciousness variables, perceived discriminationis again the THE ROLEOF GROUPCONSCIOUSNESS IN LATINOPUBLIC OPINION 443

= FIGURE1 THE IMPACTOF PERCEIVEDDISCRIMINATION ON IMMIGRATION 1

0.9

0.8

0.7

0.6

• 0.5

0.40

0.3.

0.2

0.1

0 1 2 3 PerceivedDiscrimination Values 1) Not a Problem2) A problem3) Big Problem

only dimensionto have an impact on bilingualeducation. As dren with language issues into English dominant class- perceiveddiscrimination increases, so do the odds that Lati- rooms. Conversely,citizens may support the use of Spanish nos will believe children of immigrantsshould be taught in for those with language barriersdue to its perceived bene- their native languages.Results from all four models indicate fit to Latinos generally.More in line with H2, both English that perceiveddiscrimination is the dimension of group con- proficiency and length of time in the U.S. are negatively sciousness that has the greatestinfluence on Latinopolitical correlatedwith support for bilingual education. This indi- attitudes.In addition it is clear that group consciousnesshas cates that Latinos who have lived in the U.S. for shorter a greaterrole in policy preferencesof Latinoswhen the issues periods of time and who are less proficient in English are are directlytied to the Latinocommunity. more likely to believe it is better for children of immigrants In additionto perceiveddiscrimination, both age and par- to be taught in their native language than to be taught tisanshipare significantlyand positivelycorrelated with sup- exclusively in English. port for bilingual education. Older Latinos and those who Among national origin variables,being Centralor South identify as Democratsare more likely to support the notion Americanincreases the odds of believing children of immi- that children of immigrantsshould be taught in their native grants should be taught exclusively in English. This trend language.Among political activitiesand experiences,Latino for Centraland South Americansagain supports the theory specific participationis again a contributorto Latinosalient used to explain nativity and citizenship status that recent policy support, as defense of bilingual education rises with arriving immigrants may desire that their children learn more political activitydirectly tied to the Latinocommunity. English in school due to the economic benefits associated Based on coefficients and odds ratios, cultural factors with English language proficiency.This suggests that eco- appear to be the most critical in determining support for nomic concerns may be more critical to this segment of the bilingual education, as all four variables in the cultural Latino population than cultural maintenance. In addition, factor cluster are statisticallysignificant. Interestingly,both the desire to maintain English skills may be based on a citizens and Latinosborn in the U.S. are more likely to sup- desire to prevent their children from facing discrimination port bilingual education. This contradicts the theory pro- due to their inability to speak English well. This is an moted in H3 that Latinos who stand to benefit from bilin- important finding, as Latino National Political Survey did gual education would exhibit the greatest support for the not survey this segment of the Latinopopulation. policy. This is potentially the result of non-citizens and the foreign born having a strong desire to have their children CONCLUSIONSAND DISCUSSION learn English quickly due to the economic benefits of Eng- lish proficiency in the U.S. This would motivate greater This analysishas not only shed light on the relationship support for policies that attempt quickly to assimilatechil- between group consciousnessand Latinopublic opinion, but 444 POLITICAL RESEARCH QUARTERLY

- FIGURE 2 THE IMPACT OF PERCEIVED DISCRIMINATION ON BILINGUAL EDUCATION 1

0.9

0.8

0.7

0.6

0.3

0.2

0.1

0 1 2 3 Perceived Discrimination Values 1) Not a Problem 2) A problem 3) Big Problem

has also defined the role of other contributing factors in opinion. Integration into American culture and society Latinopolitical formation toward both Latinosalient clearly influences Latino political attitudes, as demon- and more general policy areas. While the four policy areas stratedby the statisticalsignificance of length of time spent used in this analysis(immigration, bilingual education, abor- in the U.S. across all four policy areas. In addition, Latinos tion, and the death penalty)are not exhaustive,they do pro- born outside of the U.S. appear to think differently about vide a strong backdropfor this analysis.The primaryques- policy issues than Latinos who were born in the U.S., as tion addressedin this investigationof Latinosattitudes is the nativity was also significant across all four issue areas. role of group consciousness across various types of issue Finally, as expected Catholicism played a meaningful role areas. Results suggest that perceived discriminationis the in shaping Latinopolitical attitudes across the general issue dimension of group consciousnessthat plays the greatestrole areas of abortion and the death penalty Catholics clearly in determining Latino public opinion. The impact of per- support a pro-life stance in both policy areas, being more ceived discriminationon motivating support for increased likely to oppose abortion and the death penalty for con- immigrationand bilingual education among Latino salient victed murderers. policies is displayedin Figures 1 and 2 which both provide In conclusion, this analysisin no way closes the door on the predictedprobability increases associated with perceived factorsthat influence Latinopublic opinion. It does however discrimination.In both cases it is clear that the predicted provide a very important look into the role of the various probabilityof supportingthe pro-Latinostance on the policy dimensions of group consciousness in formulating Latino issue increaseswith greaterperceived discrimination. This is politicalattitudes. It is clear fromthis analysisthat while lim- consistent with previous researchthat has found that per- ited to policy areasthat are directlyrelated to ethnicity,group ceived discriminationis a key contributorto other aspects of consciousness is meaningfulto this aspect of Latinopolitical Latino political behavior such as political participation behavior. This investigation also provides some valuable (Stokes 2003; Sanchez 2006) and coalition formationwith inferences about Latino public opinion more generally,in other groups (Kaufmann2003; Garcia2000; Uhlaner 1991). particularthat culturalfactors such as nativityand length of The inability of perceived discrimination or any other time in the U.S. are the primarycauses of variationin public dimension of group consciousness to have an impact on opinion within the Latino community. Yes, we still have both generalpolicy areassuggests that the role of group con- much to learn regardingthe various political attitudesof the sciousness in Latinopublic opinion is greaterwhen the issue Latinocommunity, and how those attitudesare formulated. areais salient to the Latinocommunity. However, with the increasing availabilityof public Relative to the other variable clusters, cultural factors opinion data focused on Latinos, the prospects for greater proved to have the greatest influence on Latino public insights in this researcharea is very promising. THE ROLEOF GROUP CONSCIOUSNESSIN LATINOPUBLIC OPINION 445

APPENDIX Discriminationexperience-During thelast 5 years, haveyou, PRESENTATIONOF SURVEYITEMS afamily member;or a closefriend experienceddiscrimination AND INDEPENDENTVARIABLE MEASURES becauseof your racialor ethnicbackground, or not?And, was thatyou personallyor was that someoneelse? The values of SocioeconomicStatus/Demographics the discriminationexperience variable are; 0) no experi- ence 1) indirect experience 2) direct experience. Household Income-What is your total householdincome from all sources,before taxes? The values of the income CulturalFactors measure are a nine-point income scale ranging from < 20 thousand to >100 thousand. Nativity-Were you bornin the UnitedStates or anothercoun- Education-What is the last that in grade you completed try?The values of nativity are 0) foreign-born 1) native school?The values of the education variableare; 0) <9Pthp born. some school 2) school 3) grade 1) high high graduate Citizenship Status-Now we wouldlike to askyou about U.S. vocational 4) some 5) 6) training college college graduate citizenship.Are you, a U.S. citizen,currently applying, plan- post graduatetraining. ning to apply,or not planningto becomea U.S. citizen?The Work Status-What is status?The values of your employment values of citizenship status are 0) non-citizen 1) citizen. the work statusvariable 0) 1) are; unemployed employed. English proficiency-This measure was created from the FinancialSituation-In recent has years, your personalfinan- following survey questions; Wouldyou say you can carry cial situation worse,or aboutthe gottenbetter, gotten stayed on a conversationin English?Would you say you can reada same?The values of the financialsituation variableare; 1) newspaperor book in English?The English proficiency worse about the same 3) better. 2) scale tunes from 0) non-proficient through 3) highly Gender-What is The values of the vari- yourgender? gender proficient able are 0) female 1) male. Time Spent in the United States-How manyyears haveyou is is continuous with the Age-What your age?Age youngest livedin the UnitedStates? Time spent in the United States 18 and the oldest 90. respondent being being is continuous, with age being used to maintain non-citi- Catholicism-What is The values of your religiouspreference? zens in the analysis. Catholicismare 0)non-Catholic 1) Catholic. National PoliticalOrientations Origin To account for national variables are con- Assimilation-How importantis itfor Latinosto changeso that origin, dummy structed for Cubans, Central/South Americans, and they blendinto the largersociety as in the of a melting CaribbeanLatinos, with Mexicans as the pot?The values of the assimilationvariable are; serving compari- son All variableswere coded based on the fol- PoliticalInterest-How muchattention would you say you pay population. set of Earlier said wereHis- to politics and government?The values for the political lowing survey questions; you you or what did ancestorscome interest variable are; 0) none 1) not much 2) a fair panic Latino, country yourfamily of do with more?Each amount 3) a lot. from?Which country you identify dummy variableutilizes the same 0) non-Cuban 1) Ideology-Wouldyou sayyour viewsin mostpolitical matters are coding strategy, Cuban, 0) non-Central/South American 1) Central/South liberal,moderate, or conservative?The values of the ideology non-Caribbean Caribbean.The Caribbean variableare; 0) Liberal,1) Moderate,2) Conservative. American, 0) 1) variableincludes Latinosof both Puerto Ricanand Domini- Partisanship-In politics today, do you consideryourself a can descent. The decision to combine these two Republican,a Democrat,an Independent,or somethingelse? populations is based on the of those two countries, as well as The values of the partisanshipvariable are; 0) Republi- proximity the concentrationof those two communities in the can, 1) Independent, 2) Democrat. regional United States. PoliticalActivities and Experiences REFERENCES Latino specific participation-Please tell me whetheror not you have done each of thefollowing activities in the past ten Arce,Carlos H. 1979.Mexican Origin People in theUntied States: The years; workedas a volunteeror for payfor a Latinopolitical ChicanoSurvey Ann Arbor, MI: University of Michigan,Survey ed. Ann Con- candidate, attended a or demonstration ResearchCenter. ICPSR Arbor: Inter-University public meeting sortium for Political and Social Research. Latinoconcerns, contributed money to a regarding Latino Binder, Norman E., J. L. Polinard, and Robert D. Wrinkle. 1997. candidateor Latino The values and politicalorganization. "MexicanAmerican and Anglo Attitudes TowardImmigration distributionsof the Latino specific participationvariable Reform:A View From the Border."Social Science Quarterly 78: are;0) participationin none of these activities (1583), 1) 324-37. participation in one activity (536), 2) participation in Bolks, Sean M., Diana Evens, J. L. Polinard, and Robert D. Wrin- two activities (212), and 3) participation in all three kle.2000. "CoreBeliefs and Abortion Attitudes;A Look at Lati- activities (86). nos."Social Science Quarterly 81 (1). 446 POLITICALRESEARCH QUARTERLY

Cain, Bruce E., and D. Roderick Kiewiet. 1987. "Latinosand the Michelson, MelissaR. 2001. "TheEffect of NationalMood on Mex- 1984 Election: A ComparativePerspective." In Rodolfo de la ican AmericanPolitical Opinion."Hispanic Journal of Behavioral Garza, IgnoredVoices: Public Opinion Polls and the LatinoCom- Sciences23 (1): 57-70. munity, pp. 47-62. Austin, TX: Center for Mexican American Miller,A. H., P Gurin, G. Gurin, and O. Malanchuk.1981. "Group Studies. Consciousness and Political Participation."American Journal of Dawson, MichaelC. 1994. Behindthe Mule: Race and Classin African- PoliticalScience 25: 494-511. AmericanPolitics. Princeton, NJ: Princeton University Press. Newton, Lina. 2000. "Why Some Latinos Supported Proposition de la Garza,Rodolfo. 1987. IgnoredVoices: Public Opinion Polls and 187: Testing Economic Threat and CulturalIdentity Hypothe- the LatinoCommunity. Austin, TX:Center for MexicanAmerican ses." SocialScience Quarterly 81 (1): Studies. Olsen, M. E. 1970. Social and Political Participationof Blacks. . 1998. "Interests Not Passions: Mexican American Attitudes AmericanSociological Review 35 (4): 682-97. toward Mexico, Immigration From Mexico, and Other Issues Padilla, Felix. 1985. LatinoEthnic Consciousness: The Case of Mexi- Shaping U.S.-Mexico Relations."International Migration Review can Americansand PuertoRicans in Chicago.Notre Dame, IN: 32 (2): 401-22. Universityof Notre Dame Press. de la Garza,Rodolfo, Angelo Falcon, E Chris Garcia,and John A. Polinard,Jerry, Robert D. Wrinkle, and Rodolfo de la Garza. 1984. Garcia. 1990. "LatinoNational Political Survey, 1989-1990." "Attitudes of Mexican Americans Toward Irregular Mexican Philadelphia,PA: Temple University Press. Immigration."International Migration Review18: 782-99. de la Garza,Rodolfo O., Jerry L. Polinard,Robert D. Wrinkle, and Rodriguez, Nestor, and Rogelio T. Nunez. 1986. "An Exploration TomasLongoria. 1991. "UnderstandingIntra-Ethnic Variations: of FactorsThat Contributeto DifferentiationBetween Chicanos MexicanOrigin PopulationViews of Immigration.Social Science and Indocumentados."In Harley L. Browningand Rodolfo O. Quarterly72: 379-87. de la Garza, eds., MexicanImmigrants and MexicanAmericans: Garcia,John A. 2000. "TheLatino and AfricanAmerican Commu- An EvolvingRelation. Austin: University of Texas Press. nities: Bases for Coalition Formation and Political Action." In Sanchez, Gabriel.R. 2006. "TheRole of Group Cohesion in Polit- Gerald Jaynes, ed., Immigrationand Race: New Challengesfor ical ParticipationAmong Latinos in The United States."Ameri- AmericanDemocracy. New Haven, CT:Yale University Press. can PoliticsResearch. . 2003. Latino Politics in America: Community, Culture, and Stokes, Atiya Kai. 2003. "LatinoGroup Consciousness and Politi- Interests.Lanham, MD: Rowman and Littlefield. cal Participation."American Politics Research 31 (4): Gurin,P, Miller,A. H. Gurin,and G. Gurin. 1980. StratumIdentifi- Schmidt, Ron. 1997. "LatinoPolitics in the 1990s: A View from cation and Consciousness. Quarterly 43: 30-47. California."In E Chris Garcia,ed., PursuingPower: Latinos in the Gutierrez,David. 1995. Wallsand Mirrors.Berkeley: University of PoliticalSystem, pp. 457-62. Notre Dame, IN: University of CaliforniaPress. Notre Dame Press, 1997). Hajnal, Zoltan, and Mark Baldassare. 2001. Finding Common Tate, Katherin. 1993. FromProtest to Politics,The New BlackVoters Ground:Racial and EthnicAttitudes in California.San Francisco: in AmericanElections. New York:Sage. Public Policy Institute of California. Uhlaner, Carole. 1989. "RationalGoods and Participation:Incor- Hood, M. V, Irwin L. Morris,and KurtA. Shirkey 1997."!Quedate porating Sociability into a Theory of RationalAction." Public o Vente!":Uncovering the Determinants of Hispanic Public Choice. Opinion toward Immigration. PoliticalResearch Quarterly 50 Uhlaner, CaroleJean. 1991. "PerceivedDiscrimination and Preju- (3): 627-47. dice and the Coalition Prospects of Blacks, Latinos, and Asian Houvouras, Shannon K. 2001. "Effectsof DemographicVariables, Americans."In BryanO. Jackson and Michael B. Preston, eds. Ethnic Prejudice,and Attitudes TowardImmigration on Oppo- Racialand EthnicPolitics in California.Berkeley, CA:. Institutefor sition to BilingualEducation." HispanicJournal of BehavioralSci- GovernmentalStudies Press. ences 23: 136-52. Uhlaner, Carole Jean, and E Chris Garcia. 2002. "LatinoPublic Kaufmann, Karen. 2003. "Cracksin the Rainbow: Group Com- Opinion."In BarberaNorrander and Clyde Wilcox, eds.,Under- monality as a Basis for Latino and African-AmericanPolitical standingPublic Opinion. Washington, DC: CQ Press. Coalitions."Political Research Quarterly 56 (2): 199-210. Verba, Sidney, and Norman Nie. 1972. Participationin America. Leal, David. 2004. Latino Public Opinion? Paper presented at the New York:Harper and Row. "LatinoPolitics: The State of the Discipline,"conference spon- sored by TexasA&M University and the University of Texas at Austin. College Station, TX, April 30-May 1. Martinez, Andres. 2000. "Two Candidates Seeking Votes with Salsa."New YorkTimes, Section A, p. 30, August 16. Martinez,Anne. 2000. "EstablishedLatinos More Likely to Sup- Received: September 20, 2005 port Curbing Immigration."San Jose MercuryNews, p. 19A, Accepted for Publication: January 17, 2006 October 15. [email protected]