Racial Distancing in a Southern CityLatino Immigrantsrsquo Views of Black Americans

Paula D McClain Duke UniversityNiambi M Carter Duke UniversityVictoria M DeFrancesco Soto Duke UniversityMonique L Lyle Duke UniversityJeffrey D Grynaviski University of ChicagoShayla C Nunnally University of ConnecticutThomas J Scotto West Virginia UniversityJ Alan Kendrick St Augustinersquos CollegeGerald F Lackey University of North CarolinaKendra Davenport Cotton University of North Carolina

The United States is undergoing dramatic demographic change primarily from immigration and many of the newLatino immigrants are settling in the South This paper examines hypotheses related to attitudes of Latino immi-grants toward black Americans in a Southern city The analyses are based on a survey of black white and Latinoresidents (n = 500) The results show for the most part Latino immigrants hold negative stereotypical views ofblacks and feel that they have more in common with whites than with blacks Yet whites do not reciprocate in theirfeelings toward Latinos Latinosrsquo negative attitudes toward blacks however are modulated by a sense of linked fatewith other Latinos This research is important because the South still contains the largest population of AfricanAmericans in the United States and no section of the country has been more rigidly defined along a black-whiteracial divide How these new Latino immigrants situate themselves vis-agrave-vis black Americans has profound impli-cations for the social and political fabric of the South

Latino populations from 1990 to 2000 (US CensusBureau 2002a 2002b) Notwithstanding these majorchanges far less attention has been paid to the chang-ing demographics in the South as this region has notbeen the recipient of large numbers of Latino immi-grants in the past

Latinos by and large are an entirely new popula-tion in the South Unlike other regions of the countrythe South for the most part has had no experiencewith immigrant populations of Latin Americanorigin The large-scale settlement of Latinos in theDeep South is no more than 10ndash15 years old (DurandMassey and Charvet 2000) Most of these Southernareas are what Suro and Singer (2002) refer to as ldquonewLatino destinationsrdquo Areas with increasing Latinopopulations include cities such as Atlanta GA

The Journal of Politics Vol 68 No 3 August 2006 pp 571ndash584

copy 2006 Southern Political Science Association ISSN 0022-3816

The 2000 Census confirmed what many Ameri-cans already suspectedmdashdramatic demo-graphic change was under way in the United

States The country was becoming increasingly moreracially and ethnically diverse with Latinos being oneof the fastest growing racial and ethnic groups TheUS Census Bureau estimates that by 2050 non-His-panic whites will be only 501 of the populationwhile Latinos will be close to one-quarter of the population (US Census Bureau 2003) Some of thedemographic changes are attributed to increasingLatino immigration into the United States with areasof the South experiencing some of the most dramaticdemographic change1 A number of Southern statessuch as North Carolina Alabama and Georgiareported substantial increases in the size of their

1The USCensus Bureau estimates that 53 of the growth in the Latino population is the result of immigration

Charlotte Greensboro-Winston Salem and Raleigh-Durham NC Nashville and Memphis TN andGreenville SC among others Between 1990 and 2000the Atlanta MSA experienced a 388 increase inLatino population (55045 to 268851 comprising 7of the MSA population) Charlotte Greensboro-Winston Salem and Raleigh-Durham MSAs saw theLatino population increase 685 (9817 to 77092 for5 of MSA) 809 (6844 to 62210 for 5 of MSA)and 631 (9923 to 72580 for 6 of MSA) respec-tively Nashville and Memphis MSAs have seen a 454(7250 to 40139 for 3 of MSA) and 265 (7546 to27520 for 2 of MSA) increase respectively in theirLatino populations and the Greenville MSA under-went a 358 (5712 to 26167 for 3 of MSA) increaseover the decade (Suro and Singer 2002 12ndash13)

What accounts for this increased Latino immigra-tion into the South The limited research suggestsseveral reasons Major shifts in the global economyand new trade policies such as the 1994 North Amer-ican Free Trade Agreement (NAFTA) encouraged themigration of labor and capital restructuring regionalnational and global economies Older industries inthe South such as agriculture steel textiles furnitureand clothing were damaged by the new economywhich included foreign-owned auto plants high-techresearch and manufacturing biomedical research andnew food processing plants for poultry hogs andseafood (Griffith 1993) From the late 1970s to theearly 1990s the economy of the South outperformedall other regions of the country resulting in a need forlarge numbers of unskilled and inexpensive labor(Duchon and Murphy 2001 1 Kandel and Parrado2004) In order to fill this expanded labor need ofunskilled low-wage workers these industries began toactively recruit immigrant workers from Mexico andCentral America but primarily from Mexico (Torres2000)

Coupled with the pull of economic changes in theSouth was the push of the ongoing economic crisis inMexico This continuing crisis pushed many legal aswell as undocumented immigrants into the Southernpart of the United States (Massey Durand and Malone2002) Many of these Latino immigrants have settled insouthern towns and rural areas but large numbershave chosen urban areas as well Those who settle in urban areas tend to be employed in nonunionlow-wage jobs such as service work cleaning hotelsworking for other types of cleaning companies con-struction and landscaping firms and building mainte-nance (Massey Durand and Malone 2002)

The new Latino immigrants are entering a regionwhere race has defined the context structure and life

chances of black and white Southerners for centuriesAs VO Key highlighted in Southern Politics (1949) thepolitics of the South is the politics of race While muchhas changed since Keyrsquos book was published morethan a half-century ago few would argue that race hasceased to be a salient factor in the political and socialfabric of the region What effect is this demographicshift going to have on the structure of intergroup rela-tions in the South that has historically been entirelybased on the relationship between blacks and whitesRelations between the new Latino immigrants andestablished Southern black communities have thepotential to be fraught with conflict But it is also pos-sible these two groups will see each other as havingmuch in common as racial minorities in a Southernenvironment How these new immigrants relate to theestablished black and white populations is of criticalimportance in analyzing how they will adapt to theregion and how these established populations willadapt to them Drawing on data from a study of aSouthern city that is undergoing tremendous demo-graphic changes the findings in this paper suggest thatrelations between black Americans and Latino immi-grants are likely to be one of conflict rather than ajoining together based on shared minority status inthe South

Two Historical Cases of RacialDistancing

Given the recent nature of Latino immigration intothe South it is not surprising that little scholarlyattention has been paid to the topic although someresearch is beginning to emerge (Ciscel Smith andMendoza 2003 Cobb and Stueck 2005 Hernaacutendez-Leoacuten and Zuntildeiga 2000 2001 2003 2005 Kandel andParrado 2004 Mohl 2003 Murphy Blanchard andHill 2001 Peacock Watson and Matthews 2005Torres 2000) The exception in the literature has beenresearch on Miami and the effect of the influx of post-1959 Cubans Before Miami however there was aprior historical instance of the introduction of a newforeign population into the black-white dynamic ofthe Southmdashthe Chinese in Mississippi Their intro-duction was probably one of the first moments inpost-Civil War United States history where the black-white biracial paradigm was confronted in any realway The reaction of both of these new immigrantgroups to black Americans might give us a sense of how the new Latino immigrants will react to andposition themselves in relation to the Southern black

rsquo

American population Both Chinese immigrants tothe South during the late-nineteenth century andCuban immigrants during the mid-twentieth centurypursued a strategy of racial distancing seeing them-selves as being in economic and social competitionwith black Americans rather than as natural allies inthe fight for social and political equality

Chinese in Mississippi

The Chinese entered Southern history during Recon-struction as part of a strategy devised by Southernplanters to retain political power and economiccontrol over the newly freed blacks (Loewen 1971Quan 1982) The plantersrsquo strategy was the importa-tion and employment of Chinese men in occupationspreviously held by blacks (Quan 1982 5) From theplantersrsquo perspective the Chinese were ideal On theone hand they would take jobs from blacks forcingthem into a subordinate position while on the otherthey were barred from citizenship could not vote andthus would not present a potential political threat(Loewen 1971 22) This attempt however was anutter failure and by 1880 there were only 51 Chineseinhabitants of the Mississippi Delta (Loewen 197125)2 These numbers increased however as Chinesemen brought over Chinese brides

The Chinese were immediately confronted withthe Southrsquos racial hierarchy where they were classifiedwith blacks and subject to some of the same segrega-tion laws (Loewen 1971 23ndash24) Despite the cordialrelations Chinese apparently shared initially withblacks they forcefully resisted their social and legaldesignation with blacks In fact Mississippi Chinesefiled one of the earliest legal challenges to school seg-regation in the United States in 19243 Historians agreethat the Chinese actively sought white approval andsought to distance themselves from blacks (Loewen1971 Quan 1982) Moreover to insure their middleposition between whites and blacks the Chinese toldwhites that they were not interested in marryingwhites and were definitely not interested in marryingblacks (Loewen 1971 79) They were successful in

convincing whites that they were more like whitesthan they were like blacks (whitening up) and theyshared whitesrsquo negative attitudes toward and approvedof their treatment of blacks

During the Civil Rights Movement (CRM) rela-tions between blacks and Chinese were strained evenfurther Blacks knew the Chinese community favoredwhites and the CRM served to make that choice moreobvious In 1970 when school desegregation came toMississippi both whites and Chinese enrolled theirchildren in private academies (Loewen 1971 177) TheDelta Chinese saw the CRM as threatening the systemof racial segregation that enabled them to occupy amiddling space in the oppressive Mississippi environ-ment (Loewen 1971 Quan 1982) In brief in order toattain a certain degree of social standing the Missis-sippi Chinese had to actively distance themselves fromblacks to show whites that they had no sympathy withblacks and their plight As such the MississippiChinese began to see themselves as having more incommon with whites than with blacks and acted onthat perceived commonality

Cubans in Miami

Miami is probably the best-known instance of theintroduction of substantial numbers of a foreign pop-ulation into the South but it did not occur en masseuntil 1959 The number of Cubans in the Miami areain the 1950s was small The estimate is only about 4 (19800) of Miamirsquos 495000 population wereSpanish-language speakers in 1950 and about 70 ofthat number (13860) were estimated to be Cuban(Clark 1991) The push for the immigration ofmassive numbers of Cubans was the Cuban Revolu-tion in 1959 while the pull to the United States wasthe needs of United States Cold War foreign policyThe Cubans began to arrive in Miami in 1959 just asthe black Civil Rights Movement was getting underway and was creating opportunities for blacks acrossthe nation The arrival to a large extent of over800000 Cubans in Miami between 1959 and 1980short-circuited economic and political gains for blackMiamians (Mohl 1990)

Cubans were not the normal labor immigrantcase Most of the 1959 immigrants were from the pro-fessional class in Cuba While they initially worked inlow-paying jobs that blacks had occupied the UnitedStates government moved quickly to assist them aspart of its strategy to undermine Fidel Castro in CubaPrograms were developed that provided Cubans withfood and clothing housing assistance social servicesmedical care relocation assistance educational pro-

2According to the 2000 Census there are 3099 Chinese in Missis-sippi presently and a total Asian population including EastIndians of 18626 The most plentiful group is Vietnamese

3The Mississippi Constitution of 1890 as cited by Loewen (197167) and Quan (1982 45) stated that ldquoseparate schools shall bemaintained for children of the white and colored racesrdquo andbecause the Chinese were not of the White race they were consid-ered of the colored races In the case brought by Gong Lum agrocer in Rosedale and a US citizen (Gong Lum v Rice 1927 275US 390) the US Supreme Court upheld the Mississippi SupremeCourtrsquos decision upholding segregation of Chinese with blacks

grams job training and job placement As a resultthousands of Cuban elites were able to resume theirprofessional lives in South Florida It has been esti-mated that between 1960 and 1990 approximately $2billion was spent on Cuban resettlement activities(Mohl 1990)

The effect of massive federal financial assistance tothe new Cuban immigrants was not lost on Miamirsquosblack population who were not able to participate inthe programs established for the Cuban immigrantsAsearly as 1960 the Miami black press began to view theCuban influx as detrimental to black aspirations andachievements Blacks and Cubans began to compete for housing and residential space jobs and govern-ment services (Mohl 1990) The significant amount offederal resources going to Cubans significantly shiftedthe economic landscape in the 20 years of Cubanimmigration to Dade County black entrepreneurshipsteadily declined From 1968 to 1980 Cubans received466 of all Small Business Administration (SBA)loans in Dade County compared to only 6 to blacksAfter the four riots in the 1980s the situation worsenedwith 90 of SBA loans in Miami-Dade going toCubans and whites (Grenier and Castro 2001)

Like the Mississippi Chinese Cubans in Miamialso perceived a benefit to distancing themselves fromblack Americans The substantial assistance from andspecial status given the Cubans by the United Statesgovernment put them in a privileged position vis-agrave-vis black Americans Cubans most of whom in thefirst post-1959 wave self-identified as white perceivedthat they had little in common with blacks and situ-ated themselves alongside whites Moreover the racialstratification and racial history in Cuba was similar tothat of the United States Blacks arrived in Cuba asslaves and the post-slavery color hierarchy was one inwhich most of the social economic and political ben-efits went to those of lighter complexion while thoseCubans of African ancestry and darker skin were rel-egated to the bottom rungs of society Segregationalong with racial discrimination was prominent inpre-1959 Cuba (de la Fuente 1995 1998 Guimaraes2001 Scott 1985) Race was an old problem in Cubawith its origins in slavery and colonization and thepost-1959 white Cubans brought those attitudes withthem to the United States

Theoretical Implications

In both the Mississippi and Miami cases a system ofracial hierarchy and segregation allowed both com-munities to flourish In the Mississippi case theChinese flourished despite their status as a nonfavored

group whereas with the Cubans in Miami theirsuccess was predetermined given the conditions underwhich they entered the United States and the massivefederal government assistance If the patternsobserved in these two historical cases are predictiveit suggests that the new Latino immigrants into theSouth will distance themselves from Southern blacksRecent literature underscores the presence of this his-torical pattern In a study of black and Latino relationsin Houston TX in the 1990s Mindiola Niemann andRodriguez (2003) found that US-born Latinosexpressed more negative views of black Americansthan blacks expressed of Latinos but foreign-bornLatinos held even more negative views of black Amer-icans Moreover foreign-born Latinos were the leasttolerant of blacks

It is possible that attitudes about blacks present intheir home countries make Latino immigrants morelikely to see black Americans negatively and want todistance themselves from blacks A large literatureexists on racial hierarchy racial prejudice and dis-crimination and racial stereotypes in Latin Americathe Caribbean and Mexico (de la Cadena 2001Dulitzky 2005 Geipel 1997 Guimareas 2001Hanchard 1994 Moumlrner 1967 Sweet 1997 Wade1993 1997 Winant 1992) Literature specifically onthe racial ordering in Mexico argues that race becamea principal factor in the social and economic organi-zation of Spanish colonial society and the racial order-ing was established to benefit whites (Menchaca 2001Seed 1982) Despite the negative attitudes towardblacks on the part of Latino immigrants identified byMindiola Niemann and Rodriguez (2003) it is pos-sible that increased contact between the two groupswill reduce Latino immigrantsrsquo negative attitudestoward black Americans One approach to thinkingabout this response is through the lens of contacttheory

Contact Theory

Contact theory (or propinquity) is the springboardfor much of the literature on racial attitudes It arguesthat increased contact between two groups with negative attitudes toward each other will result in a decrease in negative attitudes (Hood and Morris1998) Originally associated with the work of Allport(1954) research on contextual determinants of racialand ethnic relations has been mixed Some scholarshave found increased contact under certain condi-tionsmdasheg interdependence common goals equalstatus and encouragement by authoritiesmdashreducesprejudice (Amir 1969 1976 Ellison and Powers 1994

rsquo

Jackman and Crane 1986 Kinder and Mendelberg1995 Powers and Ellison 1994 Sigelman and Welch1993 Stephan and Stephan 1985 Welch et al 2001)Other researchers have found that as the concentra-tion and consequently contact of two groupsincreases competition and prejudice also increase(Fossett and Kiecolt 1989 Quillian 1996) Still othershave identified the presence of both increased ani-mosity and reduced prejudice (Morris 1999 Sigelmanand Welch 1993)

Most of this research however has focused onblack and white interaction and few studies havefocused on Latino and Asian interaction with whites(Taylor 1998) Research on the effects of interminor-ity group contact on prejudice and stereotypes islimited Oliver and Wong (2003) examined attitudesof racial groups in multiracial settings finding withthe exception of Asians people who live in neighbor-hoods where their group dominates tend to harborgreater negative stereotypes about other racial minor-ity groups In other words racial stereotypes increaseas the percent of onersquos own ingroup increases in theirneighborhood Blacks and Latinos who are the mostracially isolated harbor the most negative views ofother groups but this pattern was not as pronouncedamong blacks and Latinos residing in neighborhoodsthat are racially diverse Yet they also found thathigher levels of propinquity correlated with lowerlevels of prejudice and perceived competition (Oliverand Wong 2003) Thus it appears that contact theorycan result in three different types of reactions reducedprejudice and perceived competition increased preju-dice and perceived competition or a mixture of both

Drawing from the cases of Miami and Mississippithe literature on Latino immigrant attitudes towardblack Americans and contact theory we test twohypotheses

H1 Latino immigrants hold negative stereotypes of black Americans but these stereotypes will be reduced by length of stay in the countryand increased interaction with black Americansand

H2 Latino immigrants see themselves as closer towhites than to blacks but increases in length oftime in the country and increased interactionwith black Americas will push them closer toblacks

Research Setting

Durham North Carolina is the setting for this studyThe City of Durham like many Southern locations is

undergoing demographic change Table 1 shows thechanges in the demographics of the population of theCity of Durham from 1990 to 2000 What is clear isthat the number of Latinos residing in the city hasrisen dramatically In 1990 Latinos were slightly morethan 1 of the population but by 2000 their per-centage reached 86 For decades whites were themajority in Durham (516 in 1990) but the increas-ing Latino population along with a smaller increasein the Asian population have reduced the white pro-portion to the point where in 2000 blacks and whiteswere almost equal percentages of the population455 for whites and 438 for blacks4

Durhamrsquos Latino population is antithetical to theCuban case First the current wave of Latino immi-grants does not fit into the United Statesrsquo foreignpolicy calculations of encouraging an exodus from a communist country in an effort to undermine thegovernment Thus the enormous amounts of federalaid that went to post-1959 Cubans in South Floridado not exist for these new Latino immigrantsAlthough Latinos in Durham are not benefiting fromtargeted social services as in Miami it does appearthere are special programs designed to help native

T 1 1990 and 2000 Census Data for the Cityof Durham

Change1990 2000 (+-)

Total Population 136611 187035 +

Whites 516 455 minus(70513) (85126)

Blacks 457 438 minus(62393) (81937)

American Indians 2 3 +(358) (575)

Asians 19 36 +(2676) (6815)

Latinos 13 86 +(1713) (16012)

Mexicans 308 642 +(528) (10343)

Puerto Ricans 215 67 minus(369) (696)

Cubans 88 147 +(150) (236)

Other Hispanics 389 296 minus(666) (4737)

4Both of these groups gained in absolute numbers of people butlost as a proportion of the population from 1990 to 2000

Spanish speakers receive more traditional forms ofaid Second Latinos in Durham are typically notmembers of the professional community in theirhome countries and have lower incomes and educa-tion and skill levels than the Cuban community Manyof these immigrants are from poorer countries egMexico and the countries of Central America5 There-fore Latinos and blacks might come into greater com-petition for the same jobs and social services than inMiami (for a discussion of possible job competitionsee McClain et al 2005) Third the reasons for immi-gration into the United States for recent immigrantsare primarily economic rather than political6

Why Durham North Carolina First North Car-olina has the fastest growing Latino population in thecountry It experienced an almost 500 increase in itsLatino population primarily with immigrants fromMexico skyrocketing from 76726 in 1990 to 378963in 2000 (US Census Bureau 2002a 2002b) Further-more North Carolina had the highest rate of growthin its immigrant population out of all the states in the 1990s Suro and Singer (2002) identify Raleigh-Durham as having the highest rate of Latino growthfrom 1980 to 2000mdash1180 [See figures in Table 1]Second Durham like other cities of the New Southhas experienced a decline or exit of industries whereblacks have traditionally found work As a result asubstantial portion of the black population now worksin the service industry and many (but not necessarilyall) blacks and the new Latino immigrants find them-selves competing for the same jobs Finally from aresearch and data gathering perspective Durham is ofa manageable size

Data and Measures

The analyses in this paper are based on the 2003Durham Survey of Intergroup Relations (DSIR n =500) The survey was conducted specifically for ourproject by the Center for Survey Research of the Uni-

versity of Virginia using a Computer-Aided TelephoneInterviewing (CATI) system employing random digitdialing (RDD) and dialing of directory-assisted His-panic surname sample7 A randomly generated sampleof phone numbers based on exchanges valid in theDurham North Carolina area was called An over-sample of numbers listed in the phone directory undera Hispanic surname was called at the same time8 Thesurvey was conducted from May 4 through June 222003 and interviews were conducted in both Englishand Spanish (32 of the interviews were conductedin Spanish which translates into 958 of the Latinosample opting to be interviewed in Spanish)9 Araceethnicity quota was implemented to achieve aminimum of 150 whites 150 blacks and 150 Latinosthe remaining 50 respondents were not under thisquota restriction and represent a number ofracialethnic backgrounds Interviews were completedwith 500 residents of the City of Durham for anoverall response rate of 21610 The sample of 500consists of 160 whites (32) 151 blacks (30) 167Latinos (34) 6 Asians (12) 12 who designatedtheir race as Other (24) and 1 respondent (2)who did not indicate a racial category11

5The News and Observer (Raleigh NC) identified that many of theMexican immigrants into North Carolina come primarily fromrural towns in the State of Puebla (November 29 1998 Novem-ber 30 1998) For the most part these immigrants are unskilledand poorly educated

6In a series of articles throughout 2002 chronicling the lives of arearesidents living in poverty The Herald Sun (Durham) provided apicture of life for Latinos in Durham that differed markedly fromthe Miami case Fully 26 of the more than 16000 Latinos inDurham live below the federal poverty level and in order to makea good living it is necessary for them to work more than one job(Assis and Pecquet 2002 A12)

7We recognize the problems associated with drawing a samplefrom a listing of Hispanic surnames for example missing His-panics with non-Hispanic last names and those married to non-Hispanics We also acknowledge that some Latino immigrantsmight not have phones in their homes Given the recency of theHispanic population in Durham and the high proportion ofimmigrants however we choose the sampling frame that wouldgive us the highest probabilities of getting to a Latino respondent

8Both samples were purchased from Survey Sampling Inc (SSI)of Fairfield CT

9We had the questionnaire translated by a Spanish-languageorganization in Chapel Hill NC In order to check the translationand to ensure that it tracked the English-language version thesurvey organization drew a small sample of Latinos in Durhamfor the sole purpose of checking the translation As a resultchanges were made to the translation The revised Spanish-language version was then pretested on another small sample of Latinos in Durham

10A total of 4208 phone numbers were attempted in the course ofthe survey and a total of 14014 call attempts were made TheAmerican Association for Public Opinion Research (AAPOR) ratewas calculated using the full call history of each number that wasrecorded automatically by the CATI software The response ratewas calculated according to AAPOR suggested formula RR3 withe1 = 50 and e2 = 78 We estimated e1 and e2 based on an analy-sis of residency rates and the occurrence of out-of-area householdsin our sample Partial interviews are not counted in the numera-tor of the RR3 formula

11Due to the use of the Hispanic surname sample and racialethnicquotas sampling error is more difficult to calculate The samplemay be viewed as part of two separate populations Within theRDD sample the source of 276 completions the probability of

rsquo

To assess the quality of the sample we comparedthe distribution of demographic characteristics in ourstudy with the characteristics of the population inDurham as reported in the 2000 Census Since thestudy was designed to oversample the black andLatino populations we did a detailed analysis of theextent to which each of our three main subpopulationsamples are represented in our sample the results ofwhich are not presented here12 In brief overall wefind that our sample is reasonably representative ofthe three racial subpopulations under investigationthough the distribution of respondents is slightlyolder better educated and more often female thanfound in the 2000 Census

Our data also confirm what we had suspectedmdashthe Latino population in Durham is basically animmigrant population primarily from MexicoNinety-three percent of the Latino respondents in thesample were born outside the United States Of the93 (n = 156) only about 19 were naturalized cit-izens While Mexicans were the largest portion of theLatino sample (63) Latinos from Central Americawere the next largest group (23) followed by SouthAmerican (5) Puerto Rican (4) Spanish (2)Cuban (1) and Other Latino (2) Given that weare interested in Latino immigrant attitudes most of the analyses are performed using only these 156individuals

We use two dependent variables in our analyses totest our hypotheses One dependent variablemdashLatinostereotypical views of blacksmdashis an index of threestereotype questions ranging in value from 3 (holdingthe least stereotypical views of blacks) to 15 (holdingthe most stereotypical views of blacks)13 A second

dependent variable measures which racial group(Asians blacks or whites) Latinos felt they have themost in common with (1 = have most in commonwith whites 0 = have most in common with blacks)14

We use the following predictor variables although notall are included in every equation education attain-ment measured as number of years of schooling(ranging from 0 to 25 years)15 perception of numberof blacks (Latinos) in their neighborhood16 socialcontact with blacks (Latinos)17 length of time in theUnited States as measured by a question that askedimmigrant respondents the year that they arrived18

(ranges from 0 to 35 years mean = 875 years median= 6 years)19 and gender (1 = male 0 = female) We donot include contact with blacks at work as one of thecontact indicators because of substantial missing dataEducational attainment and gender are used as demo-graphic controls

We also include the indicator of linked fate (thosebelieving that what happens to other blacks (Latinos)will have something to do with their lives)20 for severalreasons The concept of linked fate has been found tobe a powerful predictor of group consciousness polit-ical participation and political attitudes among blackAmericans (Dawson 1994 Matthews and Prothro1967 Shingles 1981 Tate 1993) The concept of linked

selection is known and the margin of error is plusmn59 Within thesurname oversample providing 244 completed interviews allhouseholds listed under a resident with a Hispanic surname wereattempted however Hispanics were included in RDD calling andnon-Hispanics were included in the oversample Non-Hispanicswith Hispanic surnames had a greater chance of selection thannon-Hispanics in the RDD sample who do not have Hispanic sur-names If we assume this to be a more or less random occurrencethen the margin of error for each of the three racialethnic groupsis roughly 8

12A copy of the full analysis is available from the lead author

13Question wording (Q1) For blacks do you think ldquowork hardrdquodescribes almost all blacks most blacks some blacks few blacksalmost no blacks (Q2) For blacks do think ldquoit is easy to get alongwith themrdquo describes (same response categories as question 1)(Q3) For blacks do you think ldquoyou can trust themrdquo describes(same response categories as question 1) We recognize that thesethree questions are the bluntest measures of stereotypes and othermore subtle measures of stereotypes exist but had to make diffi-cult choices when we had to cut the questionnaire in half in orderto stay within budget As such we decided on measuring the mostnegative stereotypes

14Question wording for Latinos (1) Of the following groups if youhad to say which one do you feel you have the most in commonwith African Americansblacks Asians or whites (2) and whichof the following groups do you feel you share the least in commonwith African Americansblacks Asians or whites We recoded theAsian category as missing since only six Latino respondentsreported feeling that Latinos have the most in common withAsians Coding for the dependent variable is 1 (have most incommon with whites) and 0 (have most in common with blacks)

15Income had initially been included in our model but wasdropped due to suspected collinearity with education

16Question wording In your neighborhood how many of the res-idents are black (Latinos) (1) none of the residents (2) a few ofthe residents (3) some of the residents (4) most of the residents(5) all of the residents

17Question wording In your social life including your friends andpeople you know from church and other social activities do youhave (4) a lot of contact (3) some contact (2) little contact or (1)no contact at all with blacks (Latinos)

18We converted this measure to years in the United States by subtracting the year arrived from 2003 (the year the survey wasconducted)

19Although our sample is of Latino immigrants some of whomare new arrivals the mean (875 years) and median (six years) ofthis predicator suggest substantial variance and is thus an ade-quate measure of length of time in country

20Question wording Do you think what happens to Hispanics inthis country will have something to do with what happens in yourlife

fate as Dawson (1994 76) suggests evolves fromsocial identity theory that suggests individuals formtheir concepts of self in part by identifying similari-ties and differences between themselves and othersIndividuals perceive a similarity in experience andtreatment or perceive a linked fate with others (Davisand Brown 2002 241) Only a few studies have exam-ined concepts similar to linked fate for Latinos Kauf-mann (2003) found that Latinos who feel close to oneanother as a group were much more likely to feel closeto black Americans In view of the significance oflinked fate for black Americans and the limited lit-erature suggesting that feelings of closeness to otherLatinos may influence Latino relations with blackAmericans we want to see if Latino linked fate existswithin Latino immigrant populations and if it doeswhat effect does it have on Latino immigrantsrsquo per-ceptions of black Americans The data indicate two-thirds (6791) of Latino immigrants feel a sense oflinked fate with other Latinos This would suggest atleast as reflected in our survey linked fate does existamong Latino immigrants Latino linked fate is meas-ured as 1 = yes and 0 = no

Although we have not specified hypotheses forblacks we examine similar equations for blacks inorder to get a sense of the reciprocal nature of the rela-tionship (n = 151) Not all indicators used for Latinoshowever are appropriate to use for blacks eg lengthof time in country as approximately 95 of blacks arenative-born We use OLS and logit regression analy-ses to test our hypotheses We use Clarify which makesit possible to compute predicted probabilities forvalues of the dependent variable while holding thepredictor variables at their means or some other valueon several of the logit analyses (King Tomz and Wittenberg 2000 Tomz Wittenberg and King 2003)

Testing and Results

Analyses for our first hypothesis indicates that theprevalence of negative stereotypes of black Americansin the Latino immigrant community is quite wide-spread and seem especially so when compared to theprevalence of white stereotypes of blacks Along eachdimension the stereotypes of blacks by Latinos aremore negative than those of white respondents (SeeTable A1 in the JOP online appendix at httpwwwjournalofpoliticsorg) We found that 589 ofthe Latino immigrants in our study reported feelingthat few or almost no blacks are hard workingapproximately one-third (325) of the Latino immi-grant respondents reported feeling that few or almost

no blacks are easy to get along with and slightly morethan a majority (569) of the Latino immigrantrespondents reported feeling that few or almost noblacks could be trusted Among whites the compara-ble figures are only 93 indicate that few or almostno blacks are hard working only 84 believe that fewor almost no blacks are easy to get along with andonly 96 feel that few or almost no blacks could betrusted Consistent with our theoretical expectationsit does appear that many Latinos hold very negativestereotypes of blacks

What factors influence the stereotypes thatLatinos hold of black Americans Table 2 presents theresults of our OLS analysis Even given the relativelysmall number of observations almost all of our pre-dictors are significant and in the hypothesized direc-tion To begin the two demographic control variableseducation and male are both statistically significantMore educated Latinos have significantly less negativestereotypes while men appear to have significantlymore negative stereotypes (perhaps reflecting a pre-disposition of men to use more extreme values onsurvey instruments) More interesting of course arethe influence of the remaining variables in the model

T 2 OLS Regression Results Predicting theExtent to Which Latino ImmigrantsHold Negative Stereotypes of Blacks

Variable Model

Main EffectsEducation minus1033

(0421)Male 7860

(3789)Linked Fate minus7472

(4225)Amount of Social Contact with Blacks minus4780

(1959)Number of Blacks in Neighborhood 2294

(1926)Length of Time in the United States minus0295

(0269)

Constant 12216N 130F Statistic 446R-Squared 1788

Note The dependent variable stereotypical views of blacksranges in value from 3 (holding the least stereotypical views ofblacks) to 15 (holding the most stereotypical views of blacks)Standard error estimates are in parenthesesp lt 01 (two-tailed test)p lt 05 (two-tailed test)p lt 10 (two-tailed test)

Charlotte Greensboro-Winston Salem and Raleigh-Durham NC Nashville and Memphis TN andGreenville SC among others Between 1990 and 2000the Atlanta MSA experienced a 388 increase inLatino population (55045 to 268851 comprising 7of the MSA population) Charlotte Greensboro-Winston Salem and Raleigh-Durham MSAs saw theLatino population increase 685 (9817 to 77092 for5 of MSA) 809 (6844 to 62210 for 5 of MSA)and 631 (9923 to 72580 for 6 of MSA) respec-tively Nashville and Memphis MSAs have seen a 454(7250 to 40139 for 3 of MSA) and 265 (7546 to27520 for 2 of MSA) increase respectively in theirLatino populations and the Greenville MSA under-went a 358 (5712 to 26167 for 3 of MSA) increaseover the decade (Suro and Singer 2002 12ndash13)

What accounts for this increased Latino immigra-tion into the South The limited research suggestsseveral reasons Major shifts in the global economyand new trade policies such as the 1994 North Amer-ican Free Trade Agreement (NAFTA) encouraged themigration of labor and capital restructuring regionalnational and global economies Older industries inthe South such as agriculture steel textiles furnitureand clothing were damaged by the new economywhich included foreign-owned auto plants high-techresearch and manufacturing biomedical research andnew food processing plants for poultry hogs andseafood (Griffith 1993) From the late 1970s to theearly 1990s the economy of the South outperformedall other regions of the country resulting in a need forlarge numbers of unskilled and inexpensive labor(Duchon and Murphy 2001 1 Kandel and Parrado2004) In order to fill this expanded labor need ofunskilled low-wage workers these industries began toactively recruit immigrant workers from Mexico andCentral America but primarily from Mexico (Torres2000)

Coupled with the pull of economic changes in theSouth was the push of the ongoing economic crisis inMexico This continuing crisis pushed many legal aswell as undocumented immigrants into the Southernpart of the United States (Massey Durand and Malone2002) Many of these Latino immigrants have settled insouthern towns and rural areas but large numbershave chosen urban areas as well Those who settle in urban areas tend to be employed in nonunionlow-wage jobs such as service work cleaning hotelsworking for other types of cleaning companies con-struction and landscaping firms and building mainte-nance (Massey Durand and Malone 2002)

The new Latino immigrants are entering a regionwhere race has defined the context structure and life

chances of black and white Southerners for centuriesAs VO Key highlighted in Southern Politics (1949) thepolitics of the South is the politics of race While muchhas changed since Keyrsquos book was published morethan a half-century ago few would argue that race hasceased to be a salient factor in the political and socialfabric of the region What effect is this demographicshift going to have on the structure of intergroup rela-tions in the South that has historically been entirelybased on the relationship between blacks and whitesRelations between the new Latino immigrants andestablished Southern black communities have thepotential to be fraught with conflict But it is also pos-sible these two groups will see each other as havingmuch in common as racial minorities in a Southernenvironment How these new immigrants relate to theestablished black and white populations is of criticalimportance in analyzing how they will adapt to theregion and how these established populations willadapt to them Drawing on data from a study of aSouthern city that is undergoing tremendous demo-graphic changes the findings in this paper suggest thatrelations between black Americans and Latino immi-grants are likely to be one of conflict rather than ajoining together based on shared minority status inthe South

Two Historical Cases of RacialDistancing

Given the recent nature of Latino immigration intothe South it is not surprising that little scholarlyattention has been paid to the topic although someresearch is beginning to emerge (Ciscel Smith andMendoza 2003 Cobb and Stueck 2005 Hernaacutendez-Leoacuten and Zuntildeiga 2000 2001 2003 2005 Kandel andParrado 2004 Mohl 2003 Murphy Blanchard andHill 2001 Peacock Watson and Matthews 2005Torres 2000) The exception in the literature has beenresearch on Miami and the effect of the influx of post-1959 Cubans Before Miami however there was aprior historical instance of the introduction of a newforeign population into the black-white dynamic ofthe Southmdashthe Chinese in Mississippi Their intro-duction was probably one of the first moments inpost-Civil War United States history where the black-white biracial paradigm was confronted in any realway The reaction of both of these new immigrantgroups to black Americans might give us a sense of how the new Latino immigrants will react to andposition themselves in relation to the Southern black

rsquo

American population Both Chinese immigrants tothe South during the late-nineteenth century andCuban immigrants during the mid-twentieth centurypursued a strategy of racial distancing seeing them-selves as being in economic and social competitionwith black Americans rather than as natural allies inthe fight for social and political equality

Chinese in Mississippi

The Chinese entered Southern history during Recon-struction as part of a strategy devised by Southernplanters to retain political power and economiccontrol over the newly freed blacks (Loewen 1971Quan 1982) The plantersrsquo strategy was the importa-tion and employment of Chinese men in occupationspreviously held by blacks (Quan 1982 5) From theplantersrsquo perspective the Chinese were ideal On theone hand they would take jobs from blacks forcingthem into a subordinate position while on the otherthey were barred from citizenship could not vote andthus would not present a potential political threat(Loewen 1971 22) This attempt however was anutter failure and by 1880 there were only 51 Chineseinhabitants of the Mississippi Delta (Loewen 197125)2 These numbers increased however as Chinesemen brought over Chinese brides

The Chinese were immediately confronted withthe Southrsquos racial hierarchy where they were classifiedwith blacks and subject to some of the same segrega-tion laws (Loewen 1971 23ndash24) Despite the cordialrelations Chinese apparently shared initially withblacks they forcefully resisted their social and legaldesignation with blacks In fact Mississippi Chinesefiled one of the earliest legal challenges to school seg-regation in the United States in 19243 Historians agreethat the Chinese actively sought white approval andsought to distance themselves from blacks (Loewen1971 Quan 1982) Moreover to insure their middleposition between whites and blacks the Chinese toldwhites that they were not interested in marryingwhites and were definitely not interested in marryingblacks (Loewen 1971 79) They were successful in

convincing whites that they were more like whitesthan they were like blacks (whitening up) and theyshared whitesrsquo negative attitudes toward and approvedof their treatment of blacks

During the Civil Rights Movement (CRM) rela-tions between blacks and Chinese were strained evenfurther Blacks knew the Chinese community favoredwhites and the CRM served to make that choice moreobvious In 1970 when school desegregation came toMississippi both whites and Chinese enrolled theirchildren in private academies (Loewen 1971 177) TheDelta Chinese saw the CRM as threatening the systemof racial segregation that enabled them to occupy amiddling space in the oppressive Mississippi environ-ment (Loewen 1971 Quan 1982) In brief in order toattain a certain degree of social standing the Missis-sippi Chinese had to actively distance themselves fromblacks to show whites that they had no sympathy withblacks and their plight As such the MississippiChinese began to see themselves as having more incommon with whites than with blacks and acted onthat perceived commonality

Cubans in Miami

Miami is probably the best-known instance of theintroduction of substantial numbers of a foreign pop-ulation into the South but it did not occur en masseuntil 1959 The number of Cubans in the Miami areain the 1950s was small The estimate is only about 4 (19800) of Miamirsquos 495000 population wereSpanish-language speakers in 1950 and about 70 ofthat number (13860) were estimated to be Cuban(Clark 1991) The push for the immigration ofmassive numbers of Cubans was the Cuban Revolu-tion in 1959 while the pull to the United States wasthe needs of United States Cold War foreign policyThe Cubans began to arrive in Miami in 1959 just asthe black Civil Rights Movement was getting underway and was creating opportunities for blacks acrossthe nation The arrival to a large extent of over800000 Cubans in Miami between 1959 and 1980short-circuited economic and political gains for blackMiamians (Mohl 1990)

Cubans were not the normal labor immigrantcase Most of the 1959 immigrants were from the pro-fessional class in Cuba While they initially worked inlow-paying jobs that blacks had occupied the UnitedStates government moved quickly to assist them aspart of its strategy to undermine Fidel Castro in CubaPrograms were developed that provided Cubans withfood and clothing housing assistance social servicesmedical care relocation assistance educational pro-

2According to the 2000 Census there are 3099 Chinese in Missis-sippi presently and a total Asian population including EastIndians of 18626 The most plentiful group is Vietnamese

3The Mississippi Constitution of 1890 as cited by Loewen (197167) and Quan (1982 45) stated that ldquoseparate schools shall bemaintained for children of the white and colored racesrdquo andbecause the Chinese were not of the White race they were consid-ered of the colored races In the case brought by Gong Lum agrocer in Rosedale and a US citizen (Gong Lum v Rice 1927 275US 390) the US Supreme Court upheld the Mississippi SupremeCourtrsquos decision upholding segregation of Chinese with blacks

grams job training and job placement As a resultthousands of Cuban elites were able to resume theirprofessional lives in South Florida It has been esti-mated that between 1960 and 1990 approximately $2billion was spent on Cuban resettlement activities(Mohl 1990)

The effect of massive federal financial assistance tothe new Cuban immigrants was not lost on Miamirsquosblack population who were not able to participate inthe programs established for the Cuban immigrantsAsearly as 1960 the Miami black press began to view theCuban influx as detrimental to black aspirations andachievements Blacks and Cubans began to compete for housing and residential space jobs and govern-ment services (Mohl 1990) The significant amount offederal resources going to Cubans significantly shiftedthe economic landscape in the 20 years of Cubanimmigration to Dade County black entrepreneurshipsteadily declined From 1968 to 1980 Cubans received466 of all Small Business Administration (SBA)loans in Dade County compared to only 6 to blacksAfter the four riots in the 1980s the situation worsenedwith 90 of SBA loans in Miami-Dade going toCubans and whites (Grenier and Castro 2001)

Like the Mississippi Chinese Cubans in Miamialso perceived a benefit to distancing themselves fromblack Americans The substantial assistance from andspecial status given the Cubans by the United Statesgovernment put them in a privileged position vis-agrave-vis black Americans Cubans most of whom in thefirst post-1959 wave self-identified as white perceivedthat they had little in common with blacks and situ-ated themselves alongside whites Moreover the racialstratification and racial history in Cuba was similar tothat of the United States Blacks arrived in Cuba asslaves and the post-slavery color hierarchy was one inwhich most of the social economic and political ben-efits went to those of lighter complexion while thoseCubans of African ancestry and darker skin were rel-egated to the bottom rungs of society Segregationalong with racial discrimination was prominent inpre-1959 Cuba (de la Fuente 1995 1998 Guimaraes2001 Scott 1985) Race was an old problem in Cubawith its origins in slavery and colonization and thepost-1959 white Cubans brought those attitudes withthem to the United States

Theoretical Implications

In both the Mississippi and Miami cases a system ofracial hierarchy and segregation allowed both com-munities to flourish In the Mississippi case theChinese flourished despite their status as a nonfavored

group whereas with the Cubans in Miami theirsuccess was predetermined given the conditions underwhich they entered the United States and the massivefederal government assistance If the patternsobserved in these two historical cases are predictiveit suggests that the new Latino immigrants into theSouth will distance themselves from Southern blacksRecent literature underscores the presence of this his-torical pattern In a study of black and Latino relationsin Houston TX in the 1990s Mindiola Niemann andRodriguez (2003) found that US-born Latinosexpressed more negative views of black Americansthan blacks expressed of Latinos but foreign-bornLatinos held even more negative views of black Amer-icans Moreover foreign-born Latinos were the leasttolerant of blacks

It is possible that attitudes about blacks present intheir home countries make Latino immigrants morelikely to see black Americans negatively and want todistance themselves from blacks A large literatureexists on racial hierarchy racial prejudice and dis-crimination and racial stereotypes in Latin Americathe Caribbean and Mexico (de la Cadena 2001Dulitzky 2005 Geipel 1997 Guimareas 2001Hanchard 1994 Moumlrner 1967 Sweet 1997 Wade1993 1997 Winant 1992) Literature specifically onthe racial ordering in Mexico argues that race becamea principal factor in the social and economic organi-zation of Spanish colonial society and the racial order-ing was established to benefit whites (Menchaca 2001Seed 1982) Despite the negative attitudes towardblacks on the part of Latino immigrants identified byMindiola Niemann and Rodriguez (2003) it is pos-sible that increased contact between the two groupswill reduce Latino immigrantsrsquo negative attitudestoward black Americans One approach to thinkingabout this response is through the lens of contacttheory

Contact Theory

Contact theory (or propinquity) is the springboardfor much of the literature on racial attitudes It arguesthat increased contact between two groups with negative attitudes toward each other will result in a decrease in negative attitudes (Hood and Morris1998) Originally associated with the work of Allport(1954) research on contextual determinants of racialand ethnic relations has been mixed Some scholarshave found increased contact under certain condi-tionsmdasheg interdependence common goals equalstatus and encouragement by authoritiesmdashreducesprejudice (Amir 1969 1976 Ellison and Powers 1994

rsquo

Jackman and Crane 1986 Kinder and Mendelberg1995 Powers and Ellison 1994 Sigelman and Welch1993 Stephan and Stephan 1985 Welch et al 2001)Other researchers have found that as the concentra-tion and consequently contact of two groupsincreases competition and prejudice also increase(Fossett and Kiecolt 1989 Quillian 1996) Still othershave identified the presence of both increased ani-mosity and reduced prejudice (Morris 1999 Sigelmanand Welch 1993)

Most of this research however has focused onblack and white interaction and few studies havefocused on Latino and Asian interaction with whites(Taylor 1998) Research on the effects of interminor-ity group contact on prejudice and stereotypes islimited Oliver and Wong (2003) examined attitudesof racial groups in multiracial settings finding withthe exception of Asians people who live in neighbor-hoods where their group dominates tend to harborgreater negative stereotypes about other racial minor-ity groups In other words racial stereotypes increaseas the percent of onersquos own ingroup increases in theirneighborhood Blacks and Latinos who are the mostracially isolated harbor the most negative views ofother groups but this pattern was not as pronouncedamong blacks and Latinos residing in neighborhoodsthat are racially diverse Yet they also found thathigher levels of propinquity correlated with lowerlevels of prejudice and perceived competition (Oliverand Wong 2003) Thus it appears that contact theorycan result in three different types of reactions reducedprejudice and perceived competition increased preju-dice and perceived competition or a mixture of both

Drawing from the cases of Miami and Mississippithe literature on Latino immigrant attitudes towardblack Americans and contact theory we test twohypotheses

H1 Latino immigrants hold negative stereotypes of black Americans but these stereotypes will be reduced by length of stay in the countryand increased interaction with black Americansand

H2 Latino immigrants see themselves as closer towhites than to blacks but increases in length oftime in the country and increased interactionwith black Americas will push them closer toblacks

Research Setting

Durham North Carolina is the setting for this studyThe City of Durham like many Southern locations is

undergoing demographic change Table 1 shows thechanges in the demographics of the population of theCity of Durham from 1990 to 2000 What is clear isthat the number of Latinos residing in the city hasrisen dramatically In 1990 Latinos were slightly morethan 1 of the population but by 2000 their per-centage reached 86 For decades whites were themajority in Durham (516 in 1990) but the increas-ing Latino population along with a smaller increasein the Asian population have reduced the white pro-portion to the point where in 2000 blacks and whiteswere almost equal percentages of the population455 for whites and 438 for blacks4

Durhamrsquos Latino population is antithetical to theCuban case First the current wave of Latino immi-grants does not fit into the United Statesrsquo foreignpolicy calculations of encouraging an exodus from a communist country in an effort to undermine thegovernment Thus the enormous amounts of federalaid that went to post-1959 Cubans in South Floridado not exist for these new Latino immigrantsAlthough Latinos in Durham are not benefiting fromtargeted social services as in Miami it does appearthere are special programs designed to help native

T 1 1990 and 2000 Census Data for the Cityof Durham

Change1990 2000 (+-)

Total Population 136611 187035 +

Whites 516 455 minus(70513) (85126)

Blacks 457 438 minus(62393) (81937)

American Indians 2 3 +(358) (575)

Asians 19 36 +(2676) (6815)

Latinos 13 86 +(1713) (16012)

Mexicans 308 642 +(528) (10343)

Puerto Ricans 215 67 minus(369) (696)

Cubans 88 147 +(150) (236)

Other Hispanics 389 296 minus(666) (4737)

4Both of these groups gained in absolute numbers of people butlost as a proportion of the population from 1990 to 2000

Spanish speakers receive more traditional forms ofaid Second Latinos in Durham are typically notmembers of the professional community in theirhome countries and have lower incomes and educa-tion and skill levels than the Cuban community Manyof these immigrants are from poorer countries egMexico and the countries of Central America5 There-fore Latinos and blacks might come into greater com-petition for the same jobs and social services than inMiami (for a discussion of possible job competitionsee McClain et al 2005) Third the reasons for immi-gration into the United States for recent immigrantsare primarily economic rather than political6

Why Durham North Carolina First North Car-olina has the fastest growing Latino population in thecountry It experienced an almost 500 increase in itsLatino population primarily with immigrants fromMexico skyrocketing from 76726 in 1990 to 378963in 2000 (US Census Bureau 2002a 2002b) Further-more North Carolina had the highest rate of growthin its immigrant population out of all the states in the 1990s Suro and Singer (2002) identify Raleigh-Durham as having the highest rate of Latino growthfrom 1980 to 2000mdash1180 [See figures in Table 1]Second Durham like other cities of the New Southhas experienced a decline or exit of industries whereblacks have traditionally found work As a result asubstantial portion of the black population now worksin the service industry and many (but not necessarilyall) blacks and the new Latino immigrants find them-selves competing for the same jobs Finally from aresearch and data gathering perspective Durham is ofa manageable size

Data and Measures

The analyses in this paper are based on the 2003Durham Survey of Intergroup Relations (DSIR n =500) The survey was conducted specifically for ourproject by the Center for Survey Research of the Uni-

versity of Virginia using a Computer-Aided TelephoneInterviewing (CATI) system employing random digitdialing (RDD) and dialing of directory-assisted His-panic surname sample7 A randomly generated sampleof phone numbers based on exchanges valid in theDurham North Carolina area was called An over-sample of numbers listed in the phone directory undera Hispanic surname was called at the same time8 Thesurvey was conducted from May 4 through June 222003 and interviews were conducted in both Englishand Spanish (32 of the interviews were conductedin Spanish which translates into 958 of the Latinosample opting to be interviewed in Spanish)9 Araceethnicity quota was implemented to achieve aminimum of 150 whites 150 blacks and 150 Latinosthe remaining 50 respondents were not under thisquota restriction and represent a number ofracialethnic backgrounds Interviews were completedwith 500 residents of the City of Durham for anoverall response rate of 21610 The sample of 500consists of 160 whites (32) 151 blacks (30) 167Latinos (34) 6 Asians (12) 12 who designatedtheir race as Other (24) and 1 respondent (2)who did not indicate a racial category11

5The News and Observer (Raleigh NC) identified that many of theMexican immigrants into North Carolina come primarily fromrural towns in the State of Puebla (November 29 1998 Novem-ber 30 1998) For the most part these immigrants are unskilledand poorly educated

6In a series of articles throughout 2002 chronicling the lives of arearesidents living in poverty The Herald Sun (Durham) provided apicture of life for Latinos in Durham that differed markedly fromthe Miami case Fully 26 of the more than 16000 Latinos inDurham live below the federal poverty level and in order to makea good living it is necessary for them to work more than one job(Assis and Pecquet 2002 A12)

7We recognize the problems associated with drawing a samplefrom a listing of Hispanic surnames for example missing His-panics with non-Hispanic last names and those married to non-Hispanics We also acknowledge that some Latino immigrantsmight not have phones in their homes Given the recency of theHispanic population in Durham and the high proportion ofimmigrants however we choose the sampling frame that wouldgive us the highest probabilities of getting to a Latino respondent

8Both samples were purchased from Survey Sampling Inc (SSI)of Fairfield CT

9We had the questionnaire translated by a Spanish-languageorganization in Chapel Hill NC In order to check the translationand to ensure that it tracked the English-language version thesurvey organization drew a small sample of Latinos in Durhamfor the sole purpose of checking the translation As a resultchanges were made to the translation The revised Spanish-language version was then pretested on another small sample of Latinos in Durham

10A total of 4208 phone numbers were attempted in the course ofthe survey and a total of 14014 call attempts were made TheAmerican Association for Public Opinion Research (AAPOR) ratewas calculated using the full call history of each number that wasrecorded automatically by the CATI software The response ratewas calculated according to AAPOR suggested formula RR3 withe1 = 50 and e2 = 78 We estimated e1 and e2 based on an analy-sis of residency rates and the occurrence of out-of-area householdsin our sample Partial interviews are not counted in the numera-tor of the RR3 formula

11Due to the use of the Hispanic surname sample and racialethnicquotas sampling error is more difficult to calculate The samplemay be viewed as part of two separate populations Within theRDD sample the source of 276 completions the probability of

rsquo

To assess the quality of the sample we comparedthe distribution of demographic characteristics in ourstudy with the characteristics of the population inDurham as reported in the 2000 Census Since thestudy was designed to oversample the black andLatino populations we did a detailed analysis of theextent to which each of our three main subpopulationsamples are represented in our sample the results ofwhich are not presented here12 In brief overall wefind that our sample is reasonably representative ofthe three racial subpopulations under investigationthough the distribution of respondents is slightlyolder better educated and more often female thanfound in the 2000 Census

Our data also confirm what we had suspectedmdashthe Latino population in Durham is basically animmigrant population primarily from MexicoNinety-three percent of the Latino respondents in thesample were born outside the United States Of the93 (n = 156) only about 19 were naturalized cit-izens While Mexicans were the largest portion of theLatino sample (63) Latinos from Central Americawere the next largest group (23) followed by SouthAmerican (5) Puerto Rican (4) Spanish (2)Cuban (1) and Other Latino (2) Given that weare interested in Latino immigrant attitudes most of the analyses are performed using only these 156individuals

We use two dependent variables in our analyses totest our hypotheses One dependent variablemdashLatinostereotypical views of blacksmdashis an index of threestereotype questions ranging in value from 3 (holdingthe least stereotypical views of blacks) to 15 (holdingthe most stereotypical views of blacks)13 A second

dependent variable measures which racial group(Asians blacks or whites) Latinos felt they have themost in common with (1 = have most in commonwith whites 0 = have most in common with blacks)14

We use the following predictor variables although notall are included in every equation education attain-ment measured as number of years of schooling(ranging from 0 to 25 years)15 perception of numberof blacks (Latinos) in their neighborhood16 socialcontact with blacks (Latinos)17 length of time in theUnited States as measured by a question that askedimmigrant respondents the year that they arrived18

(ranges from 0 to 35 years mean = 875 years median= 6 years)19 and gender (1 = male 0 = female) We donot include contact with blacks at work as one of thecontact indicators because of substantial missing dataEducational attainment and gender are used as demo-graphic controls

We also include the indicator of linked fate (thosebelieving that what happens to other blacks (Latinos)will have something to do with their lives)20 for severalreasons The concept of linked fate has been found tobe a powerful predictor of group consciousness polit-ical participation and political attitudes among blackAmericans (Dawson 1994 Matthews and Prothro1967 Shingles 1981 Tate 1993) The concept of linked

selection is known and the margin of error is plusmn59 Within thesurname oversample providing 244 completed interviews allhouseholds listed under a resident with a Hispanic surname wereattempted however Hispanics were included in RDD calling andnon-Hispanics were included in the oversample Non-Hispanicswith Hispanic surnames had a greater chance of selection thannon-Hispanics in the RDD sample who do not have Hispanic sur-names If we assume this to be a more or less random occurrencethen the margin of error for each of the three racialethnic groupsis roughly 8

12A copy of the full analysis is available from the lead author

13Question wording (Q1) For blacks do you think ldquowork hardrdquodescribes almost all blacks most blacks some blacks few blacksalmost no blacks (Q2) For blacks do think ldquoit is easy to get alongwith themrdquo describes (same response categories as question 1)(Q3) For blacks do you think ldquoyou can trust themrdquo describes(same response categories as question 1) We recognize that thesethree questions are the bluntest measures of stereotypes and othermore subtle measures of stereotypes exist but had to make diffi-cult choices when we had to cut the questionnaire in half in orderto stay within budget As such we decided on measuring the mostnegative stereotypes

14Question wording for Latinos (1) Of the following groups if youhad to say which one do you feel you have the most in commonwith African Americansblacks Asians or whites (2) and whichof the following groups do you feel you share the least in commonwith African Americansblacks Asians or whites We recoded theAsian category as missing since only six Latino respondentsreported feeling that Latinos have the most in common withAsians Coding for the dependent variable is 1 (have most incommon with whites) and 0 (have most in common with blacks)

15Income had initially been included in our model but wasdropped due to suspected collinearity with education

16Question wording In your neighborhood how many of the res-idents are black (Latinos) (1) none of the residents (2) a few ofthe residents (3) some of the residents (4) most of the residents(5) all of the residents

17Question wording In your social life including your friends andpeople you know from church and other social activities do youhave (4) a lot of contact (3) some contact (2) little contact or (1)no contact at all with blacks (Latinos)

18We converted this measure to years in the United States by subtracting the year arrived from 2003 (the year the survey wasconducted)

19Although our sample is of Latino immigrants some of whomare new arrivals the mean (875 years) and median (six years) ofthis predicator suggest substantial variance and is thus an ade-quate measure of length of time in country

20Question wording Do you think what happens to Hispanics inthis country will have something to do with what happens in yourlife

fate as Dawson (1994 76) suggests evolves fromsocial identity theory that suggests individuals formtheir concepts of self in part by identifying similari-ties and differences between themselves and othersIndividuals perceive a similarity in experience andtreatment or perceive a linked fate with others (Davisand Brown 2002 241) Only a few studies have exam-ined concepts similar to linked fate for Latinos Kauf-mann (2003) found that Latinos who feel close to oneanother as a group were much more likely to feel closeto black Americans In view of the significance oflinked fate for black Americans and the limited lit-erature suggesting that feelings of closeness to otherLatinos may influence Latino relations with blackAmericans we want to see if Latino linked fate existswithin Latino immigrant populations and if it doeswhat effect does it have on Latino immigrantsrsquo per-ceptions of black Americans The data indicate two-thirds (6791) of Latino immigrants feel a sense oflinked fate with other Latinos This would suggest atleast as reflected in our survey linked fate does existamong Latino immigrants Latino linked fate is meas-ured as 1 = yes and 0 = no

Although we have not specified hypotheses forblacks we examine similar equations for blacks inorder to get a sense of the reciprocal nature of the rela-tionship (n = 151) Not all indicators used for Latinoshowever are appropriate to use for blacks eg lengthof time in country as approximately 95 of blacks arenative-born We use OLS and logit regression analy-ses to test our hypotheses We use Clarify which makesit possible to compute predicted probabilities forvalues of the dependent variable while holding thepredictor variables at their means or some other valueon several of the logit analyses (King Tomz and Wittenberg 2000 Tomz Wittenberg and King 2003)

Testing and Results

Analyses for our first hypothesis indicates that theprevalence of negative stereotypes of black Americansin the Latino immigrant community is quite wide-spread and seem especially so when compared to theprevalence of white stereotypes of blacks Along eachdimension the stereotypes of blacks by Latinos aremore negative than those of white respondents (SeeTable A1 in the JOP online appendix at httpwwwjournalofpoliticsorg) We found that 589 ofthe Latino immigrants in our study reported feelingthat few or almost no blacks are hard workingapproximately one-third (325) of the Latino immi-grant respondents reported feeling that few or almost

no blacks are easy to get along with and slightly morethan a majority (569) of the Latino immigrantrespondents reported feeling that few or almost noblacks could be trusted Among whites the compara-ble figures are only 93 indicate that few or almostno blacks are hard working only 84 believe that fewor almost no blacks are easy to get along with andonly 96 feel that few or almost no blacks could betrusted Consistent with our theoretical expectationsit does appear that many Latinos hold very negativestereotypes of blacks

What factors influence the stereotypes thatLatinos hold of black Americans Table 2 presents theresults of our OLS analysis Even given the relativelysmall number of observations almost all of our pre-dictors are significant and in the hypothesized direc-tion To begin the two demographic control variableseducation and male are both statistically significantMore educated Latinos have significantly less negativestereotypes while men appear to have significantlymore negative stereotypes (perhaps reflecting a pre-disposition of men to use more extreme values onsurvey instruments) More interesting of course arethe influence of the remaining variables in the model

T 2 OLS Regression Results Predicting theExtent to Which Latino ImmigrantsHold Negative Stereotypes of Blacks

Variable Model

Main EffectsEducation minus1033

(0421)Male 7860

(3789)Linked Fate minus7472

(4225)Amount of Social Contact with Blacks minus4780

(1959)Number of Blacks in Neighborhood 2294

(1926)Length of Time in the United States minus0295

(0269)

Constant 12216N 130F Statistic 446R-Squared 1788

Note The dependent variable stereotypical views of blacksranges in value from 3 (holding the least stereotypical views ofblacks) to 15 (holding the most stereotypical views of blacks)Standard error estimates are in parenthesesp lt 01 (two-tailed test)p lt 05 (two-tailed test)p lt 10 (two-tailed test)

rsquo

American population Both Chinese immigrants tothe South during the late-nineteenth century andCuban immigrants during the mid-twentieth centurypursued a strategy of racial distancing seeing them-selves as being in economic and social competitionwith black Americans rather than as natural allies inthe fight for social and political equality

Chinese in Mississippi

The Chinese entered Southern history during Recon-struction as part of a strategy devised by Southernplanters to retain political power and economiccontrol over the newly freed blacks (Loewen 1971Quan 1982) The plantersrsquo strategy was the importa-tion and employment of Chinese men in occupationspreviously held by blacks (Quan 1982 5) From theplantersrsquo perspective the Chinese were ideal On theone hand they would take jobs from blacks forcingthem into a subordinate position while on the otherthey were barred from citizenship could not vote andthus would not present a potential political threat(Loewen 1971 22) This attempt however was anutter failure and by 1880 there were only 51 Chineseinhabitants of the Mississippi Delta (Loewen 197125)2 These numbers increased however as Chinesemen brought over Chinese brides

The Chinese were immediately confronted withthe Southrsquos racial hierarchy where they were classifiedwith blacks and subject to some of the same segrega-tion laws (Loewen 1971 23ndash24) Despite the cordialrelations Chinese apparently shared initially withblacks they forcefully resisted their social and legaldesignation with blacks In fact Mississippi Chinesefiled one of the earliest legal challenges to school seg-regation in the United States in 19243 Historians agreethat the Chinese actively sought white approval andsought to distance themselves from blacks (Loewen1971 Quan 1982) Moreover to insure their middleposition between whites and blacks the Chinese toldwhites that they were not interested in marryingwhites and were definitely not interested in marryingblacks (Loewen 1971 79) They were successful in

convincing whites that they were more like whitesthan they were like blacks (whitening up) and theyshared whitesrsquo negative attitudes toward and approvedof their treatment of blacks

During the Civil Rights Movement (CRM) rela-tions between blacks and Chinese were strained evenfurther Blacks knew the Chinese community favoredwhites and the CRM served to make that choice moreobvious In 1970 when school desegregation came toMississippi both whites and Chinese enrolled theirchildren in private academies (Loewen 1971 177) TheDelta Chinese saw the CRM as threatening the systemof racial segregation that enabled them to occupy amiddling space in the oppressive Mississippi environ-ment (Loewen 1971 Quan 1982) In brief in order toattain a certain degree of social standing the Missis-sippi Chinese had to actively distance themselves fromblacks to show whites that they had no sympathy withblacks and their plight As such the MississippiChinese began to see themselves as having more incommon with whites than with blacks and acted onthat perceived commonality

Cubans in Miami

Miami is probably the best-known instance of theintroduction of substantial numbers of a foreign pop-ulation into the South but it did not occur en masseuntil 1959 The number of Cubans in the Miami areain the 1950s was small The estimate is only about 4 (19800) of Miamirsquos 495000 population wereSpanish-language speakers in 1950 and about 70 ofthat number (13860) were estimated to be Cuban(Clark 1991) The push for the immigration ofmassive numbers of Cubans was the Cuban Revolu-tion in 1959 while the pull to the United States wasthe needs of United States Cold War foreign policyThe Cubans began to arrive in Miami in 1959 just asthe black Civil Rights Movement was getting underway and was creating opportunities for blacks acrossthe nation The arrival to a large extent of over800000 Cubans in Miami between 1959 and 1980short-circuited economic and political gains for blackMiamians (Mohl 1990)

Cubans were not the normal labor immigrantcase Most of the 1959 immigrants were from the pro-fessional class in Cuba While they initially worked inlow-paying jobs that blacks had occupied the UnitedStates government moved quickly to assist them aspart of its strategy to undermine Fidel Castro in CubaPrograms were developed that provided Cubans withfood and clothing housing assistance social servicesmedical care relocation assistance educational pro-

2According to the 2000 Census there are 3099 Chinese in Missis-sippi presently and a total Asian population including EastIndians of 18626 The most plentiful group is Vietnamese

3The Mississippi Constitution of 1890 as cited by Loewen (197167) and Quan (1982 45) stated that ldquoseparate schools shall bemaintained for children of the white and colored racesrdquo andbecause the Chinese were not of the White race they were consid-ered of the colored races In the case brought by Gong Lum agrocer in Rosedale and a US citizen (Gong Lum v Rice 1927 275US 390) the US Supreme Court upheld the Mississippi SupremeCourtrsquos decision upholding segregation of Chinese with blacks

grams job training and job placement As a resultthousands of Cuban elites were able to resume theirprofessional lives in South Florida It has been esti-mated that between 1960 and 1990 approximately $2billion was spent on Cuban resettlement activities(Mohl 1990)

The effect of massive federal financial assistance tothe new Cuban immigrants was not lost on Miamirsquosblack population who were not able to participate inthe programs established for the Cuban immigrantsAsearly as 1960 the Miami black press began to view theCuban influx as detrimental to black aspirations andachievements Blacks and Cubans began to compete for housing and residential space jobs and govern-ment services (Mohl 1990) The significant amount offederal resources going to Cubans significantly shiftedthe economic landscape in the 20 years of Cubanimmigration to Dade County black entrepreneurshipsteadily declined From 1968 to 1980 Cubans received466 of all Small Business Administration (SBA)loans in Dade County compared to only 6 to blacksAfter the four riots in the 1980s the situation worsenedwith 90 of SBA loans in Miami-Dade going toCubans and whites (Grenier and Castro 2001)

Like the Mississippi Chinese Cubans in Miamialso perceived a benefit to distancing themselves fromblack Americans The substantial assistance from andspecial status given the Cubans by the United Statesgovernment put them in a privileged position vis-agrave-vis black Americans Cubans most of whom in thefirst post-1959 wave self-identified as white perceivedthat they had little in common with blacks and situ-ated themselves alongside whites Moreover the racialstratification and racial history in Cuba was similar tothat of the United States Blacks arrived in Cuba asslaves and the post-slavery color hierarchy was one inwhich most of the social economic and political ben-efits went to those of lighter complexion while thoseCubans of African ancestry and darker skin were rel-egated to the bottom rungs of society Segregationalong with racial discrimination was prominent inpre-1959 Cuba (de la Fuente 1995 1998 Guimaraes2001 Scott 1985) Race was an old problem in Cubawith its origins in slavery and colonization and thepost-1959 white Cubans brought those attitudes withthem to the United States

Theoretical Implications

In both the Mississippi and Miami cases a system ofracial hierarchy and segregation allowed both com-munities to flourish In the Mississippi case theChinese flourished despite their status as a nonfavored

group whereas with the Cubans in Miami theirsuccess was predetermined given the conditions underwhich they entered the United States and the massivefederal government assistance If the patternsobserved in these two historical cases are predictiveit suggests that the new Latino immigrants into theSouth will distance themselves from Southern blacksRecent literature underscores the presence of this his-torical pattern In a study of black and Latino relationsin Houston TX in the 1990s Mindiola Niemann andRodriguez (2003) found that US-born Latinosexpressed more negative views of black Americansthan blacks expressed of Latinos but foreign-bornLatinos held even more negative views of black Amer-icans Moreover foreign-born Latinos were the leasttolerant of blacks

It is possible that attitudes about blacks present intheir home countries make Latino immigrants morelikely to see black Americans negatively and want todistance themselves from blacks A large literatureexists on racial hierarchy racial prejudice and dis-crimination and racial stereotypes in Latin Americathe Caribbean and Mexico (de la Cadena 2001Dulitzky 2005 Geipel 1997 Guimareas 2001Hanchard 1994 Moumlrner 1967 Sweet 1997 Wade1993 1997 Winant 1992) Literature specifically onthe racial ordering in Mexico argues that race becamea principal factor in the social and economic organi-zation of Spanish colonial society and the racial order-ing was established to benefit whites (Menchaca 2001Seed 1982) Despite the negative attitudes towardblacks on the part of Latino immigrants identified byMindiola Niemann and Rodriguez (2003) it is pos-sible that increased contact between the two groupswill reduce Latino immigrantsrsquo negative attitudestoward black Americans One approach to thinkingabout this response is through the lens of contacttheory

Contact Theory

Contact theory (or propinquity) is the springboardfor much of the literature on racial attitudes It arguesthat increased contact between two groups with negative attitudes toward each other will result in a decrease in negative attitudes (Hood and Morris1998) Originally associated with the work of Allport(1954) research on contextual determinants of racialand ethnic relations has been mixed Some scholarshave found increased contact under certain condi-tionsmdasheg interdependence common goals equalstatus and encouragement by authoritiesmdashreducesprejudice (Amir 1969 1976 Ellison and Powers 1994

rsquo

Jackman and Crane 1986 Kinder and Mendelberg1995 Powers and Ellison 1994 Sigelman and Welch1993 Stephan and Stephan 1985 Welch et al 2001)Other researchers have found that as the concentra-tion and consequently contact of two groupsincreases competition and prejudice also increase(Fossett and Kiecolt 1989 Quillian 1996) Still othershave identified the presence of both increased ani-mosity and reduced prejudice (Morris 1999 Sigelmanand Welch 1993)

Most of this research however has focused onblack and white interaction and few studies havefocused on Latino and Asian interaction with whites(Taylor 1998) Research on the effects of interminor-ity group contact on prejudice and stereotypes islimited Oliver and Wong (2003) examined attitudesof racial groups in multiracial settings finding withthe exception of Asians people who live in neighbor-hoods where their group dominates tend to harborgreater negative stereotypes about other racial minor-ity groups In other words racial stereotypes increaseas the percent of onersquos own ingroup increases in theirneighborhood Blacks and Latinos who are the mostracially isolated harbor the most negative views ofother groups but this pattern was not as pronouncedamong blacks and Latinos residing in neighborhoodsthat are racially diverse Yet they also found thathigher levels of propinquity correlated with lowerlevels of prejudice and perceived competition (Oliverand Wong 2003) Thus it appears that contact theorycan result in three different types of reactions reducedprejudice and perceived competition increased preju-dice and perceived competition or a mixture of both

Drawing from the cases of Miami and Mississippithe literature on Latino immigrant attitudes towardblack Americans and contact theory we test twohypotheses

H1 Latino immigrants hold negative stereotypes of black Americans but these stereotypes will be reduced by length of stay in the countryand increased interaction with black Americansand

H2 Latino immigrants see themselves as closer towhites than to blacks but increases in length oftime in the country and increased interactionwith black Americas will push them closer toblacks

Research Setting

Durham North Carolina is the setting for this studyThe City of Durham like many Southern locations is

undergoing demographic change Table 1 shows thechanges in the demographics of the population of theCity of Durham from 1990 to 2000 What is clear isthat the number of Latinos residing in the city hasrisen dramatically In 1990 Latinos were slightly morethan 1 of the population but by 2000 their per-centage reached 86 For decades whites were themajority in Durham (516 in 1990) but the increas-ing Latino population along with a smaller increasein the Asian population have reduced the white pro-portion to the point where in 2000 blacks and whiteswere almost equal percentages of the population455 for whites and 438 for blacks4

Durhamrsquos Latino population is antithetical to theCuban case First the current wave of Latino immi-grants does not fit into the United Statesrsquo foreignpolicy calculations of encouraging an exodus from a communist country in an effort to undermine thegovernment Thus the enormous amounts of federalaid that went to post-1959 Cubans in South Floridado not exist for these new Latino immigrantsAlthough Latinos in Durham are not benefiting fromtargeted social services as in Miami it does appearthere are special programs designed to help native

T 1 1990 and 2000 Census Data for the Cityof Durham

Change1990 2000 (+-)

Total Population 136611 187035 +

Whites 516 455 minus(70513) (85126)

Blacks 457 438 minus(62393) (81937)

American Indians 2 3 +(358) (575)

Asians 19 36 +(2676) (6815)

Latinos 13 86 +(1713) (16012)

Mexicans 308 642 +(528) (10343)

Puerto Ricans 215 67 minus(369) (696)

Cubans 88 147 +(150) (236)

Other Hispanics 389 296 minus(666) (4737)

4Both of these groups gained in absolute numbers of people butlost as a proportion of the population from 1990 to 2000

Spanish speakers receive more traditional forms ofaid Second Latinos in Durham are typically notmembers of the professional community in theirhome countries and have lower incomes and educa-tion and skill levels than the Cuban community Manyof these immigrants are from poorer countries egMexico and the countries of Central America5 There-fore Latinos and blacks might come into greater com-petition for the same jobs and social services than inMiami (for a discussion of possible job competitionsee McClain et al 2005) Third the reasons for immi-gration into the United States for recent immigrantsare primarily economic rather than political6

Why Durham North Carolina First North Car-olina has the fastest growing Latino population in thecountry It experienced an almost 500 increase in itsLatino population primarily with immigrants fromMexico skyrocketing from 76726 in 1990 to 378963in 2000 (US Census Bureau 2002a 2002b) Further-more North Carolina had the highest rate of growthin its immigrant population out of all the states in the 1990s Suro and Singer (2002) identify Raleigh-Durham as having the highest rate of Latino growthfrom 1980 to 2000mdash1180 [See figures in Table 1]Second Durham like other cities of the New Southhas experienced a decline or exit of industries whereblacks have traditionally found work As a result asubstantial portion of the black population now worksin the service industry and many (but not necessarilyall) blacks and the new Latino immigrants find them-selves competing for the same jobs Finally from aresearch and data gathering perspective Durham is ofa manageable size

Data and Measures

The analyses in this paper are based on the 2003Durham Survey of Intergroup Relations (DSIR n =500) The survey was conducted specifically for ourproject by the Center for Survey Research of the Uni-

versity of Virginia using a Computer-Aided TelephoneInterviewing (CATI) system employing random digitdialing (RDD) and dialing of directory-assisted His-panic surname sample7 A randomly generated sampleof phone numbers based on exchanges valid in theDurham North Carolina area was called An over-sample of numbers listed in the phone directory undera Hispanic surname was called at the same time8 Thesurvey was conducted from May 4 through June 222003 and interviews were conducted in both Englishand Spanish (32 of the interviews were conductedin Spanish which translates into 958 of the Latinosample opting to be interviewed in Spanish)9 Araceethnicity quota was implemented to achieve aminimum of 150 whites 150 blacks and 150 Latinosthe remaining 50 respondents were not under thisquota restriction and represent a number ofracialethnic backgrounds Interviews were completedwith 500 residents of the City of Durham for anoverall response rate of 21610 The sample of 500consists of 160 whites (32) 151 blacks (30) 167Latinos (34) 6 Asians (12) 12 who designatedtheir race as Other (24) and 1 respondent (2)who did not indicate a racial category11

5The News and Observer (Raleigh NC) identified that many of theMexican immigrants into North Carolina come primarily fromrural towns in the State of Puebla (November 29 1998 Novem-ber 30 1998) For the most part these immigrants are unskilledand poorly educated

6In a series of articles throughout 2002 chronicling the lives of arearesidents living in poverty The Herald Sun (Durham) provided apicture of life for Latinos in Durham that differed markedly fromthe Miami case Fully 26 of the more than 16000 Latinos inDurham live below the federal poverty level and in order to makea good living it is necessary for them to work more than one job(Assis and Pecquet 2002 A12)

7We recognize the problems associated with drawing a samplefrom a listing of Hispanic surnames for example missing His-panics with non-Hispanic last names and those married to non-Hispanics We also acknowledge that some Latino immigrantsmight not have phones in their homes Given the recency of theHispanic population in Durham and the high proportion ofimmigrants however we choose the sampling frame that wouldgive us the highest probabilities of getting to a Latino respondent

8Both samples were purchased from Survey Sampling Inc (SSI)of Fairfield CT

9We had the questionnaire translated by a Spanish-languageorganization in Chapel Hill NC In order to check the translationand to ensure that it tracked the English-language version thesurvey organization drew a small sample of Latinos in Durhamfor the sole purpose of checking the translation As a resultchanges were made to the translation The revised Spanish-language version was then pretested on another small sample of Latinos in Durham

10A total of 4208 phone numbers were attempted in the course ofthe survey and a total of 14014 call attempts were made TheAmerican Association for Public Opinion Research (AAPOR) ratewas calculated using the full call history of each number that wasrecorded automatically by the CATI software The response ratewas calculated according to AAPOR suggested formula RR3 withe1 = 50 and e2 = 78 We estimated e1 and e2 based on an analy-sis of residency rates and the occurrence of out-of-area householdsin our sample Partial interviews are not counted in the numera-tor of the RR3 formula

11Due to the use of the Hispanic surname sample and racialethnicquotas sampling error is more difficult to calculate The samplemay be viewed as part of two separate populations Within theRDD sample the source of 276 completions the probability of

rsquo

To assess the quality of the sample we comparedthe distribution of demographic characteristics in ourstudy with the characteristics of the population inDurham as reported in the 2000 Census Since thestudy was designed to oversample the black andLatino populations we did a detailed analysis of theextent to which each of our three main subpopulationsamples are represented in our sample the results ofwhich are not presented here12 In brief overall wefind that our sample is reasonably representative ofthe three racial subpopulations under investigationthough the distribution of respondents is slightlyolder better educated and more often female thanfound in the 2000 Census

Our data also confirm what we had suspectedmdashthe Latino population in Durham is basically animmigrant population primarily from MexicoNinety-three percent of the Latino respondents in thesample were born outside the United States Of the93 (n = 156) only about 19 were naturalized cit-izens While Mexicans were the largest portion of theLatino sample (63) Latinos from Central Americawere the next largest group (23) followed by SouthAmerican (5) Puerto Rican (4) Spanish (2)Cuban (1) and Other Latino (2) Given that weare interested in Latino immigrant attitudes most of the analyses are performed using only these 156individuals

We use two dependent variables in our analyses totest our hypotheses One dependent variablemdashLatinostereotypical views of blacksmdashis an index of threestereotype questions ranging in value from 3 (holdingthe least stereotypical views of blacks) to 15 (holdingthe most stereotypical views of blacks)13 A second

dependent variable measures which racial group(Asians blacks or whites) Latinos felt they have themost in common with (1 = have most in commonwith whites 0 = have most in common with blacks)14

We use the following predictor variables although notall are included in every equation education attain-ment measured as number of years of schooling(ranging from 0 to 25 years)15 perception of numberof blacks (Latinos) in their neighborhood16 socialcontact with blacks (Latinos)17 length of time in theUnited States as measured by a question that askedimmigrant respondents the year that they arrived18

(ranges from 0 to 35 years mean = 875 years median= 6 years)19 and gender (1 = male 0 = female) We donot include contact with blacks at work as one of thecontact indicators because of substantial missing dataEducational attainment and gender are used as demo-graphic controls

We also include the indicator of linked fate (thosebelieving that what happens to other blacks (Latinos)will have something to do with their lives)20 for severalreasons The concept of linked fate has been found tobe a powerful predictor of group consciousness polit-ical participation and political attitudes among blackAmericans (Dawson 1994 Matthews and Prothro1967 Shingles 1981 Tate 1993) The concept of linked

selection is known and the margin of error is plusmn59 Within thesurname oversample providing 244 completed interviews allhouseholds listed under a resident with a Hispanic surname wereattempted however Hispanics were included in RDD calling andnon-Hispanics were included in the oversample Non-Hispanicswith Hispanic surnames had a greater chance of selection thannon-Hispanics in the RDD sample who do not have Hispanic sur-names If we assume this to be a more or less random occurrencethen the margin of error for each of the three racialethnic groupsis roughly 8

12A copy of the full analysis is available from the lead author

13Question wording (Q1) For blacks do you think ldquowork hardrdquodescribes almost all blacks most blacks some blacks few blacksalmost no blacks (Q2) For blacks do think ldquoit is easy to get alongwith themrdquo describes (same response categories as question 1)(Q3) For blacks do you think ldquoyou can trust themrdquo describes(same response categories as question 1) We recognize that thesethree questions are the bluntest measures of stereotypes and othermore subtle measures of stereotypes exist but had to make diffi-cult choices when we had to cut the questionnaire in half in orderto stay within budget As such we decided on measuring the mostnegative stereotypes

14Question wording for Latinos (1) Of the following groups if youhad to say which one do you feel you have the most in commonwith African Americansblacks Asians or whites (2) and whichof the following groups do you feel you share the least in commonwith African Americansblacks Asians or whites We recoded theAsian category as missing since only six Latino respondentsreported feeling that Latinos have the most in common withAsians Coding for the dependent variable is 1 (have most incommon with whites) and 0 (have most in common with blacks)

15Income had initially been included in our model but wasdropped due to suspected collinearity with education

16Question wording In your neighborhood how many of the res-idents are black (Latinos) (1) none of the residents (2) a few ofthe residents (3) some of the residents (4) most of the residents(5) all of the residents

17Question wording In your social life including your friends andpeople you know from church and other social activities do youhave (4) a lot of contact (3) some contact (2) little contact or (1)no contact at all with blacks (Latinos)

18We converted this measure to years in the United States by subtracting the year arrived from 2003 (the year the survey wasconducted)

19Although our sample is of Latino immigrants some of whomare new arrivals the mean (875 years) and median (six years) ofthis predicator suggest substantial variance and is thus an ade-quate measure of length of time in country

20Question wording Do you think what happens to Hispanics inthis country will have something to do with what happens in yourlife

fate as Dawson (1994 76) suggests evolves fromsocial identity theory that suggests individuals formtheir concepts of self in part by identifying similari-ties and differences between themselves and othersIndividuals perceive a similarity in experience andtreatment or perceive a linked fate with others (Davisand Brown 2002 241) Only a few studies have exam-ined concepts similar to linked fate for Latinos Kauf-mann (2003) found that Latinos who feel close to oneanother as a group were much more likely to feel closeto black Americans In view of the significance oflinked fate for black Americans and the limited lit-erature suggesting that feelings of closeness to otherLatinos may influence Latino relations with blackAmericans we want to see if Latino linked fate existswithin Latino immigrant populations and if it doeswhat effect does it have on Latino immigrantsrsquo per-ceptions of black Americans The data indicate two-thirds (6791) of Latino immigrants feel a sense oflinked fate with other Latinos This would suggest atleast as reflected in our survey linked fate does existamong Latino immigrants Latino linked fate is meas-ured as 1 = yes and 0 = no

Although we have not specified hypotheses forblacks we examine similar equations for blacks inorder to get a sense of the reciprocal nature of the rela-tionship (n = 151) Not all indicators used for Latinoshowever are appropriate to use for blacks eg lengthof time in country as approximately 95 of blacks arenative-born We use OLS and logit regression analy-ses to test our hypotheses We use Clarify which makesit possible to compute predicted probabilities forvalues of the dependent variable while holding thepredictor variables at their means or some other valueon several of the logit analyses (King Tomz and Wittenberg 2000 Tomz Wittenberg and King 2003)

Testing and Results

Analyses for our first hypothesis indicates that theprevalence of negative stereotypes of black Americansin the Latino immigrant community is quite wide-spread and seem especially so when compared to theprevalence of white stereotypes of blacks Along eachdimension the stereotypes of blacks by Latinos aremore negative than those of white respondents (SeeTable A1 in the JOP online appendix at httpwwwjournalofpoliticsorg) We found that 589 ofthe Latino immigrants in our study reported feelingthat few or almost no blacks are hard workingapproximately one-third (325) of the Latino immi-grant respondents reported feeling that few or almost

no blacks are easy to get along with and slightly morethan a majority (569) of the Latino immigrantrespondents reported feeling that few or almost noblacks could be trusted Among whites the compara-ble figures are only 93 indicate that few or almostno blacks are hard working only 84 believe that fewor almost no blacks are easy to get along with andonly 96 feel that few or almost no blacks could betrusted Consistent with our theoretical expectationsit does appear that many Latinos hold very negativestereotypes of blacks

What factors influence the stereotypes thatLatinos hold of black Americans Table 2 presents theresults of our OLS analysis Even given the relativelysmall number of observations almost all of our pre-dictors are significant and in the hypothesized direc-tion To begin the two demographic control variableseducation and male are both statistically significantMore educated Latinos have significantly less negativestereotypes while men appear to have significantlymore negative stereotypes (perhaps reflecting a pre-disposition of men to use more extreme values onsurvey instruments) More interesting of course arethe influence of the remaining variables in the model

T 2 OLS Regression Results Predicting theExtent to Which Latino ImmigrantsHold Negative Stereotypes of Blacks

Variable Model

Main EffectsEducation minus1033

(0421)Male 7860

(3789)Linked Fate minus7472

(4225)Amount of Social Contact with Blacks minus4780

(1959)Number of Blacks in Neighborhood 2294

(1926)Length of Time in the United States minus0295

(0269)

Constant 12216N 130F Statistic 446R-Squared 1788

Note The dependent variable stereotypical views of blacksranges in value from 3 (holding the least stereotypical views ofblacks) to 15 (holding the most stereotypical views of blacks)Standard error estimates are in parenthesesp lt 01 (two-tailed test)p lt 05 (two-tailed test)p lt 10 (two-tailed test)

grams job training and job placement As a resultthousands of Cuban elites were able to resume theirprofessional lives in South Florida It has been esti-mated that between 1960 and 1990 approximately $2billion was spent on Cuban resettlement activities(Mohl 1990)

The effect of massive federal financial assistance tothe new Cuban immigrants was not lost on Miamirsquosblack population who were not able to participate inthe programs established for the Cuban immigrantsAsearly as 1960 the Miami black press began to view theCuban influx as detrimental to black aspirations andachievements Blacks and Cubans began to compete for housing and residential space jobs and govern-ment services (Mohl 1990) The significant amount offederal resources going to Cubans significantly shiftedthe economic landscape in the 20 years of Cubanimmigration to Dade County black entrepreneurshipsteadily declined From 1968 to 1980 Cubans received466 of all Small Business Administration (SBA)loans in Dade County compared to only 6 to blacksAfter the four riots in the 1980s the situation worsenedwith 90 of SBA loans in Miami-Dade going toCubans and whites (Grenier and Castro 2001)

Like the Mississippi Chinese Cubans in Miamialso perceived a benefit to distancing themselves fromblack Americans The substantial assistance from andspecial status given the Cubans by the United Statesgovernment put them in a privileged position vis-agrave-vis black Americans Cubans most of whom in thefirst post-1959 wave self-identified as white perceivedthat they had little in common with blacks and situ-ated themselves alongside whites Moreover the racialstratification and racial history in Cuba was similar tothat of the United States Blacks arrived in Cuba asslaves and the post-slavery color hierarchy was one inwhich most of the social economic and political ben-efits went to those of lighter complexion while thoseCubans of African ancestry and darker skin were rel-egated to the bottom rungs of society Segregationalong with racial discrimination was prominent inpre-1959 Cuba (de la Fuente 1995 1998 Guimaraes2001 Scott 1985) Race was an old problem in Cubawith its origins in slavery and colonization and thepost-1959 white Cubans brought those attitudes withthem to the United States

Theoretical Implications

In both the Mississippi and Miami cases a system ofracial hierarchy and segregation allowed both com-munities to flourish In the Mississippi case theChinese flourished despite their status as a nonfavored

group whereas with the Cubans in Miami theirsuccess was predetermined given the conditions underwhich they entered the United States and the massivefederal government assistance If the patternsobserved in these two historical cases are predictiveit suggests that the new Latino immigrants into theSouth will distance themselves from Southern blacksRecent literature underscores the presence of this his-torical pattern In a study of black and Latino relationsin Houston TX in the 1990s Mindiola Niemann andRodriguez (2003) found that US-born Latinosexpressed more negative views of black Americansthan blacks expressed of Latinos but foreign-bornLatinos held even more negative views of black Amer-icans Moreover foreign-born Latinos were the leasttolerant of blacks

It is possible that attitudes about blacks present intheir home countries make Latino immigrants morelikely to see black Americans negatively and want todistance themselves from blacks A large literatureexists on racial hierarchy racial prejudice and dis-crimination and racial stereotypes in Latin Americathe Caribbean and Mexico (de la Cadena 2001Dulitzky 2005 Geipel 1997 Guimareas 2001Hanchard 1994 Moumlrner 1967 Sweet 1997 Wade1993 1997 Winant 1992) Literature specifically onthe racial ordering in Mexico argues that race becamea principal factor in the social and economic organi-zation of Spanish colonial society and the racial order-ing was established to benefit whites (Menchaca 2001Seed 1982) Despite the negative attitudes towardblacks on the part of Latino immigrants identified byMindiola Niemann and Rodriguez (2003) it is pos-sible that increased contact between the two groupswill reduce Latino immigrantsrsquo negative attitudestoward black Americans One approach to thinkingabout this response is through the lens of contacttheory

Contact Theory

Contact theory (or propinquity) is the springboardfor much of the literature on racial attitudes It arguesthat increased contact between two groups with negative attitudes toward each other will result in a decrease in negative attitudes (Hood and Morris1998) Originally associated with the work of Allport(1954) research on contextual determinants of racialand ethnic relations has been mixed Some scholarshave found increased contact under certain condi-tionsmdasheg interdependence common goals equalstatus and encouragement by authoritiesmdashreducesprejudice (Amir 1969 1976 Ellison and Powers 1994

rsquo

Jackman and Crane 1986 Kinder and Mendelberg1995 Powers and Ellison 1994 Sigelman and Welch1993 Stephan and Stephan 1985 Welch et al 2001)Other researchers have found that as the concentra-tion and consequently contact of two groupsincreases competition and prejudice also increase(Fossett and Kiecolt 1989 Quillian 1996) Still othershave identified the presence of both increased ani-mosity and reduced prejudice (Morris 1999 Sigelmanand Welch 1993)

Most of this research however has focused onblack and white interaction and few studies havefocused on Latino and Asian interaction with whites(Taylor 1998) Research on the effects of interminor-ity group contact on prejudice and stereotypes islimited Oliver and Wong (2003) examined attitudesof racial groups in multiracial settings finding withthe exception of Asians people who live in neighbor-hoods where their group dominates tend to harborgreater negative stereotypes about other racial minor-ity groups In other words racial stereotypes increaseas the percent of onersquos own ingroup increases in theirneighborhood Blacks and Latinos who are the mostracially isolated harbor the most negative views ofother groups but this pattern was not as pronouncedamong blacks and Latinos residing in neighborhoodsthat are racially diverse Yet they also found thathigher levels of propinquity correlated with lowerlevels of prejudice and perceived competition (Oliverand Wong 2003) Thus it appears that contact theorycan result in three different types of reactions reducedprejudice and perceived competition increased preju-dice and perceived competition or a mixture of both

Drawing from the cases of Miami and Mississippithe literature on Latino immigrant attitudes towardblack Americans and contact theory we test twohypotheses

H1 Latino immigrants hold negative stereotypes of black Americans but these stereotypes will be reduced by length of stay in the countryand increased interaction with black Americansand

H2 Latino immigrants see themselves as closer towhites than to blacks but increases in length oftime in the country and increased interactionwith black Americas will push them closer toblacks

Research Setting

Durham North Carolina is the setting for this studyThe City of Durham like many Southern locations is

undergoing demographic change Table 1 shows thechanges in the demographics of the population of theCity of Durham from 1990 to 2000 What is clear isthat the number of Latinos residing in the city hasrisen dramatically In 1990 Latinos were slightly morethan 1 of the population but by 2000 their per-centage reached 86 For decades whites were themajority in Durham (516 in 1990) but the increas-ing Latino population along with a smaller increasein the Asian population have reduced the white pro-portion to the point where in 2000 blacks and whiteswere almost equal percentages of the population455 for whites and 438 for blacks4

Durhamrsquos Latino population is antithetical to theCuban case First the current wave of Latino immi-grants does not fit into the United Statesrsquo foreignpolicy calculations of encouraging an exodus from a communist country in an effort to undermine thegovernment Thus the enormous amounts of federalaid that went to post-1959 Cubans in South Floridado not exist for these new Latino immigrantsAlthough Latinos in Durham are not benefiting fromtargeted social services as in Miami it does appearthere are special programs designed to help native

T 1 1990 and 2000 Census Data for the Cityof Durham

Change1990 2000 (+-)

Total Population 136611 187035 +

Whites 516 455 minus(70513) (85126)

Blacks 457 438 minus(62393) (81937)

American Indians 2 3 +(358) (575)

Asians 19 36 +(2676) (6815)

Latinos 13 86 +(1713) (16012)

Mexicans 308 642 +(528) (10343)

Puerto Ricans 215 67 minus(369) (696)

Cubans 88 147 +(150) (236)

Other Hispanics 389 296 minus(666) (4737)

4Both of these groups gained in absolute numbers of people butlost as a proportion of the population from 1990 to 2000

Spanish speakers receive more traditional forms ofaid Second Latinos in Durham are typically notmembers of the professional community in theirhome countries and have lower incomes and educa-tion and skill levels than the Cuban community Manyof these immigrants are from poorer countries egMexico and the countries of Central America5 There-fore Latinos and blacks might come into greater com-petition for the same jobs and social services than inMiami (for a discussion of possible job competitionsee McClain et al 2005) Third the reasons for immi-gration into the United States for recent immigrantsare primarily economic rather than political6

Why Durham North Carolina First North Car-olina has the fastest growing Latino population in thecountry It experienced an almost 500 increase in itsLatino population primarily with immigrants fromMexico skyrocketing from 76726 in 1990 to 378963in 2000 (US Census Bureau 2002a 2002b) Further-more North Carolina had the highest rate of growthin its immigrant population out of all the states in the 1990s Suro and Singer (2002) identify Raleigh-Durham as having the highest rate of Latino growthfrom 1980 to 2000mdash1180 [See figures in Table 1]Second Durham like other cities of the New Southhas experienced a decline or exit of industries whereblacks have traditionally found work As a result asubstantial portion of the black population now worksin the service industry and many (but not necessarilyall) blacks and the new Latino immigrants find them-selves competing for the same jobs Finally from aresearch and data gathering perspective Durham is ofa manageable size

Data and Measures

The analyses in this paper are based on the 2003Durham Survey of Intergroup Relations (DSIR n =500) The survey was conducted specifically for ourproject by the Center for Survey Research of the Uni-

versity of Virginia using a Computer-Aided TelephoneInterviewing (CATI) system employing random digitdialing (RDD) and dialing of directory-assisted His-panic surname sample7 A randomly generated sampleof phone numbers based on exchanges valid in theDurham North Carolina area was called An over-sample of numbers listed in the phone directory undera Hispanic surname was called at the same time8 Thesurvey was conducted from May 4 through June 222003 and interviews were conducted in both Englishand Spanish (32 of the interviews were conductedin Spanish which translates into 958 of the Latinosample opting to be interviewed in Spanish)9 Araceethnicity quota was implemented to achieve aminimum of 150 whites 150 blacks and 150 Latinosthe remaining 50 respondents were not under thisquota restriction and represent a number ofracialethnic backgrounds Interviews were completedwith 500 residents of the City of Durham for anoverall response rate of 21610 The sample of 500consists of 160 whites (32) 151 blacks (30) 167Latinos (34) 6 Asians (12) 12 who designatedtheir race as Other (24) and 1 respondent (2)who did not indicate a racial category11

5The News and Observer (Raleigh NC) identified that many of theMexican immigrants into North Carolina come primarily fromrural towns in the State of Puebla (November 29 1998 Novem-ber 30 1998) For the most part these immigrants are unskilledand poorly educated

6In a series of articles throughout 2002 chronicling the lives of arearesidents living in poverty The Herald Sun (Durham) provided apicture of life for Latinos in Durham that differed markedly fromthe Miami case Fully 26 of the more than 16000 Latinos inDurham live below the federal poverty level and in order to makea good living it is necessary for them to work more than one job(Assis and Pecquet 2002 A12)

7We recognize the problems associated with drawing a samplefrom a listing of Hispanic surnames for example missing His-panics with non-Hispanic last names and those married to non-Hispanics We also acknowledge that some Latino immigrantsmight not have phones in their homes Given the recency of theHispanic population in Durham and the high proportion ofimmigrants however we choose the sampling frame that wouldgive us the highest probabilities of getting to a Latino respondent

8Both samples were purchased from Survey Sampling Inc (SSI)of Fairfield CT

9We had the questionnaire translated by a Spanish-languageorganization in Chapel Hill NC In order to check the translationand to ensure that it tracked the English-language version thesurvey organization drew a small sample of Latinos in Durhamfor the sole purpose of checking the translation As a resultchanges were made to the translation The revised Spanish-language version was then pretested on another small sample of Latinos in Durham

10A total of 4208 phone numbers were attempted in the course ofthe survey and a total of 14014 call attempts were made TheAmerican Association for Public Opinion Research (AAPOR) ratewas calculated using the full call history of each number that wasrecorded automatically by the CATI software The response ratewas calculated according to AAPOR suggested formula RR3 withe1 = 50 and e2 = 78 We estimated e1 and e2 based on an analy-sis of residency rates and the occurrence of out-of-area householdsin our sample Partial interviews are not counted in the numera-tor of the RR3 formula

11Due to the use of the Hispanic surname sample and racialethnicquotas sampling error is more difficult to calculate The samplemay be viewed as part of two separate populations Within theRDD sample the source of 276 completions the probability of

rsquo

To assess the quality of the sample we comparedthe distribution of demographic characteristics in ourstudy with the characteristics of the population inDurham as reported in the 2000 Census Since thestudy was designed to oversample the black andLatino populations we did a detailed analysis of theextent to which each of our three main subpopulationsamples are represented in our sample the results ofwhich are not presented here12 In brief overall wefind that our sample is reasonably representative ofthe three racial subpopulations under investigationthough the distribution of respondents is slightlyolder better educated and more often female thanfound in the 2000 Census

Our data also confirm what we had suspectedmdashthe Latino population in Durham is basically animmigrant population primarily from MexicoNinety-three percent of the Latino respondents in thesample were born outside the United States Of the93 (n = 156) only about 19 were naturalized cit-izens While Mexicans were the largest portion of theLatino sample (63) Latinos from Central Americawere the next largest group (23) followed by SouthAmerican (5) Puerto Rican (4) Spanish (2)Cuban (1) and Other Latino (2) Given that weare interested in Latino immigrant attitudes most of the analyses are performed using only these 156individuals

We use two dependent variables in our analyses totest our hypotheses One dependent variablemdashLatinostereotypical views of blacksmdashis an index of threestereotype questions ranging in value from 3 (holdingthe least stereotypical views of blacks) to 15 (holdingthe most stereotypical views of blacks)13 A second

dependent variable measures which racial group(Asians blacks or whites) Latinos felt they have themost in common with (1 = have most in commonwith whites 0 = have most in common with blacks)14

We use the following predictor variables although notall are included in every equation education attain-ment measured as number of years of schooling(ranging from 0 to 25 years)15 perception of numberof blacks (Latinos) in their neighborhood16 socialcontact with blacks (Latinos)17 length of time in theUnited States as measured by a question that askedimmigrant respondents the year that they arrived18

(ranges from 0 to 35 years mean = 875 years median= 6 years)19 and gender (1 = male 0 = female) We donot include contact with blacks at work as one of thecontact indicators because of substantial missing dataEducational attainment and gender are used as demo-graphic controls

We also include the indicator of linked fate (thosebelieving that what happens to other blacks (Latinos)will have something to do with their lives)20 for severalreasons The concept of linked fate has been found tobe a powerful predictor of group consciousness polit-ical participation and political attitudes among blackAmericans (Dawson 1994 Matthews and Prothro1967 Shingles 1981 Tate 1993) The concept of linked

selection is known and the margin of error is plusmn59 Within thesurname oversample providing 244 completed interviews allhouseholds listed under a resident with a Hispanic surname wereattempted however Hispanics were included in RDD calling andnon-Hispanics were included in the oversample Non-Hispanicswith Hispanic surnames had a greater chance of selection thannon-Hispanics in the RDD sample who do not have Hispanic sur-names If we assume this to be a more or less random occurrencethen the margin of error for each of the three racialethnic groupsis roughly 8

12A copy of the full analysis is available from the lead author

13Question wording (Q1) For blacks do you think ldquowork hardrdquodescribes almost all blacks most blacks some blacks few blacksalmost no blacks (Q2) For blacks do think ldquoit is easy to get alongwith themrdquo describes (same response categories as question 1)(Q3) For blacks do you think ldquoyou can trust themrdquo describes(same response categories as question 1) We recognize that thesethree questions are the bluntest measures of stereotypes and othermore subtle measures of stereotypes exist but had to make diffi-cult choices when we had to cut the questionnaire in half in orderto stay within budget As such we decided on measuring the mostnegative stereotypes

14Question wording for Latinos (1) Of the following groups if youhad to say which one do you feel you have the most in commonwith African Americansblacks Asians or whites (2) and whichof the following groups do you feel you share the least in commonwith African Americansblacks Asians or whites We recoded theAsian category as missing since only six Latino respondentsreported feeling that Latinos have the most in common withAsians Coding for the dependent variable is 1 (have most incommon with whites) and 0 (have most in common with blacks)

15Income had initially been included in our model but wasdropped due to suspected collinearity with education

16Question wording In your neighborhood how many of the res-idents are black (Latinos) (1) none of the residents (2) a few ofthe residents (3) some of the residents (4) most of the residents(5) all of the residents

17Question wording In your social life including your friends andpeople you know from church and other social activities do youhave (4) a lot of contact (3) some contact (2) little contact or (1)no contact at all with blacks (Latinos)

18We converted this measure to years in the United States by subtracting the year arrived from 2003 (the year the survey wasconducted)

19Although our sample is of Latino immigrants some of whomare new arrivals the mean (875 years) and median (six years) ofthis predicator suggest substantial variance and is thus an ade-quate measure of length of time in country

20Question wording Do you think what happens to Hispanics inthis country will have something to do with what happens in yourlife

fate as Dawson (1994 76) suggests evolves fromsocial identity theory that suggests individuals formtheir concepts of self in part by identifying similari-ties and differences between themselves and othersIndividuals perceive a similarity in experience andtreatment or perceive a linked fate with others (Davisand Brown 2002 241) Only a few studies have exam-ined concepts similar to linked fate for Latinos Kauf-mann (2003) found that Latinos who feel close to oneanother as a group were much more likely to feel closeto black Americans In view of the significance oflinked fate for black Americans and the limited lit-erature suggesting that feelings of closeness to otherLatinos may influence Latino relations with blackAmericans we want to see if Latino linked fate existswithin Latino immigrant populations and if it doeswhat effect does it have on Latino immigrantsrsquo per-ceptions of black Americans The data indicate two-thirds (6791) of Latino immigrants feel a sense oflinked fate with other Latinos This would suggest atleast as reflected in our survey linked fate does existamong Latino immigrants Latino linked fate is meas-ured as 1 = yes and 0 = no

Although we have not specified hypotheses forblacks we examine similar equations for blacks inorder to get a sense of the reciprocal nature of the rela-tionship (n = 151) Not all indicators used for Latinoshowever are appropriate to use for blacks eg lengthof time in country as approximately 95 of blacks arenative-born We use OLS and logit regression analy-ses to test our hypotheses We use Clarify which makesit possible to compute predicted probabilities forvalues of the dependent variable while holding thepredictor variables at their means or some other valueon several of the logit analyses (King Tomz and Wittenberg 2000 Tomz Wittenberg and King 2003)

Testing and Results

Analyses for our first hypothesis indicates that theprevalence of negative stereotypes of black Americansin the Latino immigrant community is quite wide-spread and seem especially so when compared to theprevalence of white stereotypes of blacks Along eachdimension the stereotypes of blacks by Latinos aremore negative than those of white respondents (SeeTable A1 in the JOP online appendix at httpwwwjournalofpoliticsorg) We found that 589 ofthe Latino immigrants in our study reported feelingthat few or almost no blacks are hard workingapproximately one-third (325) of the Latino immi-grant respondents reported feeling that few or almost

no blacks are easy to get along with and slightly morethan a majority (569) of the Latino immigrantrespondents reported feeling that few or almost noblacks could be trusted Among whites the compara-ble figures are only 93 indicate that few or almostno blacks are hard working only 84 believe that fewor almost no blacks are easy to get along with andonly 96 feel that few or almost no blacks could betrusted Consistent with our theoretical expectationsit does appear that many Latinos hold very negativestereotypes of blacks

What factors influence the stereotypes thatLatinos hold of black Americans Table 2 presents theresults of our OLS analysis Even given the relativelysmall number of observations almost all of our pre-dictors are significant and in the hypothesized direc-tion To begin the two demographic control variableseducation and male are both statistically significantMore educated Latinos have significantly less negativestereotypes while men appear to have significantlymore negative stereotypes (perhaps reflecting a pre-disposition of men to use more extreme values onsurvey instruments) More interesting of course arethe influence of the remaining variables in the model

T 2 OLS Regression Results Predicting theExtent to Which Latino ImmigrantsHold Negative Stereotypes of Blacks

Variable Model

Main EffectsEducation minus1033

(0421)Male 7860

(3789)Linked Fate minus7472

(4225)Amount of Social Contact with Blacks minus4780

(1959)Number of Blacks in Neighborhood 2294

(1926)Length of Time in the United States minus0295

(0269)

Constant 12216N 130F Statistic 446R-Squared 1788

Note The dependent variable stereotypical views of blacksranges in value from 3 (holding the least stereotypical views ofblacks) to 15 (holding the most stereotypical views of blacks)Standard error estimates are in parenthesesp lt 01 (two-tailed test)p lt 05 (two-tailed test)p lt 10 (two-tailed test)

rsquo

Jackman and Crane 1986 Kinder and Mendelberg1995 Powers and Ellison 1994 Sigelman and Welch1993 Stephan and Stephan 1985 Welch et al 2001)Other researchers have found that as the concentra-tion and consequently contact of two groupsincreases competition and prejudice also increase(Fossett and Kiecolt 1989 Quillian 1996) Still othershave identified the presence of both increased ani-mosity and reduced prejudice (Morris 1999 Sigelmanand Welch 1993)

Most of this research however has focused onblack and white interaction and few studies havefocused on Latino and Asian interaction with whites(Taylor 1998) Research on the effects of interminor-ity group contact on prejudice and stereotypes islimited Oliver and Wong (2003) examined attitudesof racial groups in multiracial settings finding withthe exception of Asians people who live in neighbor-hoods where their group dominates tend to harborgreater negative stereotypes about other racial minor-ity groups In other words racial stereotypes increaseas the percent of onersquos own ingroup increases in theirneighborhood Blacks and Latinos who are the mostracially isolated harbor the most negative views ofother groups but this pattern was not as pronouncedamong blacks and Latinos residing in neighborhoodsthat are racially diverse Yet they also found thathigher levels of propinquity correlated with lowerlevels of prejudice and perceived competition (Oliverand Wong 2003) Thus it appears that contact theorycan result in three different types of reactions reducedprejudice and perceived competition increased preju-dice and perceived competition or a mixture of both

Drawing from the cases of Miami and Mississippithe literature on Latino immigrant attitudes towardblack Americans and contact theory we test twohypotheses

H1 Latino immigrants hold negative stereotypes of black Americans but these stereotypes will be reduced by length of stay in the countryand increased interaction with black Americansand

H2 Latino immigrants see themselves as closer towhites than to blacks but increases in length oftime in the country and increased interactionwith black Americas will push them closer toblacks

Research Setting

Durham North Carolina is the setting for this studyThe City of Durham like many Southern locations is

undergoing demographic change Table 1 shows thechanges in the demographics of the population of theCity of Durham from 1990 to 2000 What is clear isthat the number of Latinos residing in the city hasrisen dramatically In 1990 Latinos were slightly morethan 1 of the population but by 2000 their per-centage reached 86 For decades whites were themajority in Durham (516 in 1990) but the increas-ing Latino population along with a smaller increasein the Asian population have reduced the white pro-portion to the point where in 2000 blacks and whiteswere almost equal percentages of the population455 for whites and 438 for blacks4

Durhamrsquos Latino population is antithetical to theCuban case First the current wave of Latino immi-grants does not fit into the United Statesrsquo foreignpolicy calculations of encouraging an exodus from a communist country in an effort to undermine thegovernment Thus the enormous amounts of federalaid that went to post-1959 Cubans in South Floridado not exist for these new Latino immigrantsAlthough Latinos in Durham are not benefiting fromtargeted social services as in Miami it does appearthere are special programs designed to help native

T 1 1990 and 2000 Census Data for the Cityof Durham

Change1990 2000 (+-)

Total Population 136611 187035 +

Whites 516 455 minus(70513) (85126)

Blacks 457 438 minus(62393) (81937)

American Indians 2 3 +(358) (575)

Asians 19 36 +(2676) (6815)

Latinos 13 86 +(1713) (16012)

Mexicans 308 642 +(528) (10343)

Puerto Ricans 215 67 minus(369) (696)

Cubans 88 147 +(150) (236)

Other Hispanics 389 296 minus(666) (4737)

4Both of these groups gained in absolute numbers of people butlost as a proportion of the population from 1990 to 2000

Spanish speakers receive more traditional forms ofaid Second Latinos in Durham are typically notmembers of the professional community in theirhome countries and have lower incomes and educa-tion and skill levels than the Cuban community Manyof these immigrants are from poorer countries egMexico and the countries of Central America5 There-fore Latinos and blacks might come into greater com-petition for the same jobs and social services than inMiami (for a discussion of possible job competitionsee McClain et al 2005) Third the reasons for immi-gration into the United States for recent immigrantsare primarily economic rather than political6

Why Durham North Carolina First North Car-olina has the fastest growing Latino population in thecountry It experienced an almost 500 increase in itsLatino population primarily with immigrants fromMexico skyrocketing from 76726 in 1990 to 378963in 2000 (US Census Bureau 2002a 2002b) Further-more North Carolina had the highest rate of growthin its immigrant population out of all the states in the 1990s Suro and Singer (2002) identify Raleigh-Durham as having the highest rate of Latino growthfrom 1980 to 2000mdash1180 [See figures in Table 1]Second Durham like other cities of the New Southhas experienced a decline or exit of industries whereblacks have traditionally found work As a result asubstantial portion of the black population now worksin the service industry and many (but not necessarilyall) blacks and the new Latino immigrants find them-selves competing for the same jobs Finally from aresearch and data gathering perspective Durham is ofa manageable size

Data and Measures

The analyses in this paper are based on the 2003Durham Survey of Intergroup Relations (DSIR n =500) The survey was conducted specifically for ourproject by the Center for Survey Research of the Uni-

versity of Virginia using a Computer-Aided TelephoneInterviewing (CATI) system employing random digitdialing (RDD) and dialing of directory-assisted His-panic surname sample7 A randomly generated sampleof phone numbers based on exchanges valid in theDurham North Carolina area was called An over-sample of numbers listed in the phone directory undera Hispanic surname was called at the same time8 Thesurvey was conducted from May 4 through June 222003 and interviews were conducted in both Englishand Spanish (32 of the interviews were conductedin Spanish which translates into 958 of the Latinosample opting to be interviewed in Spanish)9 Araceethnicity quota was implemented to achieve aminimum of 150 whites 150 blacks and 150 Latinosthe remaining 50 respondents were not under thisquota restriction and represent a number ofracialethnic backgrounds Interviews were completedwith 500 residents of the City of Durham for anoverall response rate of 21610 The sample of 500consists of 160 whites (32) 151 blacks (30) 167Latinos (34) 6 Asians (12) 12 who designatedtheir race as Other (24) and 1 respondent (2)who did not indicate a racial category11

5The News and Observer (Raleigh NC) identified that many of theMexican immigrants into North Carolina come primarily fromrural towns in the State of Puebla (November 29 1998 Novem-ber 30 1998) For the most part these immigrants are unskilledand poorly educated

6In a series of articles throughout 2002 chronicling the lives of arearesidents living in poverty The Herald Sun (Durham) provided apicture of life for Latinos in Durham that differed markedly fromthe Miami case Fully 26 of the more than 16000 Latinos inDurham live below the federal poverty level and in order to makea good living it is necessary for them to work more than one job(Assis and Pecquet 2002 A12)

7We recognize the problems associated with drawing a samplefrom a listing of Hispanic surnames for example missing His-panics with non-Hispanic last names and those married to non-Hispanics We also acknowledge that some Latino immigrantsmight not have phones in their homes Given the recency of theHispanic population in Durham and the high proportion ofimmigrants however we choose the sampling frame that wouldgive us the highest probabilities of getting to a Latino respondent

8Both samples were purchased from Survey Sampling Inc (SSI)of Fairfield CT

9We had the questionnaire translated by a Spanish-languageorganization in Chapel Hill NC In order to check the translationand to ensure that it tracked the English-language version thesurvey organization drew a small sample of Latinos in Durhamfor the sole purpose of checking the translation As a resultchanges were made to the translation The revised Spanish-language version was then pretested on another small sample of Latinos in Durham

10A total of 4208 phone numbers were attempted in the course ofthe survey and a total of 14014 call attempts were made TheAmerican Association for Public Opinion Research (AAPOR) ratewas calculated using the full call history of each number that wasrecorded automatically by the CATI software The response ratewas calculated according to AAPOR suggested formula RR3 withe1 = 50 and e2 = 78 We estimated e1 and e2 based on an analy-sis of residency rates and the occurrence of out-of-area householdsin our sample Partial interviews are not counted in the numera-tor of the RR3 formula

11Due to the use of the Hispanic surname sample and racialethnicquotas sampling error is more difficult to calculate The samplemay be viewed as part of two separate populations Within theRDD sample the source of 276 completions the probability of

rsquo

To assess the quality of the sample we comparedthe distribution of demographic characteristics in ourstudy with the characteristics of the population inDurham as reported in the 2000 Census Since thestudy was designed to oversample the black andLatino populations we did a detailed analysis of theextent to which each of our three main subpopulationsamples are represented in our sample the results ofwhich are not presented here12 In brief overall wefind that our sample is reasonably representative ofthe three racial subpopulations under investigationthough the distribution of respondents is slightlyolder better educated and more often female thanfound in the 2000 Census

Our data also confirm what we had suspectedmdashthe Latino population in Durham is basically animmigrant population primarily from MexicoNinety-three percent of the Latino respondents in thesample were born outside the United States Of the93 (n = 156) only about 19 were naturalized cit-izens While Mexicans were the largest portion of theLatino sample (63) Latinos from Central Americawere the next largest group (23) followed by SouthAmerican (5) Puerto Rican (4) Spanish (2)Cuban (1) and Other Latino (2) Given that weare interested in Latino immigrant attitudes most of the analyses are performed using only these 156individuals

We use two dependent variables in our analyses totest our hypotheses One dependent variablemdashLatinostereotypical views of blacksmdashis an index of threestereotype questions ranging in value from 3 (holdingthe least stereotypical views of blacks) to 15 (holdingthe most stereotypical views of blacks)13 A second

dependent variable measures which racial group(Asians blacks or whites) Latinos felt they have themost in common with (1 = have most in commonwith whites 0 = have most in common with blacks)14

We use the following predictor variables although notall are included in every equation education attain-ment measured as number of years of schooling(ranging from 0 to 25 years)15 perception of numberof blacks (Latinos) in their neighborhood16 socialcontact with blacks (Latinos)17 length of time in theUnited States as measured by a question that askedimmigrant respondents the year that they arrived18

(ranges from 0 to 35 years mean = 875 years median= 6 years)19 and gender (1 = male 0 = female) We donot include contact with blacks at work as one of thecontact indicators because of substantial missing dataEducational attainment and gender are used as demo-graphic controls

We also include the indicator of linked fate (thosebelieving that what happens to other blacks (Latinos)will have something to do with their lives)20 for severalreasons The concept of linked fate has been found tobe a powerful predictor of group consciousness polit-ical participation and political attitudes among blackAmericans (Dawson 1994 Matthews and Prothro1967 Shingles 1981 Tate 1993) The concept of linked

selection is known and the margin of error is plusmn59 Within thesurname oversample providing 244 completed interviews allhouseholds listed under a resident with a Hispanic surname wereattempted however Hispanics were included in RDD calling andnon-Hispanics were included in the oversample Non-Hispanicswith Hispanic surnames had a greater chance of selection thannon-Hispanics in the RDD sample who do not have Hispanic sur-names If we assume this to be a more or less random occurrencethen the margin of error for each of the three racialethnic groupsis roughly 8

12A copy of the full analysis is available from the lead author

13Question wording (Q1) For blacks do you think ldquowork hardrdquodescribes almost all blacks most blacks some blacks few blacksalmost no blacks (Q2) For blacks do think ldquoit is easy to get alongwith themrdquo describes (same response categories as question 1)(Q3) For blacks do you think ldquoyou can trust themrdquo describes(same response categories as question 1) We recognize that thesethree questions are the bluntest measures of stereotypes and othermore subtle measures of stereotypes exist but had to make diffi-cult choices when we had to cut the questionnaire in half in orderto stay within budget As such we decided on measuring the mostnegative stereotypes

14Question wording for Latinos (1) Of the following groups if youhad to say which one do you feel you have the most in commonwith African Americansblacks Asians or whites (2) and whichof the following groups do you feel you share the least in commonwith African Americansblacks Asians or whites We recoded theAsian category as missing since only six Latino respondentsreported feeling that Latinos have the most in common withAsians Coding for the dependent variable is 1 (have most incommon with whites) and 0 (have most in common with blacks)

15Income had initially been included in our model but wasdropped due to suspected collinearity with education

16Question wording In your neighborhood how many of the res-idents are black (Latinos) (1) none of the residents (2) a few ofthe residents (3) some of the residents (4) most of the residents(5) all of the residents

17Question wording In your social life including your friends andpeople you know from church and other social activities do youhave (4) a lot of contact (3) some contact (2) little contact or (1)no contact at all with blacks (Latinos)

18We converted this measure to years in the United States by subtracting the year arrived from 2003 (the year the survey wasconducted)

19Although our sample is of Latino immigrants some of whomare new arrivals the mean (875 years) and median (six years) ofthis predicator suggest substantial variance and is thus an ade-quate measure of length of time in country

20Question wording Do you think what happens to Hispanics inthis country will have something to do with what happens in yourlife

fate as Dawson (1994 76) suggests evolves fromsocial identity theory that suggests individuals formtheir concepts of self in part by identifying similari-ties and differences between themselves and othersIndividuals perceive a similarity in experience andtreatment or perceive a linked fate with others (Davisand Brown 2002 241) Only a few studies have exam-ined concepts similar to linked fate for Latinos Kauf-mann (2003) found that Latinos who feel close to oneanother as a group were much more likely to feel closeto black Americans In view of the significance oflinked fate for black Americans and the limited lit-erature suggesting that feelings of closeness to otherLatinos may influence Latino relations with blackAmericans we want to see if Latino linked fate existswithin Latino immigrant populations and if it doeswhat effect does it have on Latino immigrantsrsquo per-ceptions of black Americans The data indicate two-thirds (6791) of Latino immigrants feel a sense oflinked fate with other Latinos This would suggest atleast as reflected in our survey linked fate does existamong Latino immigrants Latino linked fate is meas-ured as 1 = yes and 0 = no

Although we have not specified hypotheses forblacks we examine similar equations for blacks inorder to get a sense of the reciprocal nature of the rela-tionship (n = 151) Not all indicators used for Latinoshowever are appropriate to use for blacks eg lengthof time in country as approximately 95 of blacks arenative-born We use OLS and logit regression analy-ses to test our hypotheses We use Clarify which makesit possible to compute predicted probabilities forvalues of the dependent variable while holding thepredictor variables at their means or some other valueon several of the logit analyses (King Tomz and Wittenberg 2000 Tomz Wittenberg and King 2003)

Testing and Results

Analyses for our first hypothesis indicates that theprevalence of negative stereotypes of black Americansin the Latino immigrant community is quite wide-spread and seem especially so when compared to theprevalence of white stereotypes of blacks Along eachdimension the stereotypes of blacks by Latinos aremore negative than those of white respondents (SeeTable A1 in the JOP online appendix at httpwwwjournalofpoliticsorg) We found that 589 ofthe Latino immigrants in our study reported feelingthat few or almost no blacks are hard workingapproximately one-third (325) of the Latino immi-grant respondents reported feeling that few or almost

no blacks are easy to get along with and slightly morethan a majority (569) of the Latino immigrantrespondents reported feeling that few or almost noblacks could be trusted Among whites the compara-ble figures are only 93 indicate that few or almostno blacks are hard working only 84 believe that fewor almost no blacks are easy to get along with andonly 96 feel that few or almost no blacks could betrusted Consistent with our theoretical expectationsit does appear that many Latinos hold very negativestereotypes of blacks

What factors influence the stereotypes thatLatinos hold of black Americans Table 2 presents theresults of our OLS analysis Even given the relativelysmall number of observations almost all of our pre-dictors are significant and in the hypothesized direc-tion To begin the two demographic control variableseducation and male are both statistically significantMore educated Latinos have significantly less negativestereotypes while men appear to have significantlymore negative stereotypes (perhaps reflecting a pre-disposition of men to use more extreme values onsurvey instruments) More interesting of course arethe influence of the remaining variables in the model

T 2 OLS Regression Results Predicting theExtent to Which Latino ImmigrantsHold Negative Stereotypes of Blacks

Variable Model

Main EffectsEducation minus1033

(0421)Male 7860

(3789)Linked Fate minus7472

(4225)Amount of Social Contact with Blacks minus4780

(1959)Number of Blacks in Neighborhood 2294

(1926)Length of Time in the United States minus0295

(0269)

Constant 12216N 130F Statistic 446R-Squared 1788

Note The dependent variable stereotypical views of blacksranges in value from 3 (holding the least stereotypical views ofblacks) to 15 (holding the most stereotypical views of blacks)Standard error estimates are in parenthesesp lt 01 (two-tailed test)p lt 05 (two-tailed test)p lt 10 (two-tailed test)

Spanish speakers receive more traditional forms ofaid Second Latinos in Durham are typically notmembers of the professional community in theirhome countries and have lower incomes and educa-tion and skill levels than the Cuban community Manyof these immigrants are from poorer countries egMexico and the countries of Central America5 There-fore Latinos and blacks might come into greater com-petition for the same jobs and social services than inMiami (for a discussion of possible job competitionsee McClain et al 2005) Third the reasons for immi-gration into the United States for recent immigrantsare primarily economic rather than political6

Why Durham North Carolina First North Car-olina has the fastest growing Latino population in thecountry It experienced an almost 500 increase in itsLatino population primarily with immigrants fromMexico skyrocketing from 76726 in 1990 to 378963in 2000 (US Census Bureau 2002a 2002b) Further-more North Carolina had the highest rate of growthin its immigrant population out of all the states in the 1990s Suro and Singer (2002) identify Raleigh-Durham as having the highest rate of Latino growthfrom 1980 to 2000mdash1180 [See figures in Table 1]Second Durham like other cities of the New Southhas experienced a decline or exit of industries whereblacks have traditionally found work As a result asubstantial portion of the black population now worksin the service industry and many (but not necessarilyall) blacks and the new Latino immigrants find them-selves competing for the same jobs Finally from aresearch and data gathering perspective Durham is ofa manageable size

Data and Measures

The analyses in this paper are based on the 2003Durham Survey of Intergroup Relations (DSIR n =500) The survey was conducted specifically for ourproject by the Center for Survey Research of the Uni-

versity of Virginia using a Computer-Aided TelephoneInterviewing (CATI) system employing random digitdialing (RDD) and dialing of directory-assisted His-panic surname sample7 A randomly generated sampleof phone numbers based on exchanges valid in theDurham North Carolina area was called An over-sample of numbers listed in the phone directory undera Hispanic surname was called at the same time8 Thesurvey was conducted from May 4 through June 222003 and interviews were conducted in both Englishand Spanish (32 of the interviews were conductedin Spanish which translates into 958 of the Latinosample opting to be interviewed in Spanish)9 Araceethnicity quota was implemented to achieve aminimum of 150 whites 150 blacks and 150 Latinosthe remaining 50 respondents were not under thisquota restriction and represent a number ofracialethnic backgrounds Interviews were completedwith 500 residents of the City of Durham for anoverall response rate of 21610 The sample of 500consists of 160 whites (32) 151 blacks (30) 167Latinos (34) 6 Asians (12) 12 who designatedtheir race as Other (24) and 1 respondent (2)who did not indicate a racial category11

5The News and Observer (Raleigh NC) identified that many of theMexican immigrants into North Carolina come primarily fromrural towns in the State of Puebla (November 29 1998 Novem-ber 30 1998) For the most part these immigrants are unskilledand poorly educated

6In a series of articles throughout 2002 chronicling the lives of arearesidents living in poverty The Herald Sun (Durham) provided apicture of life for Latinos in Durham that differed markedly fromthe Miami case Fully 26 of the more than 16000 Latinos inDurham live below the federal poverty level and in order to makea good living it is necessary for them to work more than one job(Assis and Pecquet 2002 A12)

7We recognize the problems associated with drawing a samplefrom a listing of Hispanic surnames for example missing His-panics with non-Hispanic last names and those married to non-Hispanics We also acknowledge that some Latino immigrantsmight not have phones in their homes Given the recency of theHispanic population in Durham and the high proportion ofimmigrants however we choose the sampling frame that wouldgive us the highest probabilities of getting to a Latino respondent

8Both samples were purchased from Survey Sampling Inc (SSI)of Fairfield CT

9We had the questionnaire translated by a Spanish-languageorganization in Chapel Hill NC In order to check the translationand to ensure that it tracked the English-language version thesurvey organization drew a small sample of Latinos in Durhamfor the sole purpose of checking the translation As a resultchanges were made to the translation The revised Spanish-language version was then pretested on another small sample of Latinos in Durham

10A total of 4208 phone numbers were attempted in the course ofthe survey and a total of 14014 call attempts were made TheAmerican Association for Public Opinion Research (AAPOR) ratewas calculated using the full call history of each number that wasrecorded automatically by the CATI software The response ratewas calculated according to AAPOR suggested formula RR3 withe1 = 50 and e2 = 78 We estimated e1 and e2 based on an analy-sis of residency rates and the occurrence of out-of-area householdsin our sample Partial interviews are not counted in the numera-tor of the RR3 formula

11Due to the use of the Hispanic surname sample and racialethnicquotas sampling error is more difficult to calculate The samplemay be viewed as part of two separate populations Within theRDD sample the source of 276 completions the probability of

rsquo

To assess the quality of the sample we comparedthe distribution of demographic characteristics in ourstudy with the characteristics of the population inDurham as reported in the 2000 Census Since thestudy was designed to oversample the black andLatino populations we did a detailed analysis of theextent to which each of our three main subpopulationsamples are represented in our sample the results ofwhich are not presented here12 In brief overall wefind that our sample is reasonably representative ofthe three racial subpopulations under investigationthough the distribution of respondents is slightlyolder better educated and more often female thanfound in the 2000 Census

Our data also confirm what we had suspectedmdashthe Latino population in Durham is basically animmigrant population primarily from MexicoNinety-three percent of the Latino respondents in thesample were born outside the United States Of the93 (n = 156) only about 19 were naturalized cit-izens While Mexicans were the largest portion of theLatino sample (63) Latinos from Central Americawere the next largest group (23) followed by SouthAmerican (5) Puerto Rican (4) Spanish (2)Cuban (1) and Other Latino (2) Given that weare interested in Latino immigrant attitudes most of the analyses are performed using only these 156individuals

We use two dependent variables in our analyses totest our hypotheses One dependent variablemdashLatinostereotypical views of blacksmdashis an index of threestereotype questions ranging in value from 3 (holdingthe least stereotypical views of blacks) to 15 (holdingthe most stereotypical views of blacks)13 A second

dependent variable measures which racial group(Asians blacks or whites) Latinos felt they have themost in common with (1 = have most in commonwith whites 0 = have most in common with blacks)14

We use the following predictor variables although notall are included in every equation education attain-ment measured as number of years of schooling(ranging from 0 to 25 years)15 perception of numberof blacks (Latinos) in their neighborhood16 socialcontact with blacks (Latinos)17 length of time in theUnited States as measured by a question that askedimmigrant respondents the year that they arrived18

(ranges from 0 to 35 years mean = 875 years median= 6 years)19 and gender (1 = male 0 = female) We donot include contact with blacks at work as one of thecontact indicators because of substantial missing dataEducational attainment and gender are used as demo-graphic controls

We also include the indicator of linked fate (thosebelieving that what happens to other blacks (Latinos)will have something to do with their lives)20 for severalreasons The concept of linked fate has been found tobe a powerful predictor of group consciousness polit-ical participation and political attitudes among blackAmericans (Dawson 1994 Matthews and Prothro1967 Shingles 1981 Tate 1993) The concept of linked

selection is known and the margin of error is plusmn59 Within thesurname oversample providing 244 completed interviews allhouseholds listed under a resident with a Hispanic surname wereattempted however Hispanics were included in RDD calling andnon-Hispanics were included in the oversample Non-Hispanicswith Hispanic surnames had a greater chance of selection thannon-Hispanics in the RDD sample who do not have Hispanic sur-names If we assume this to be a more or less random occurrencethen the margin of error for each of the three racialethnic groupsis roughly 8

12A copy of the full analysis is available from the lead author

13Question wording (Q1) For blacks do you think ldquowork hardrdquodescribes almost all blacks most blacks some blacks few blacksalmost no blacks (Q2) For blacks do think ldquoit is easy to get alongwith themrdquo describes (same response categories as question 1)(Q3) For blacks do you think ldquoyou can trust themrdquo describes(same response categories as question 1) We recognize that thesethree questions are the bluntest measures of stereotypes and othermore subtle measures of stereotypes exist but had to make diffi-cult choices when we had to cut the questionnaire in half in orderto stay within budget As such we decided on measuring the mostnegative stereotypes

14Question wording for Latinos (1) Of the following groups if youhad to say which one do you feel you have the most in commonwith African Americansblacks Asians or whites (2) and whichof the following groups do you feel you share the least in commonwith African Americansblacks Asians or whites We recoded theAsian category as missing since only six Latino respondentsreported feeling that Latinos have the most in common withAsians Coding for the dependent variable is 1 (have most incommon with whites) and 0 (have most in common with blacks)

15Income had initially been included in our model but wasdropped due to suspected collinearity with education

16Question wording In your neighborhood how many of the res-idents are black (Latinos) (1) none of the residents (2) a few ofthe residents (3) some of the residents (4) most of the residents(5) all of the residents

17Question wording In your social life including your friends andpeople you know from church and other social activities do youhave (4) a lot of contact (3) some contact (2) little contact or (1)no contact at all with blacks (Latinos)

18We converted this measure to years in the United States by subtracting the year arrived from 2003 (the year the survey wasconducted)

19Although our sample is of Latino immigrants some of whomare new arrivals the mean (875 years) and median (six years) ofthis predicator suggest substantial variance and is thus an ade-quate measure of length of time in country

20Question wording Do you think what happens to Hispanics inthis country will have something to do with what happens in yourlife

fate as Dawson (1994 76) suggests evolves fromsocial identity theory that suggests individuals formtheir concepts of self in part by identifying similari-ties and differences between themselves and othersIndividuals perceive a similarity in experience andtreatment or perceive a linked fate with others (Davisand Brown 2002 241) Only a few studies have exam-ined concepts similar to linked fate for Latinos Kauf-mann (2003) found that Latinos who feel close to oneanother as a group were much more likely to feel closeto black Americans In view of the significance oflinked fate for black Americans and the limited lit-erature suggesting that feelings of closeness to otherLatinos may influence Latino relations with blackAmericans we want to see if Latino linked fate existswithin Latino immigrant populations and if it doeswhat effect does it have on Latino immigrantsrsquo per-ceptions of black Americans The data indicate two-thirds (6791) of Latino immigrants feel a sense oflinked fate with other Latinos This would suggest atleast as reflected in our survey linked fate does existamong Latino immigrants Latino linked fate is meas-ured as 1 = yes and 0 = no

Although we have not specified hypotheses forblacks we examine similar equations for blacks inorder to get a sense of the reciprocal nature of the rela-tionship (n = 151) Not all indicators used for Latinoshowever are appropriate to use for blacks eg lengthof time in country as approximately 95 of blacks arenative-born We use OLS and logit regression analy-ses to test our hypotheses We use Clarify which makesit possible to compute predicted probabilities forvalues of the dependent variable while holding thepredictor variables at their means or some other valueon several of the logit analyses (King Tomz and Wittenberg 2000 Tomz Wittenberg and King 2003)

Testing and Results

Analyses for our first hypothesis indicates that theprevalence of negative stereotypes of black Americansin the Latino immigrant community is quite wide-spread and seem especially so when compared to theprevalence of white stereotypes of blacks Along eachdimension the stereotypes of blacks by Latinos aremore negative than those of white respondents (SeeTable A1 in the JOP online appendix at httpwwwjournalofpoliticsorg) We found that 589 ofthe Latino immigrants in our study reported feelingthat few or almost no blacks are hard workingapproximately one-third (325) of the Latino immi-grant respondents reported feeling that few or almost

no blacks are easy to get along with and slightly morethan a majority (569) of the Latino immigrantrespondents reported feeling that few or almost noblacks could be trusted Among whites the compara-ble figures are only 93 indicate that few or almostno blacks are hard working only 84 believe that fewor almost no blacks are easy to get along with andonly 96 feel that few or almost no blacks could betrusted Consistent with our theoretical expectationsit does appear that many Latinos hold very negativestereotypes of blacks

What factors influence the stereotypes thatLatinos hold of black Americans Table 2 presents theresults of our OLS analysis Even given the relativelysmall number of observations almost all of our pre-dictors are significant and in the hypothesized direc-tion To begin the two demographic control variableseducation and male are both statistically significantMore educated Latinos have significantly less negativestereotypes while men appear to have significantlymore negative stereotypes (perhaps reflecting a pre-disposition of men to use more extreme values onsurvey instruments) More interesting of course arethe influence of the remaining variables in the model

T 2 OLS Regression Results Predicting theExtent to Which Latino ImmigrantsHold Negative Stereotypes of Blacks

Variable Model

Main EffectsEducation minus1033

(0421)Male 7860

(3789)Linked Fate minus7472

(4225)Amount of Social Contact with Blacks minus4780

(1959)Number of Blacks in Neighborhood 2294

(1926)Length of Time in the United States minus0295

(0269)

Constant 12216N 130F Statistic 446R-Squared 1788

Note The dependent variable stereotypical views of blacksranges in value from 3 (holding the least stereotypical views ofblacks) to 15 (holding the most stereotypical views of blacks)Standard error estimates are in parenthesesp lt 01 (two-tailed test)p lt 05 (two-tailed test)p lt 10 (two-tailed test)

rsquo

To assess the quality of the sample we comparedthe distribution of demographic characteristics in ourstudy with the characteristics of the population inDurham as reported in the 2000 Census Since thestudy was designed to oversample the black andLatino populations we did a detailed analysis of theextent to which each of our three main subpopulationsamples are represented in our sample the results ofwhich are not presented here12 In brief overall wefind that our sample is reasonably representative ofthe three racial subpopulations under investigationthough the distribution of respondents is slightlyolder better educated and more often female thanfound in the 2000 Census

Our data also confirm what we had suspectedmdashthe Latino population in Durham is basically animmigrant population primarily from MexicoNinety-three percent of the Latino respondents in thesample were born outside the United States Of the93 (n = 156) only about 19 were naturalized cit-izens While Mexicans were the largest portion of theLatino sample (63) Latinos from Central Americawere the next largest group (23) followed by SouthAmerican (5) Puerto Rican (4) Spanish (2)Cuban (1) and Other Latino (2) Given that weare interested in Latino immigrant attitudes most of the analyses are performed using only these 156individuals

We use two dependent variables in our analyses totest our hypotheses One dependent variablemdashLatinostereotypical views of blacksmdashis an index of threestereotype questions ranging in value from 3 (holdingthe least stereotypical views of blacks) to 15 (holdingthe most stereotypical views of blacks)13 A second

dependent variable measures which racial group(Asians blacks or whites) Latinos felt they have themost in common with (1 = have most in commonwith whites 0 = have most in common with blacks)14

We use the following predictor variables although notall are included in every equation education attain-ment measured as number of years of schooling(ranging from 0 to 25 years)15 perception of numberof blacks (Latinos) in their neighborhood16 socialcontact with blacks (Latinos)17 length of time in theUnited States as measured by a question that askedimmigrant respondents the year that they arrived18

(ranges from 0 to 35 years mean = 875 years median= 6 years)19 and gender (1 = male 0 = female) We donot include contact with blacks at work as one of thecontact indicators because of substantial missing dataEducational attainment and gender are used as demo-graphic controls

We also include the indicator of linked fate (thosebelieving that what happens to other blacks (Latinos)will have something to do with their lives)20 for severalreasons The concept of linked fate has been found tobe a powerful predictor of group consciousness polit-ical participation and political attitudes among blackAmericans (Dawson 1994 Matthews and Prothro1967 Shingles 1981 Tate 1993) The concept of linked

selection is known and the margin of error is plusmn59 Within thesurname oversample providing 244 completed interviews allhouseholds listed under a resident with a Hispanic surname wereattempted however Hispanics were included in RDD calling andnon-Hispanics were included in the oversample Non-Hispanicswith Hispanic surnames had a greater chance of selection thannon-Hispanics in the RDD sample who do not have Hispanic sur-names If we assume this to be a more or less random occurrencethen the margin of error for each of the three racialethnic groupsis roughly 8

12A copy of the full analysis is available from the lead author

13Question wording (Q1) For blacks do you think ldquowork hardrdquodescribes almost all blacks most blacks some blacks few blacksalmost no blacks (Q2) For blacks do think ldquoit is easy to get alongwith themrdquo describes (same response categories as question 1)(Q3) For blacks do you think ldquoyou can trust themrdquo describes(same response categories as question 1) We recognize that thesethree questions are the bluntest measures of stereotypes and othermore subtle measures of stereotypes exist but had to make diffi-cult choices when we had to cut the questionnaire in half in orderto stay within budget As such we decided on measuring the mostnegative stereotypes

14Question wording for Latinos (1) Of the following groups if youhad to say which one do you feel you have the most in commonwith African Americansblacks Asians or whites (2) and whichof the following groups do you feel you share the least in commonwith African Americansblacks Asians or whites We recoded theAsian category as missing since only six Latino respondentsreported feeling that Latinos have the most in common withAsians Coding for the dependent variable is 1 (have most incommon with whites) and 0 (have most in common with blacks)

15Income had initially been included in our model but wasdropped due to suspected collinearity with education

16Question wording In your neighborhood how many of the res-idents are black (Latinos) (1) none of the residents (2) a few ofthe residents (3) some of the residents (4) most of the residents(5) all of the residents

17Question wording In your social life including your friends andpeople you know from church and other social activities do youhave (4) a lot of contact (3) some contact (2) little contact or (1)no contact at all with blacks (Latinos)

18We converted this measure to years in the United States by subtracting the year arrived from 2003 (the year the survey wasconducted)

19Although our sample is of Latino immigrants some of whomare new arrivals the mean (875 years) and median (six years) ofthis predicator suggest substantial variance and is thus an ade-quate measure of length of time in country

20Question wording Do you think what happens to Hispanics inthis country will have something to do with what happens in yourlife

fate as Dawson (1994 76) suggests evolves fromsocial identity theory that suggests individuals formtheir concepts of self in part by identifying similari-ties and differences between themselves and othersIndividuals perceive a similarity in experience andtreatment or perceive a linked fate with others (Davisand Brown 2002 241) Only a few studies have exam-ined concepts similar to linked fate for Latinos Kauf-mann (2003) found that Latinos who feel close to oneanother as a group were much more likely to feel closeto black Americans In view of the significance oflinked fate for black Americans and the limited lit-erature suggesting that feelings of closeness to otherLatinos may influence Latino relations with blackAmericans we want to see if Latino linked fate existswithin Latino immigrant populations and if it doeswhat effect does it have on Latino immigrantsrsquo per-ceptions of black Americans The data indicate two-thirds (6791) of Latino immigrants feel a sense oflinked fate with other Latinos This would suggest atleast as reflected in our survey linked fate does existamong Latino immigrants Latino linked fate is meas-ured as 1 = yes and 0 = no

Although we have not specified hypotheses forblacks we examine similar equations for blacks inorder to get a sense of the reciprocal nature of the rela-tionship (n = 151) Not all indicators used for Latinoshowever are appropriate to use for blacks eg lengthof time in country as approximately 95 of blacks arenative-born We use OLS and logit regression analy-ses to test our hypotheses We use Clarify which makesit possible to compute predicted probabilities forvalues of the dependent variable while holding thepredictor variables at their means or some other valueon several of the logit analyses (King Tomz and Wittenberg 2000 Tomz Wittenberg and King 2003)

Testing and Results

Analyses for our first hypothesis indicates that theprevalence of negative stereotypes of black Americansin the Latino immigrant community is quite wide-spread and seem especially so when compared to theprevalence of white stereotypes of blacks Along eachdimension the stereotypes of blacks by Latinos aremore negative than those of white respondents (SeeTable A1 in the JOP online appendix at httpwwwjournalofpoliticsorg) We found that 589 ofthe Latino immigrants in our study reported feelingthat few or almost no blacks are hard workingapproximately one-third (325) of the Latino immi-grant respondents reported feeling that few or almost

no blacks are easy to get along with and slightly morethan a majority (569) of the Latino immigrantrespondents reported feeling that few or almost noblacks could be trusted Among whites the compara-ble figures are only 93 indicate that few or almostno blacks are hard working only 84 believe that fewor almost no blacks are easy to get along with andonly 96 feel that few or almost no blacks could betrusted Consistent with our theoretical expectationsit does appear that many Latinos hold very negativestereotypes of blacks

What factors influence the stereotypes thatLatinos hold of black Americans Table 2 presents theresults of our OLS analysis Even given the relativelysmall number of observations almost all of our pre-dictors are significant and in the hypothesized direc-tion To begin the two demographic control variableseducation and male are both statistically significantMore educated Latinos have significantly less negativestereotypes while men appear to have significantlymore negative stereotypes (perhaps reflecting a pre-disposition of men to use more extreme values onsurvey instruments) More interesting of course arethe influence of the remaining variables in the model

T 2 OLS Regression Results Predicting theExtent to Which Latino ImmigrantsHold Negative Stereotypes of Blacks

Variable Model

Main EffectsEducation minus1033

(0421)Male 7860

(3789)Linked Fate minus7472

(4225)Amount of Social Contact with Blacks minus4780

(1959)Number of Blacks in Neighborhood 2294

(1926)Length of Time in the United States minus0295

(0269)

Constant 12216N 130F Statistic 446R-Squared 1788

Note The dependent variable stereotypical views of blacksranges in value from 3 (holding the least stereotypical views ofblacks) to 15 (holding the most stereotypical views of blacks)Standard error estimates are in parenthesesp lt 01 (two-tailed test)p lt 05 (two-tailed test)p lt 10 (two-tailed test)

fate as Dawson (1994 76) suggests evolves fromsocial identity theory that suggests individuals formtheir concepts of self in part by identifying similari-ties and differences between themselves and othersIndividuals perceive a similarity in experience andtreatment or perceive a linked fate with others (Davisand Brown 2002 241) Only a few studies have exam-ined concepts similar to linked fate for Latinos Kauf-mann (2003) found that Latinos who feel close to oneanother as a group were much more likely to feel closeto black Americans In view of the significance oflinked fate for black Americans and the limited lit-erature suggesting that feelings of closeness to otherLatinos may influence Latino relations with blackAmericans we want to see if Latino linked fate existswithin Latino immigrant populations and if it doeswhat effect does it have on Latino immigrantsrsquo per-ceptions of black Americans The data indicate two-thirds (6791) of Latino immigrants feel a sense oflinked fate with other Latinos This would suggest atleast as reflected in our survey linked fate does existamong Latino immigrants Latino linked fate is meas-ured as 1 = yes and 0 = no

Although we have not specified hypotheses forblacks we examine similar equations for blacks inorder to get a sense of the reciprocal nature of the rela-tionship (n = 151) Not all indicators used for Latinoshowever are appropriate to use for blacks eg lengthof time in country as approximately 95 of blacks arenative-born We use OLS and logit regression analy-ses to test our hypotheses We use Clarify which makesit possible to compute predicted probabilities forvalues of the dependent variable while holding thepredictor variables at their means or some other valueon several of the logit analyses (King Tomz and Wittenberg 2000 Tomz Wittenberg and King 2003)

Testing and Results

Analyses for our first hypothesis indicates that theprevalence of negative stereotypes of black Americansin the Latino immigrant community is quite wide-spread and seem especially so when compared to theprevalence of white stereotypes of blacks Along eachdimension the stereotypes of blacks by Latinos aremore negative than those of white respondents (SeeTable A1 in the JOP online appendix at httpwwwjournalofpoliticsorg) We found that 589 ofthe Latino immigrants in our study reported feelingthat few or almost no blacks are hard workingapproximately one-third (325) of the Latino immi-grant respondents reported feeling that few or almost

no blacks are easy to get along with and slightly morethan a majority (569) of the Latino immigrantrespondents reported feeling that few or almost noblacks could be trusted Among whites the compara-ble figures are only 93 indicate that few or almostno blacks are hard working only 84 believe that fewor almost no blacks are easy to get along with andonly 96 feel that few or almost no blacks could betrusted Consistent with our theoretical expectationsit does appear that many Latinos hold very negativestereotypes of blacks

What factors influence the stereotypes thatLatinos hold of black Americans Table 2 presents theresults of our OLS analysis Even given the relativelysmall number of observations almost all of our pre-dictors are significant and in the hypothesized direc-tion To begin the two demographic control variableseducation and male are both statistically significantMore educated Latinos have significantly less negativestereotypes while men appear to have significantlymore negative stereotypes (perhaps reflecting a pre-disposition of men to use more extreme values onsurvey instruments) More interesting of course arethe influence of the remaining variables in the model

T 2 OLS Regression Results Predicting theExtent to Which Latino ImmigrantsHold Negative Stereotypes of Blacks

Variable Model

Main EffectsEducation minus1033

(0421)Male 7860

(3789)Linked Fate minus7472

(4225)Amount of Social Contact with Blacks minus4780

(1959)Number of Blacks in Neighborhood 2294

(1926)Length of Time in the United States minus0295

(0269)

Constant 12216N 130F Statistic 446R-Squared 1788

Note The dependent variable stereotypical views of blacksranges in value from 3 (holding the least stereotypical views ofblacks) to 15 (holding the most stereotypical views of blacks)Standard error estimates are in parenthesesp lt 01 (two-tailed test)p lt 05 (two-tailed test)p lt 10 (two-tailed test)

rsquo

Linked Fate has a statistically significant and negativeeffect on unfavorable stereotypes of black Americanswhich suggests that a sense of group identity withother Latinos is a robust corrective to negative stereo-types of blacks Similarly length of time in the UnitedStates has a negative but in this case insignificanteffect on stereotypes of blacks Consistent with earlierstudies on the effects of contact among groups ourresults are mixed The amount of social contact alsohas a significant negative influence on Latinosrsquo stereo-types of blacks meaning more social contact predictsmore positive attitudes toward blacks yet contrary toour expectations the number of blacks in the neigh-borhood has a positive but insignificant coefficient21

(These two indicators are not collinear in the zero-order case) The explanation perhaps is that con-trolling for the ameliorating effects on negativestereotypes of social contact having black neighborsmay simply become a better marker for competitionover housing jobs and other resources which couldexplain the greater prejudice of Latinos living inneighborhoods with more blacks

One might think that the cause of Latinosrsquo nega-tive opinions about blacks is the transmission of prej-udice from Southern whites but our data do notsupport this notion If you recall white respondentsreport less negative stereotypes of blacks than doLatinos To the extent that these responses are honestlycommunicated then it seems unlikely that Latinoprejudices are transmitted by whites As we knowmany racial attitudes are subject to social desirabilityeffects so the values of Southern whites reflected inour survey may not be an accurate reflection of whatpeople really think While not able to dismiss thisnotion entirely it seems peculiar to us that whiteswould transmit negative stereotypes of blacks to otherminorities in face-to-face interactions especiallywhen they would not give these opinions over thephone which might indicate their discomfort atgiving voice to these stereotypes Second a necessarycondition for whites to communicate negative stereo-types to Latinos would seem to be that the longerLatinos were in the United States interacting with thecountryrsquos white population the more prejudiced theywould have to become We find in our multivariateanalysis that Latinos who spend more time in theUnited States have less negative stereotypes of blacks(although the small sample size prevents us from

claiming statistical significance) which would beinconsistent with a model based on the communica-tion of values from Southern whites

Alternatively Latino prejudice toward blacks maybe reciprocated prejudice That is perhaps Latinosrsquounfavorable opinions of blacks are to compensate insome way for prejudicial attitudes of blacks towardLatinos Our data show that there is little support forthis conjecture (See bottom of Table A1 in the JOPonline appendix) The overwhelming impression con-veyed in the data is that blacks view Latinos muchmore favorably than Latinos view blacks Almostthree-fourths (719) of blacks feel most or almost allLatinos are hardworking two-fifths (428) believemost or almost all Latinos are easy to get along withand only one-third (326) indicate almost no or fewLatinos could be trusted On the other hand only92 of Latinos feel that most or almost all blacks arehard working only 265 feel that most or almost allblacks are easy to get along with and only 8 believethat most or almost all blacks can be trusted It doesnot appear therefore that Latino prejudices are nec-essarily a result of hostility from Southern blacks

Although blacks hold fewer stereotypes ofLatinos stereotypes do exist as evidenced by the per-centages presented in the previous paragraph TheOLS results for the factors contributing to blackstereotypes of Latinos are similar in ways to those forLatinos with several distinct and interesting differ-ences (The results are not shown here but can befound in Table A2 in the JOP online appendix) Blackswith higher levels of education hold fewer stereotypi-cal views of Latinos and unlike Latino males blackmen hold fewer stereotypes of Latinos than do blackwomen Although linked fate and social contact withLatinos are not statistically significant they are bothin the direction that we would have hypothesizedBlacks with linked fate and social contact with Latinosappear to hold fewer stereotypes of Latinos Interest-ingly just as in the Latino model blacks with moreneighborhood contact with Latinos appear to holdmore negative stereotypes of Latinos than blacks withless neighborhood contact This coefficient howeveris not statistically significant

For our second hypothesis we began by consider-ing which racial groupmdashAsians blacks or whitesmdashLatino immigrants in our study reported feeling theyhave the most in common with and the least incommon with The data reveal that the overwhelmingmajority of our Latino immigrant respondents(783) feel that they have the most in common withwhites and the least in common with blacks (528)Curiously the relatively warm feelings toward whites

21We reran the models including interaction terms of linked fatewith neighborhood interaction with blacks and education withneighborhood interaction with blacks individually and in combi-nation None of the interaction terms were statistically significant

and the coolness toward blacks among Latino respon-dents are not mutual (See Table A3 in the JOP onlineappendix) The data show that while 459 of whiterespondents see themselves as having the most incommon with blacks just 222 of whites see them-selves as having the most in common with LatinosSimilarly a plurality of whites (475) sees themselvesas having the least in common with Latinos Blackrespondents on the other hand do not generally feel as distant from Latinos as Latinos feel from blacks About half of the blacks in the sample reporthaving the most in common with Latinos (496)while an almost similar proportion (455) reporthaving the most in common with whites Just 227of blacks report having the least in common withLatinos22

To test our expectation that increases in length oftime in the United States and interaction with blackAmericans move Latino immigrants closer to blackswe used a logit model with our dependent variablebeing a measure of which group blacks or whitesLatino immigrants feel they have the most in commonwith In this instance we found including both contactmeasures simultaneously social contact and blacks inthe neighborhood resulted in both being statisticallyinsignificant but since the number of blacks in therespondentrsquos neighborhood was closer to significancethan social contact we decided to include it in ourmodel In addition to our primary independent vari-ables we also added a measure of the extent to whichour Latino immigrant respondents hold stereotypicalviews of black Americans We used the dependentvariable of the stereotype index as a predictor in thisanalysis The Latino linked fate variable is alsoincluded as a predictor

Our results for this hypothesis are shown in Table3 Our analyses do not reveal a statistically significantrelationship between educational attainment andlength of time in the United States with closeness to aparticular racial group The coefficients however onboth variables are in the predicted direction thus sug-gesting the relationship that we hypothesized but notstatistically significant The coefficient on the numberof blacks in a respondentrsquos neighborhood seems to reveala relationship that is opposite of what we hypothe-sizedmdashthe greater the number of blacks in the neigh-borhood the more likely Latino immigrants are to feelthey have more in common with whitesmdashbut it is also

statistically insignificant23 The most statistically sig-nificant variables in our analysis are the stereotypeindex and linked fate with the results revealing thatthose Latino immigrants with linked fate and with lessstereotypical views of black Americans are more likelyto feel they have the most in common with blacks andleast likely to believe they have the most in commonwith whites Once again a sense of group identityamong Latino immigrants contributes significantly toa sense of connection with black Americans

We compute percentage changes in probability aswe move from one value on our independent variablesto another Substantively therefore an average Latinoimmigrant (ie with the mean levels of educationlength of time in the United States neighborhoodinteraction with black Americans and perception ofblack Americans on the stereotype index) with linkedfate is approximately 19 more likely than an averageLatino without linked fate to feel that she has the mostin common with black Americans Though the rela-

22Given the small numbers of Asians in the Durham area (andnationally) it is interesting to note that between 30 and 40 ofall races reported having the least in common and 32 of whitesreported themselves as having the most in common with Asians

T 3 Logit Results for Which Group LatinoRespondents Report Having the Most inCommon With

Variables Coefficients

Education minus0735(0576)

Linked Fate minus2005(8703)

Length of Time in the United States minus0301(0393)

Number of Blacks in Neighborhood 2897(2826)

Stereotypical Views of Blacks 2293(Higher values indicate holding more (1269)stereotypical views of blacks)

Constant 9166N 102Chi-Squared 1421Pseudo R2 1407

Note The dependent variable for this model is coded such that 1indicates having the most in common with whites and 0 indicateshaving the most in common with blacks Standard error estimatesare in parenthesesp lt 01 (two-tailed test)p lt 05 (two-tailed test)p lt 10 (two-tailed test)

23As with the first hypothesis we included interaction terms oflinked fate with neighborhood contact with blacks and educationwith neighborhood contact with blacks individually and in combination None of the interaction terms was statistically significant

rsquo

tionship between perceptions of black Americans onthe stereotype index and closeness to whites or blacksis statistically significant the substantive effect isminuscule Latino immigrants with highly stereotyp-ical views of black Americans are less than 1 (22)more likely than other Latino immigrants to feel theyhave the most in common with whites24

Discussion and Conclusion

We posited two hypothesesmdashLatino immigrants holdnegative stereotypes of blacks and Latino immigrantsfeel they have the most in common with whites bothattitudes that may be mitigated by contact factors andlength of time in the countrymdashas the core questionsfor this paper Some aspects of our hypotheses havebeen confirmed and other portions have not pannedout What we see is that contact theory and racial dis-tancing are not mutually exclusive

For the most part Latino immigrants in Durhamhold negative stereotypical views of blacks withLatino males holding more stereotypical views ofblack Americans than do Latinas The presence of neg-ative stereotypes of black Americans among Latinoimmigrants is consistent with the findings of Mindi-ola Niemann and Rodriguez (2002) This is a trou-bling finding that raises additional questions Giventhat length of stay in the United States appears to beunrelated to the strength of negative stereotypes aboutblacks this finding suggests that Latino immigrantsmight possibly bring views of the racial hierarchies intheir own countries with them to the United StatesSince the research on race and Latin America andMexico identifies blacks as representing the bottomrungs of society and the presence of the process ofldquowhitening uprdquo we assume that they might bring prej-udicial attitudes with them

On the other hand several factors do appear toreduce Latino immigrantsrsquo negative stereotypes ofblack Americansmdashincreases in education a sense oflinked fate on the part of Latino immigrants withother Latinos and more social interaction with blacksIn this instance contact theory as defined as socialcontact seems to bring about a positive change in atti-tude Moreover the presence of linked fate mitigatesnegative attitudes The significance of this predictorhas not been studied to any significant extent in theLatino politics literature and clearly deserves addi-

tional study something we plan to do as we continueour work on this project

In addition to holding negative stereotypical viewsof black Americans Latino immigrants do indeed feelthat they have more in common with whites than withblacks Moreover living in the same neighborhoods asblacks contrary to our expectations appears to rein-force the view on the part of Latino immigrants thatthey have more in common with whites and the leastin common with blacks Thus while social contact inthe previous hypothesis reduces negative stereotypicalviews Latino immigrants living in the same neigh-borhoods with blacks pushes them farther away fromblacks and closer to whites In this instance contacttheory does not appear to work when the contactcomes about by living in the same neighborhood Ifas some scholars suggest the longer some Latinoimmigrants remain in the United States the morelikely they are to begin to see themselves as ldquocollectiveblacksrdquo then our results suggest that this is not nec-essarily the case at least for the Latino immigrants inour Southern location (Bonilla-Silva 2004) While ourstatistical analyses shows that Latinosrsquo negative stereo-types of blacks and the likelihood that Latinos iden-tify more with whites than with blacks decreases withlength of stay in this country the effects are statisti-cally uncertain It may be that with a considerablylarger number of observations one might find a significant decline in attachments to whites but thevery small magnitude of these relationships does notprovide this argument with much traction Againhowever the presence of Latino immigrant linked fatemoves Latino immigrants closer to blacks and awayfrom whites This finding is consistent with Kauf-mannrsquos (2003) findings on Latinos in generalmdashLatinos who feel closer to other Latinos are more likelyto feel closer to blacks

Our findings on the dimension of negative stereo-types of black Americans held by Latino immigrantsare not merely a confirmation of previous results butrepresent a significant difference from previous find-ings While our result may conform to that of Mindi-ola Niemann and Rodriguez (2002) the context isdifferent and the possible effects more profoundMindiola at al examined attitudes in Houston Texaswhere Latino immigrants were integrating into anexisting Latino American population with a longhistory in Texas in general and Houston in particularAttitudes of Latino immigrants could possibly bemoderated and ameliorated through interaction withnative-born Latino American populations This situa-tion does not exist however in the South whereAmerican Latino communities are not present to any

24For blacks none of the predictors are statistically significant(Results are not shown) Moreover the direction of the coefficientsis opposite what we might have hypothesized

great extent Therefore this means Latino immigrantshave no reference point for black Americans otherthan their own attitudes which might have beenformed prior to their arrival in the United States andwhich do not appear to dissipate with increased inter-action with blacks Yet the finding that these negativeattitudes are modulated by a sense of linked fate sug-gests possibilities for the formation of connections toblack Americans in the absence of the presence of anextant American Latino community

Whatrsquos more these findings are importantbecause the South still contains the largest populationof black Americans in the United States and is con-sidered their ldquoregional homelandrdquo Moreover theSouth has suffered through some of the most politi-cized battles over race relations in recent history Noother section of the country has been as rigidlydefined along the black-white racial divide as has theSouth While we did not test directly Keyrsquos assertionsof over 50 years ago of the continued salience of racein the South the black-white divide in the South iscertain to shape the attitudes and incentives of Latinosin the region and how these new Latino immigrantssituate themselves vis-agrave-vis black Americans has pro-found implications for the social and political fabricof the South

What do these findings mean for Southern poli-tics and the politics of race We must be cautious indrawing broad conclusions about the future politics ofthe South from a one-city study with a small surveysample Yet our findings are suggestive of possible pat-terns that might be exhibited in other parts of theSouth and implicative of future trends

Latino immigrantsrsquo negative views of black Amer-icans most likely brought with them from their homecountries and reinforced rather than reduced byneighborhood interactions with blacks suggest thatthese new Latino immigrants may behave in wayssimilar to the Chinese in Mississippi in the mid-nineteenth century and the Cubans in Miami in themid-twentieth centurymdashidentification with whitesdistancing themselves from blacks and feeling noresponsibility to rectify the continuing inequalities of black Americans Given the increasing number ofLatino immigrants in the South and the possibilitythat over time their numbers might rival or evensurpass black Americans in the region if large por-tions of Latino immigrants maintain negative atti-tudes of black Americans where will this leave blacksWill blacks find that they must not only makedemands on whites for continued progress but alsomount a fight on another front against Latinos Orwill Latino immigrants begin to see themselves as

closer to blacks the longer they reside in the UnitedStates as some scholars suggest Clearly in order toanswer this question we need to do similar work in aseries of Southern locations But based on our resultsin this paper which are by no means definitive but arehighly suggestive we fear the former rather than thelatter is the future of the politics of race in the South

Acknowledgment

This is a revised version of a paper presented at theannual meeting of the Midwest Political Science Asso-ciation Chicago IL April 14ndash18 2004 The surveyreported in this paper was funded by a grant from TheFord Foundation (St Benedict the Black Meets theVirgin of Guadalupe Project Grant 1025-1445) Wethank the Foundation and Dr Melvin Oliver formerVice President of Asset Building and CommunityDevelopment now Dean of Social Science at the University of California Santa Barbara for theirsupport of our research We would also like to thankthe three anonymous reviewers whose commentspushed us to think more critically about this impor-tant topic We are very appreciative of their efforts onour behalf

Manuscript submitted 9 May 2005Manuscript accepted for publication 18 October 2005

References

Allport Gordon W 1954 The Nature of Prejudice CambridgeAddison-Wesley

Amir Yehuda 1969 ldquoContact Hypothesis in Ethnic RelationsrdquoPsychological Bulletin 71 319ndash42

Amir Yehuda 1976 ldquoThe Role of Intergroup Contact in Changeof Prejudice and Ethnic Relationsrdquo In Toward the Eliminationof Racism ed Phyllis A Katz New York Pergamonpp 245ndash308

Assis Claudia and Julian Pecquet 2002 ldquoHispanicsrsquo Search for aBetter Life Pushes Durham into Povertyrdquo The Herald Sun 25September A12

Bonilla-Silva Eduardo 2004 ldquoFrom Bi-Racial to Tri-RacialTowards a New System of Racial Stratification in the USArdquoEthnic and Racial Studies 27 (6) 931ndash50

Ciscel David H Barbara Ellen Smith and Marcela Mendoza2003 ldquoGhosts in the Global Machine New Immigrants and theRedefinition of Workrdquo Journal of Economic Issues 37 (2)333ndash41

Clark Juan M 1991 ldquoThe Social Impact of Cuban Immigrationin Floridardquo In Cuban Exiles in Florida Their Presence and Con-tributions eds Antonio Jorge Jamie Suchlicki and AdolfoLeyva de Varona University of Miami Research Institute forCuban Studies pp 39ndash61

rsquo

Cobb James C and William Stueck eds 2005 Globalization andthe American South Athens University of Georgia Press

Davis Darren W and Ronald E Brown 2002 ldquoThe Antipathy ofBlack Nationalism Behavioral and Attitudinal Implications ofan African American Ideologyrdquo American Journal of PoliticalScience 46 (2) 239ndash53

Dawson Michael C 1994 Behind the Mule Race and Class inAfrican-American Politics Princeton Princeton UniversityPress

de la Cadena Marisol 2001 ldquoReconstructing Race RacismCulture and Mestizaje in Latin Americardquo NACLA Report on theAmericas 34 (6) 16ndash23

de la Fuente Alejandro 1995 ldquoRace and Inequality in Cuba1899ndash1981rdquo Journal of Contemporary History 30 (1) 131ndash69

de la Fuente Alejandro 1998 ldquoRace National Discourse and Politics in Cubardquo Latin American Perspectives 25 (3) 43ndash70

Dulitzky Ariel E 2005 ldquoA Region in Denial Racial Discrimina-tion and Racism in Latin Americardquo In Neither Enemies norFriends Latinos Blacks Afro-Latinos eds Anani Dzidzienyoand Suzanne Oboler New York Palgrave-Macmillan 39ndash59

Duchon Deborah A and Arthur D Murphy 2001 ldquoIntroduc-tion From Patrones and Caciques to Good Ole Boysrdquo In LatinoWorkers in the Contemporary South eds Arthur D MurphyColleen Blanchard and Jennifer A Hill Athens University ofGeorgia Press pp 1ndash9

Durand Jorge Douglas S Massey and Fernando Charvet 2000ldquoThe Changing Geography of Mexican Immigration to theUnited States 1910ndash1996rdquo Social Science Quarterly 81 (1)1ndash16

Ellison Christopher G and Daniel A Powers 1994 ldquoThe ContactHypothesis and Racial Attitudes Among Black AmericansrdquoSocial Science Quarterly 75 (2) 385ndash400

Fossett Mark A and K Jill Kiecolt 1989 ldquoThe Relative Size ofMinority Populations and White Racial Attitudesrdquo SocialScience Quarterly 70 (4) 820ndash35

Geipel John 1997 ldquoBrazilrsquos African Legacyrdquo History Today 47 (8)18ndash24

Gong Lum v Rice 1927 275 US 78

Grenier Guillermo J and Max Castro 2001 ldquoBlacks and Cubansin Miami The Negative Consequences of the Cuban Enclaveon Ethnic Relationsrdquo In Governing American Cities Inter-Ethnic Coalitions Competition and Conflict ed Michael Jones-Correa New York Russell Sage Foundation pp 137ndash57

Griffith David 1993 Jonesrsquos Minimal Low-Wage Labor in theUnited States Albany State University of New York Press

Guimaraes Antonio Sergio 2001 ldquoRace Class and Color BehindBrazilrsquos lsquoRacial Democracyrsquordquo NACLA Report on the Americas 34(6) 38ndash41

Hanchard Michael G 1994 Orpheus and Power The MovimentoNegro of Rio de Janeiro and Satildeo Paulo Brazil 1945ndash1988Princeton Princeton University Press

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2000 ldquo lsquoMakingCarpet by the Milersquo The Emergence of a Mexican ImmigrantCommunity in an Industrial Region of the USHistoric SouthrdquoSocial Science Quarterly 81 (1) 49ndash66

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2001 ldquoA New Desti-nation for an Old Migration Origins Trajectories and LaborMarket Incorporation of Latinos in Dalton Georgiardquo In LatinoWorkers in the Contemporary South eds Arthur D MurphyColleen Blanchard and Jennifer A Hill Athens University ofGeorgia Press pp 127ndash35

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2003 ldquoMexicanImmigrant Communities in the South and Social Capital The

Case of Dalton Georgiardquo Southern Rural Sociology 19 (1)20ndash45

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2005 ldquoAppalachiaMeets Aztlaacuten Mexican Immigration and Intergroup Relationsin Dalton Georgiardquo In New Destinations Mexican Immigrationin the United States eds Rubeacuten Hernaacutendez-Leoacuten and VictorZuntildeiga New York Russell Sage Foundation pp 244ndash73

Hood M W III and Irwin L Morris 1998 ldquoGive Us Your TiredYour Poor But Make Sure They Have a Green Card TheEffects of Documented and Undocumented Migrant Contexton Anglo Opinion Toward Immigrationrdquo Political Behavior 20(1) 1ndash15

Jackman Mary R and Marie Crane 1986 ldquorsquoSome of my BestFriends are Black rsquo Interracial Friendship and WhitesrsquoRacial Attitudesrdquo Public Opinion Quarterly 50 (4) 459ndash86

Kandel William and Emilio A Parrado 2004 ldquoHispanics in theAmerican South and the Transformation of the Poultry Indus-tryrdquo In Hispanic Spaces Latino Places Community and CulturalDiversity in Contemporary America ed Daniel D ArreolaAustin University of Texas Press pp 255ndash76

Kaufmann Karen M 2003 ldquoCracks in the Rainbow Group Commonality as a Basis for Latino and African-AmericanPolitical Coalitionsrdquo Political Research Quarterly 56 (2)199ndash210

Key V O 1949 Southern Politics New York Vintage Books

Kinder Donald R and Tali Mendelberg 1995 ldquoCracks in Ameri-can Apartheid The Political Impact of Prejudice AmongDesegregated Whitesrdquo The Journal of Politics 57 (2) 402ndash24

King Gary Michael Tomz and Jason Wittenberg 2000 ldquoMakingthe Most of Statistical Analyses Improving Interpretation andPresentationrdquo American Journal of Political Science 44 (2)347ndash61

Loewen James W 1971 The Mississippi Chinese Between Blackand White Cambridge Harvard University Press

Massey Douglas Jorge Durand and Nolan J Malone 2002Beyond Smoke and Mirrors Mexican Immigration in an Era ofEconomic Integration New York Russell Sage Foundation

Matthews Donald R and James W Prothro 1967 Negroes and theNew Southern Politics New York Harcourt Brace amp World Inc

McClain Paula D Monique L Lyle Niambi M Carter Gerald FLackey Kendra Davenport Cotton Shayla C NunnallyThomas J Scotto Jeffrey D Grynaviski and J Alan Kendrick2005 ldquoBlack Americans and Latino Immigrants in a SouthernCity Friendly Neighbors or Economic Competitorsrdquo Type-script Duke University

Menchaca Martha 2001 Recovering History Constructing RaceThe Indian Black and White Roots of Mexican AmericansAustin University of Texas Press

Mindiola Jr Tatcho Yolanda Flores Niemann and NestorRodriguez 2002 Black-Brown Relations and StereotypesAustin University of Texas Press

Mohl Raymond A 1990 ldquoOn the Edge Blacks and Hispanics inMetropolitan Miami since 1959rdquo Florida Historical Quarterly69 (1) 37ndash57

Mohl Raymond A 2003 ldquoGlobalization Latinization and theNuevo New Southrdquo Journal of American Ethnic History 22 (4)31ndash66

Moumlrner Magnus ed 1967 Race Mixture in the History of LatinAmerica Boston Little Brown and Company

Morris Alan 1999 ldquoRace Relations and Racism in a RaciallyDiverse Inner City Neighbourhood A Case Study of HillbrowJohannesburgrdquo Journal of Southern African Studies 25 (4)667ndash94

Murphy Arthur D Colleen Blanchard and Jennifer A Hill 2001Latino Workers in the Contemporary South Athens Universityof Georgia Press

Oliver J Eric and Janelle Wong 2003 ldquoInter-group Prejudice inMultiethnic Settingsrdquo American Journal of Political Science 47(4) 567ndash82

Peacock James L Harry L Watson and Carrie R Matthews ed2005 The American South in a Global World Chapel Hill Uni-versity of North Carolina Press

Powers Daniel and Christopher Ellison 1994 ldquoInterracialContact and Black Racial Attitudesrdquo Social Forces 74 (1)205ndash27

Quan Robert Seto 1982 Lotus Among the Magnolias The Missis-sippi Chinese Jackson University Press of Mississippi

Quillian Lincoln 1996 ldquoGroup Threat and Regional Change inAttitudes Towards African Americansrdquo American Journal ofSociology 102 (3) 816ndash60

Rice et al v Gong Lum et al 1925 139 Miss 760 104 So 105

Scott Rebecca J 1985 Slave Emancipation in Cuba The Transitionto Free Labor 1860ndash1899 Princeton Princeton UniversityPress

Seed Patricia 1982 ldquoSocial Dimensions of Race Mexico City1753rdquo The Hispanic American Historical Review 62 (4)569ndash606

Shingles Richard D 1981 ldquoBlack Consciousness and Political Par-ticipation The Missing Linkrdquo American Political Science Review75 (2) 76ndash90

Sigelman Lee and Susan Welch 1993 ldquoThe Contact HypothesisRevisited Black-White Interaction and Positive Racial Atti-tudesrdquo Social Forces 71 (3) 781ndash95

Stephan Walter G and Cookie White Stephan 1985 ldquoIntergroupAnxietyrdquo The Journal of Social Issues 41 (3) 157ndash75

Suro Roberto and Audrey Singer 2002 Latino Growth in Metro-politan America Changing Patterns New Locations Washing-ton Brookings Institution Center on Urban and MetropolitanPolicy and The Pew Hispanic Center

Sweet James H 1997 ldquoThe Iberian Roots of American RacistThoughtrdquo The William and Mary Quarterly 54 (1) 143ndash66

Tate Katherine 1993 From Protest to Politics The New Black Votersin American Elections (enlarged edition) Cambridge HarvardUniversity Press and Russell Sage Foundation

Taylor Marylee C 1998 ldquoHow White Attitudes Vary with theRacial Composition of Local Populations Numbers CountrdquoAmerican Sociological Review 63 (4) 512ndash35

Tomz Michael Jason Wittenberg and Gary King 2003 CLARIFYSoftware for Interpreting and Presenting Statistical ResultsVersion 21 Stanford University University of Wisconsinand Harvard University httpgkingharvardedustatsshtml(August 9 2003)

Torres Cruz C 2000 ldquoEmerging Latino Communities A NewChallenge for the Rural Southrdquo The Southern Rural Develop-ment Center No 12 (August) 1ndash8

US Census Bureau 2002a Census of PopulationhttpfactfindercensusgovservlettQTTable (August 132002)

US Census Bureau 2002b Redistricting File httpcenstatscensusgovdataNC160379000pdf (August 132002)

US Census Bureau 2003 US Interim Projections httpwwwcensusgovusinterimproj (September 25 2003)

Wade Peter 1993 Blackness and Race Mixture The Dynamics ofRacial Identity in Colombia Baltimore The Johns HopkinsUniversity Press

Wade Peter 1997 Race and Ethnicity in Latin America LondonPluto Press

Welch Susan Lee Sigelman Timothy Bledsoe and MichaelCombs 2001 Race and Place Race Relations in an AmericanCity New York Cambridge University Press

Winant Howard 1992 ldquoRethinking Race in Brazilrdquo Journal ofLatin American Studies 24 (1) 173ndash92

Paula D McClain is professor of political scienceDuke University Durham NC 27708-0204 NiambiM Carter is PhD candidate in political science DukeUniversity Durham NC 27708-0204 Victoria MDeFrancesco Soto is PhD candidate in politicalscience Duke University Durham NC 27708-0204Monique L Lyle is PhD candidate in political scienceDuke University Durham NC 27708-0204 Jeffrey DGrynaviski is assistant professor of political scienceUniversity of Chicago Chicago IL 60637 Shayla CNunnally is assistant professor of political science andInstitute of African American Studies University ofConnecticut Storrs CT 06269-1024 Thomas J Scottois assistant professor of political science West VirginiaUniversity Morganstown WV 26506 J Alan Kendrickis assistant professor of political science St Augus-tinersquos College Raleigh NC 27610 Gerald F Lackey isgraduate student in sociology University of NorthCarolina-Chapel Hill Chapel Hill NC 27599-3210Kendra Davenport Cotton is PhD candidate in polit-ical science University of North Carolina-Chapel HillChapel Hill NC 27599-3210

and the coolness toward blacks among Latino respon-dents are not mutual (See Table A3 in the JOP onlineappendix) The data show that while 459 of whiterespondents see themselves as having the most incommon with blacks just 222 of whites see them-selves as having the most in common with LatinosSimilarly a plurality of whites (475) sees themselvesas having the least in common with Latinos Blackrespondents on the other hand do not generally feel as distant from Latinos as Latinos feel from blacks About half of the blacks in the sample reporthaving the most in common with Latinos (496)while an almost similar proportion (455) reporthaving the most in common with whites Just 227of blacks report having the least in common withLatinos22

To test our expectation that increases in length oftime in the United States and interaction with blackAmericans move Latino immigrants closer to blackswe used a logit model with our dependent variablebeing a measure of which group blacks or whitesLatino immigrants feel they have the most in commonwith In this instance we found including both contactmeasures simultaneously social contact and blacks inthe neighborhood resulted in both being statisticallyinsignificant but since the number of blacks in therespondentrsquos neighborhood was closer to significancethan social contact we decided to include it in ourmodel In addition to our primary independent vari-ables we also added a measure of the extent to whichour Latino immigrant respondents hold stereotypicalviews of black Americans We used the dependentvariable of the stereotype index as a predictor in thisanalysis The Latino linked fate variable is alsoincluded as a predictor

Our results for this hypothesis are shown in Table3 Our analyses do not reveal a statistically significantrelationship between educational attainment andlength of time in the United States with closeness to aparticular racial group The coefficients however onboth variables are in the predicted direction thus sug-gesting the relationship that we hypothesized but notstatistically significant The coefficient on the numberof blacks in a respondentrsquos neighborhood seems to reveala relationship that is opposite of what we hypothe-sizedmdashthe greater the number of blacks in the neigh-borhood the more likely Latino immigrants are to feelthey have more in common with whitesmdashbut it is also

statistically insignificant23 The most statistically sig-nificant variables in our analysis are the stereotypeindex and linked fate with the results revealing thatthose Latino immigrants with linked fate and with lessstereotypical views of black Americans are more likelyto feel they have the most in common with blacks andleast likely to believe they have the most in commonwith whites Once again a sense of group identityamong Latino immigrants contributes significantly toa sense of connection with black Americans

We compute percentage changes in probability aswe move from one value on our independent variablesto another Substantively therefore an average Latinoimmigrant (ie with the mean levels of educationlength of time in the United States neighborhoodinteraction with black Americans and perception ofblack Americans on the stereotype index) with linkedfate is approximately 19 more likely than an averageLatino without linked fate to feel that she has the mostin common with black Americans Though the rela-

22Given the small numbers of Asians in the Durham area (andnationally) it is interesting to note that between 30 and 40 ofall races reported having the least in common and 32 of whitesreported themselves as having the most in common with Asians

T 3 Logit Results for Which Group LatinoRespondents Report Having the Most inCommon With

Variables Coefficients

Education minus0735(0576)

Linked Fate minus2005(8703)

Length of Time in the United States minus0301(0393)

Number of Blacks in Neighborhood 2897(2826)

Stereotypical Views of Blacks 2293(Higher values indicate holding more (1269)stereotypical views of blacks)

Constant 9166N 102Chi-Squared 1421Pseudo R2 1407

Note The dependent variable for this model is coded such that 1indicates having the most in common with whites and 0 indicateshaving the most in common with blacks Standard error estimatesare in parenthesesp lt 01 (two-tailed test)p lt 05 (two-tailed test)p lt 10 (two-tailed test)

23As with the first hypothesis we included interaction terms oflinked fate with neighborhood contact with blacks and educationwith neighborhood contact with blacks individually and in combination None of the interaction terms was statistically significant

rsquo

tionship between perceptions of black Americans onthe stereotype index and closeness to whites or blacksis statistically significant the substantive effect isminuscule Latino immigrants with highly stereotyp-ical views of black Americans are less than 1 (22)more likely than other Latino immigrants to feel theyhave the most in common with whites24

Discussion and Conclusion

We posited two hypothesesmdashLatino immigrants holdnegative stereotypes of blacks and Latino immigrantsfeel they have the most in common with whites bothattitudes that may be mitigated by contact factors andlength of time in the countrymdashas the core questionsfor this paper Some aspects of our hypotheses havebeen confirmed and other portions have not pannedout What we see is that contact theory and racial dis-tancing are not mutually exclusive

For the most part Latino immigrants in Durhamhold negative stereotypical views of blacks withLatino males holding more stereotypical views ofblack Americans than do Latinas The presence of neg-ative stereotypes of black Americans among Latinoimmigrants is consistent with the findings of Mindi-ola Niemann and Rodriguez (2002) This is a trou-bling finding that raises additional questions Giventhat length of stay in the United States appears to beunrelated to the strength of negative stereotypes aboutblacks this finding suggests that Latino immigrantsmight possibly bring views of the racial hierarchies intheir own countries with them to the United StatesSince the research on race and Latin America andMexico identifies blacks as representing the bottomrungs of society and the presence of the process ofldquowhitening uprdquo we assume that they might bring prej-udicial attitudes with them

On the other hand several factors do appear toreduce Latino immigrantsrsquo negative stereotypes ofblack Americansmdashincreases in education a sense oflinked fate on the part of Latino immigrants withother Latinos and more social interaction with blacksIn this instance contact theory as defined as socialcontact seems to bring about a positive change in atti-tude Moreover the presence of linked fate mitigatesnegative attitudes The significance of this predictorhas not been studied to any significant extent in theLatino politics literature and clearly deserves addi-

tional study something we plan to do as we continueour work on this project

In addition to holding negative stereotypical viewsof black Americans Latino immigrants do indeed feelthat they have more in common with whites than withblacks Moreover living in the same neighborhoods asblacks contrary to our expectations appears to rein-force the view on the part of Latino immigrants thatthey have more in common with whites and the leastin common with blacks Thus while social contact inthe previous hypothesis reduces negative stereotypicalviews Latino immigrants living in the same neigh-borhoods with blacks pushes them farther away fromblacks and closer to whites In this instance contacttheory does not appear to work when the contactcomes about by living in the same neighborhood Ifas some scholars suggest the longer some Latinoimmigrants remain in the United States the morelikely they are to begin to see themselves as ldquocollectiveblacksrdquo then our results suggest that this is not nec-essarily the case at least for the Latino immigrants inour Southern location (Bonilla-Silva 2004) While ourstatistical analyses shows that Latinosrsquo negative stereo-types of blacks and the likelihood that Latinos iden-tify more with whites than with blacks decreases withlength of stay in this country the effects are statisti-cally uncertain It may be that with a considerablylarger number of observations one might find a significant decline in attachments to whites but thevery small magnitude of these relationships does notprovide this argument with much traction Againhowever the presence of Latino immigrant linked fatemoves Latino immigrants closer to blacks and awayfrom whites This finding is consistent with Kauf-mannrsquos (2003) findings on Latinos in generalmdashLatinos who feel closer to other Latinos are more likelyto feel closer to blacks

Our findings on the dimension of negative stereo-types of black Americans held by Latino immigrantsare not merely a confirmation of previous results butrepresent a significant difference from previous find-ings While our result may conform to that of Mindi-ola Niemann and Rodriguez (2002) the context isdifferent and the possible effects more profoundMindiola at al examined attitudes in Houston Texaswhere Latino immigrants were integrating into anexisting Latino American population with a longhistory in Texas in general and Houston in particularAttitudes of Latino immigrants could possibly bemoderated and ameliorated through interaction withnative-born Latino American populations This situa-tion does not exist however in the South whereAmerican Latino communities are not present to any

24For blacks none of the predictors are statistically significant(Results are not shown) Moreover the direction of the coefficientsis opposite what we might have hypothesized

great extent Therefore this means Latino immigrantshave no reference point for black Americans otherthan their own attitudes which might have beenformed prior to their arrival in the United States andwhich do not appear to dissipate with increased inter-action with blacks Yet the finding that these negativeattitudes are modulated by a sense of linked fate sug-gests possibilities for the formation of connections toblack Americans in the absence of the presence of anextant American Latino community

Whatrsquos more these findings are importantbecause the South still contains the largest populationof black Americans in the United States and is con-sidered their ldquoregional homelandrdquo Moreover theSouth has suffered through some of the most politi-cized battles over race relations in recent history Noother section of the country has been as rigidlydefined along the black-white racial divide as has theSouth While we did not test directly Keyrsquos assertionsof over 50 years ago of the continued salience of racein the South the black-white divide in the South iscertain to shape the attitudes and incentives of Latinosin the region and how these new Latino immigrantssituate themselves vis-agrave-vis black Americans has pro-found implications for the social and political fabricof the South

What do these findings mean for Southern poli-tics and the politics of race We must be cautious indrawing broad conclusions about the future politics ofthe South from a one-city study with a small surveysample Yet our findings are suggestive of possible pat-terns that might be exhibited in other parts of theSouth and implicative of future trends

Latino immigrantsrsquo negative views of black Amer-icans most likely brought with them from their homecountries and reinforced rather than reduced byneighborhood interactions with blacks suggest thatthese new Latino immigrants may behave in wayssimilar to the Chinese in Mississippi in the mid-nineteenth century and the Cubans in Miami in themid-twentieth centurymdashidentification with whitesdistancing themselves from blacks and feeling noresponsibility to rectify the continuing inequalities of black Americans Given the increasing number ofLatino immigrants in the South and the possibilitythat over time their numbers might rival or evensurpass black Americans in the region if large por-tions of Latino immigrants maintain negative atti-tudes of black Americans where will this leave blacksWill blacks find that they must not only makedemands on whites for continued progress but alsomount a fight on another front against Latinos Orwill Latino immigrants begin to see themselves as

closer to blacks the longer they reside in the UnitedStates as some scholars suggest Clearly in order toanswer this question we need to do similar work in aseries of Southern locations But based on our resultsin this paper which are by no means definitive but arehighly suggestive we fear the former rather than thelatter is the future of the politics of race in the South

Acknowledgment

This is a revised version of a paper presented at theannual meeting of the Midwest Political Science Asso-ciation Chicago IL April 14ndash18 2004 The surveyreported in this paper was funded by a grant from TheFord Foundation (St Benedict the Black Meets theVirgin of Guadalupe Project Grant 1025-1445) Wethank the Foundation and Dr Melvin Oliver formerVice President of Asset Building and CommunityDevelopment now Dean of Social Science at the University of California Santa Barbara for theirsupport of our research We would also like to thankthe three anonymous reviewers whose commentspushed us to think more critically about this impor-tant topic We are very appreciative of their efforts onour behalf

Manuscript submitted 9 May 2005Manuscript accepted for publication 18 October 2005

References

Allport Gordon W 1954 The Nature of Prejudice CambridgeAddison-Wesley

Amir Yehuda 1969 ldquoContact Hypothesis in Ethnic RelationsrdquoPsychological Bulletin 71 319ndash42

Amir Yehuda 1976 ldquoThe Role of Intergroup Contact in Changeof Prejudice and Ethnic Relationsrdquo In Toward the Eliminationof Racism ed Phyllis A Katz New York Pergamonpp 245ndash308

Assis Claudia and Julian Pecquet 2002 ldquoHispanicsrsquo Search for aBetter Life Pushes Durham into Povertyrdquo The Herald Sun 25September A12

Bonilla-Silva Eduardo 2004 ldquoFrom Bi-Racial to Tri-RacialTowards a New System of Racial Stratification in the USArdquoEthnic and Racial Studies 27 (6) 931ndash50

Ciscel David H Barbara Ellen Smith and Marcela Mendoza2003 ldquoGhosts in the Global Machine New Immigrants and theRedefinition of Workrdquo Journal of Economic Issues 37 (2)333ndash41

Clark Juan M 1991 ldquoThe Social Impact of Cuban Immigrationin Floridardquo In Cuban Exiles in Florida Their Presence and Con-tributions eds Antonio Jorge Jamie Suchlicki and AdolfoLeyva de Varona University of Miami Research Institute forCuban Studies pp 39ndash61

rsquo

Cobb James C and William Stueck eds 2005 Globalization andthe American South Athens University of Georgia Press

Davis Darren W and Ronald E Brown 2002 ldquoThe Antipathy ofBlack Nationalism Behavioral and Attitudinal Implications ofan African American Ideologyrdquo American Journal of PoliticalScience 46 (2) 239ndash53

Dawson Michael C 1994 Behind the Mule Race and Class inAfrican-American Politics Princeton Princeton UniversityPress

de la Cadena Marisol 2001 ldquoReconstructing Race RacismCulture and Mestizaje in Latin Americardquo NACLA Report on theAmericas 34 (6) 16ndash23

de la Fuente Alejandro 1995 ldquoRace and Inequality in Cuba1899ndash1981rdquo Journal of Contemporary History 30 (1) 131ndash69

de la Fuente Alejandro 1998 ldquoRace National Discourse and Politics in Cubardquo Latin American Perspectives 25 (3) 43ndash70

Dulitzky Ariel E 2005 ldquoA Region in Denial Racial Discrimina-tion and Racism in Latin Americardquo In Neither Enemies norFriends Latinos Blacks Afro-Latinos eds Anani Dzidzienyoand Suzanne Oboler New York Palgrave-Macmillan 39ndash59

Duchon Deborah A and Arthur D Murphy 2001 ldquoIntroduc-tion From Patrones and Caciques to Good Ole Boysrdquo In LatinoWorkers in the Contemporary South eds Arthur D MurphyColleen Blanchard and Jennifer A Hill Athens University ofGeorgia Press pp 1ndash9

Durand Jorge Douglas S Massey and Fernando Charvet 2000ldquoThe Changing Geography of Mexican Immigration to theUnited States 1910ndash1996rdquo Social Science Quarterly 81 (1)1ndash16

Ellison Christopher G and Daniel A Powers 1994 ldquoThe ContactHypothesis and Racial Attitudes Among Black AmericansrdquoSocial Science Quarterly 75 (2) 385ndash400

Fossett Mark A and K Jill Kiecolt 1989 ldquoThe Relative Size ofMinority Populations and White Racial Attitudesrdquo SocialScience Quarterly 70 (4) 820ndash35

Geipel John 1997 ldquoBrazilrsquos African Legacyrdquo History Today 47 (8)18ndash24

Gong Lum v Rice 1927 275 US 78

Grenier Guillermo J and Max Castro 2001 ldquoBlacks and Cubansin Miami The Negative Consequences of the Cuban Enclaveon Ethnic Relationsrdquo In Governing American Cities Inter-Ethnic Coalitions Competition and Conflict ed Michael Jones-Correa New York Russell Sage Foundation pp 137ndash57

Griffith David 1993 Jonesrsquos Minimal Low-Wage Labor in theUnited States Albany State University of New York Press

Guimaraes Antonio Sergio 2001 ldquoRace Class and Color BehindBrazilrsquos lsquoRacial Democracyrsquordquo NACLA Report on the Americas 34(6) 38ndash41

Hanchard Michael G 1994 Orpheus and Power The MovimentoNegro of Rio de Janeiro and Satildeo Paulo Brazil 1945ndash1988Princeton Princeton University Press

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2000 ldquo lsquoMakingCarpet by the Milersquo The Emergence of a Mexican ImmigrantCommunity in an Industrial Region of the USHistoric SouthrdquoSocial Science Quarterly 81 (1) 49ndash66

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2001 ldquoA New Desti-nation for an Old Migration Origins Trajectories and LaborMarket Incorporation of Latinos in Dalton Georgiardquo In LatinoWorkers in the Contemporary South eds Arthur D MurphyColleen Blanchard and Jennifer A Hill Athens University ofGeorgia Press pp 127ndash35

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2003 ldquoMexicanImmigrant Communities in the South and Social Capital The

Case of Dalton Georgiardquo Southern Rural Sociology 19 (1)20ndash45

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2005 ldquoAppalachiaMeets Aztlaacuten Mexican Immigration and Intergroup Relationsin Dalton Georgiardquo In New Destinations Mexican Immigrationin the United States eds Rubeacuten Hernaacutendez-Leoacuten and VictorZuntildeiga New York Russell Sage Foundation pp 244ndash73

Hood M W III and Irwin L Morris 1998 ldquoGive Us Your TiredYour Poor But Make Sure They Have a Green Card TheEffects of Documented and Undocumented Migrant Contexton Anglo Opinion Toward Immigrationrdquo Political Behavior 20(1) 1ndash15

Jackman Mary R and Marie Crane 1986 ldquorsquoSome of my BestFriends are Black rsquo Interracial Friendship and WhitesrsquoRacial Attitudesrdquo Public Opinion Quarterly 50 (4) 459ndash86

Kandel William and Emilio A Parrado 2004 ldquoHispanics in theAmerican South and the Transformation of the Poultry Indus-tryrdquo In Hispanic Spaces Latino Places Community and CulturalDiversity in Contemporary America ed Daniel D ArreolaAustin University of Texas Press pp 255ndash76

Kaufmann Karen M 2003 ldquoCracks in the Rainbow Group Commonality as a Basis for Latino and African-AmericanPolitical Coalitionsrdquo Political Research Quarterly 56 (2)199ndash210

Key V O 1949 Southern Politics New York Vintage Books

Kinder Donald R and Tali Mendelberg 1995 ldquoCracks in Ameri-can Apartheid The Political Impact of Prejudice AmongDesegregated Whitesrdquo The Journal of Politics 57 (2) 402ndash24

King Gary Michael Tomz and Jason Wittenberg 2000 ldquoMakingthe Most of Statistical Analyses Improving Interpretation andPresentationrdquo American Journal of Political Science 44 (2)347ndash61

Loewen James W 1971 The Mississippi Chinese Between Blackand White Cambridge Harvard University Press

Massey Douglas Jorge Durand and Nolan J Malone 2002Beyond Smoke and Mirrors Mexican Immigration in an Era ofEconomic Integration New York Russell Sage Foundation

Matthews Donald R and James W Prothro 1967 Negroes and theNew Southern Politics New York Harcourt Brace amp World Inc

McClain Paula D Monique L Lyle Niambi M Carter Gerald FLackey Kendra Davenport Cotton Shayla C NunnallyThomas J Scotto Jeffrey D Grynaviski and J Alan Kendrick2005 ldquoBlack Americans and Latino Immigrants in a SouthernCity Friendly Neighbors or Economic Competitorsrdquo Type-script Duke University

Menchaca Martha 2001 Recovering History Constructing RaceThe Indian Black and White Roots of Mexican AmericansAustin University of Texas Press

Mindiola Jr Tatcho Yolanda Flores Niemann and NestorRodriguez 2002 Black-Brown Relations and StereotypesAustin University of Texas Press

Mohl Raymond A 1990 ldquoOn the Edge Blacks and Hispanics inMetropolitan Miami since 1959rdquo Florida Historical Quarterly69 (1) 37ndash57

Mohl Raymond A 2003 ldquoGlobalization Latinization and theNuevo New Southrdquo Journal of American Ethnic History 22 (4)31ndash66

Moumlrner Magnus ed 1967 Race Mixture in the History of LatinAmerica Boston Little Brown and Company

Morris Alan 1999 ldquoRace Relations and Racism in a RaciallyDiverse Inner City Neighbourhood A Case Study of HillbrowJohannesburgrdquo Journal of Southern African Studies 25 (4)667ndash94

Murphy Arthur D Colleen Blanchard and Jennifer A Hill 2001Latino Workers in the Contemporary South Athens Universityof Georgia Press

Oliver J Eric and Janelle Wong 2003 ldquoInter-group Prejudice inMultiethnic Settingsrdquo American Journal of Political Science 47(4) 567ndash82

Peacock James L Harry L Watson and Carrie R Matthews ed2005 The American South in a Global World Chapel Hill Uni-versity of North Carolina Press

Powers Daniel and Christopher Ellison 1994 ldquoInterracialContact and Black Racial Attitudesrdquo Social Forces 74 (1)205ndash27

Quan Robert Seto 1982 Lotus Among the Magnolias The Missis-sippi Chinese Jackson University Press of Mississippi

Quillian Lincoln 1996 ldquoGroup Threat and Regional Change inAttitudes Towards African Americansrdquo American Journal ofSociology 102 (3) 816ndash60

Rice et al v Gong Lum et al 1925 139 Miss 760 104 So 105

Scott Rebecca J 1985 Slave Emancipation in Cuba The Transitionto Free Labor 1860ndash1899 Princeton Princeton UniversityPress

Seed Patricia 1982 ldquoSocial Dimensions of Race Mexico City1753rdquo The Hispanic American Historical Review 62 (4)569ndash606

Shingles Richard D 1981 ldquoBlack Consciousness and Political Par-ticipation The Missing Linkrdquo American Political Science Review75 (2) 76ndash90

Sigelman Lee and Susan Welch 1993 ldquoThe Contact HypothesisRevisited Black-White Interaction and Positive Racial Atti-tudesrdquo Social Forces 71 (3) 781ndash95

Stephan Walter G and Cookie White Stephan 1985 ldquoIntergroupAnxietyrdquo The Journal of Social Issues 41 (3) 157ndash75

Suro Roberto and Audrey Singer 2002 Latino Growth in Metro-politan America Changing Patterns New Locations Washing-ton Brookings Institution Center on Urban and MetropolitanPolicy and The Pew Hispanic Center

Sweet James H 1997 ldquoThe Iberian Roots of American RacistThoughtrdquo The William and Mary Quarterly 54 (1) 143ndash66

Tate Katherine 1993 From Protest to Politics The New Black Votersin American Elections (enlarged edition) Cambridge HarvardUniversity Press and Russell Sage Foundation

Taylor Marylee C 1998 ldquoHow White Attitudes Vary with theRacial Composition of Local Populations Numbers CountrdquoAmerican Sociological Review 63 (4) 512ndash35

Tomz Michael Jason Wittenberg and Gary King 2003 CLARIFYSoftware for Interpreting and Presenting Statistical ResultsVersion 21 Stanford University University of Wisconsinand Harvard University httpgkingharvardedustatsshtml(August 9 2003)

Torres Cruz C 2000 ldquoEmerging Latino Communities A NewChallenge for the Rural Southrdquo The Southern Rural Develop-ment Center No 12 (August) 1ndash8

US Census Bureau 2002a Census of PopulationhttpfactfindercensusgovservlettQTTable (August 132002)

US Census Bureau 2002b Redistricting File httpcenstatscensusgovdataNC160379000pdf (August 132002)

US Census Bureau 2003 US Interim Projections httpwwwcensusgovusinterimproj (September 25 2003)

Wade Peter 1993 Blackness and Race Mixture The Dynamics ofRacial Identity in Colombia Baltimore The Johns HopkinsUniversity Press

Wade Peter 1997 Race and Ethnicity in Latin America LondonPluto Press

Welch Susan Lee Sigelman Timothy Bledsoe and MichaelCombs 2001 Race and Place Race Relations in an AmericanCity New York Cambridge University Press

Winant Howard 1992 ldquoRethinking Race in Brazilrdquo Journal ofLatin American Studies 24 (1) 173ndash92

Paula D McClain is professor of political scienceDuke University Durham NC 27708-0204 NiambiM Carter is PhD candidate in political science DukeUniversity Durham NC 27708-0204 Victoria MDeFrancesco Soto is PhD candidate in politicalscience Duke University Durham NC 27708-0204Monique L Lyle is PhD candidate in political scienceDuke University Durham NC 27708-0204 Jeffrey DGrynaviski is assistant professor of political scienceUniversity of Chicago Chicago IL 60637 Shayla CNunnally is assistant professor of political science andInstitute of African American Studies University ofConnecticut Storrs CT 06269-1024 Thomas J Scottois assistant professor of political science West VirginiaUniversity Morganstown WV 26506 J Alan Kendrickis assistant professor of political science St Augus-tinersquos College Raleigh NC 27610 Gerald F Lackey isgraduate student in sociology University of NorthCarolina-Chapel Hill Chapel Hill NC 27599-3210Kendra Davenport Cotton is PhD candidate in polit-ical science University of North Carolina-Chapel HillChapel Hill NC 27599-3210

rsquo

tionship between perceptions of black Americans onthe stereotype index and closeness to whites or blacksis statistically significant the substantive effect isminuscule Latino immigrants with highly stereotyp-ical views of black Americans are less than 1 (22)more likely than other Latino immigrants to feel theyhave the most in common with whites24

Discussion and Conclusion

We posited two hypothesesmdashLatino immigrants holdnegative stereotypes of blacks and Latino immigrantsfeel they have the most in common with whites bothattitudes that may be mitigated by contact factors andlength of time in the countrymdashas the core questionsfor this paper Some aspects of our hypotheses havebeen confirmed and other portions have not pannedout What we see is that contact theory and racial dis-tancing are not mutually exclusive

For the most part Latino immigrants in Durhamhold negative stereotypical views of blacks withLatino males holding more stereotypical views ofblack Americans than do Latinas The presence of neg-ative stereotypes of black Americans among Latinoimmigrants is consistent with the findings of Mindi-ola Niemann and Rodriguez (2002) This is a trou-bling finding that raises additional questions Giventhat length of stay in the United States appears to beunrelated to the strength of negative stereotypes aboutblacks this finding suggests that Latino immigrantsmight possibly bring views of the racial hierarchies intheir own countries with them to the United StatesSince the research on race and Latin America andMexico identifies blacks as representing the bottomrungs of society and the presence of the process ofldquowhitening uprdquo we assume that they might bring prej-udicial attitudes with them

On the other hand several factors do appear toreduce Latino immigrantsrsquo negative stereotypes ofblack Americansmdashincreases in education a sense oflinked fate on the part of Latino immigrants withother Latinos and more social interaction with blacksIn this instance contact theory as defined as socialcontact seems to bring about a positive change in atti-tude Moreover the presence of linked fate mitigatesnegative attitudes The significance of this predictorhas not been studied to any significant extent in theLatino politics literature and clearly deserves addi-

tional study something we plan to do as we continueour work on this project

In addition to holding negative stereotypical viewsof black Americans Latino immigrants do indeed feelthat they have more in common with whites than withblacks Moreover living in the same neighborhoods asblacks contrary to our expectations appears to rein-force the view on the part of Latino immigrants thatthey have more in common with whites and the leastin common with blacks Thus while social contact inthe previous hypothesis reduces negative stereotypicalviews Latino immigrants living in the same neigh-borhoods with blacks pushes them farther away fromblacks and closer to whites In this instance contacttheory does not appear to work when the contactcomes about by living in the same neighborhood Ifas some scholars suggest the longer some Latinoimmigrants remain in the United States the morelikely they are to begin to see themselves as ldquocollectiveblacksrdquo then our results suggest that this is not nec-essarily the case at least for the Latino immigrants inour Southern location (Bonilla-Silva 2004) While ourstatistical analyses shows that Latinosrsquo negative stereo-types of blacks and the likelihood that Latinos iden-tify more with whites than with blacks decreases withlength of stay in this country the effects are statisti-cally uncertain It may be that with a considerablylarger number of observations one might find a significant decline in attachments to whites but thevery small magnitude of these relationships does notprovide this argument with much traction Againhowever the presence of Latino immigrant linked fatemoves Latino immigrants closer to blacks and awayfrom whites This finding is consistent with Kauf-mannrsquos (2003) findings on Latinos in generalmdashLatinos who feel closer to other Latinos are more likelyto feel closer to blacks

Our findings on the dimension of negative stereo-types of black Americans held by Latino immigrantsare not merely a confirmation of previous results butrepresent a significant difference from previous find-ings While our result may conform to that of Mindi-ola Niemann and Rodriguez (2002) the context isdifferent and the possible effects more profoundMindiola at al examined attitudes in Houston Texaswhere Latino immigrants were integrating into anexisting Latino American population with a longhistory in Texas in general and Houston in particularAttitudes of Latino immigrants could possibly bemoderated and ameliorated through interaction withnative-born Latino American populations This situa-tion does not exist however in the South whereAmerican Latino communities are not present to any

24For blacks none of the predictors are statistically significant(Results are not shown) Moreover the direction of the coefficientsis opposite what we might have hypothesized

great extent Therefore this means Latino immigrantshave no reference point for black Americans otherthan their own attitudes which might have beenformed prior to their arrival in the United States andwhich do not appear to dissipate with increased inter-action with blacks Yet the finding that these negativeattitudes are modulated by a sense of linked fate sug-gests possibilities for the formation of connections toblack Americans in the absence of the presence of anextant American Latino community

Whatrsquos more these findings are importantbecause the South still contains the largest populationof black Americans in the United States and is con-sidered their ldquoregional homelandrdquo Moreover theSouth has suffered through some of the most politi-cized battles over race relations in recent history Noother section of the country has been as rigidlydefined along the black-white racial divide as has theSouth While we did not test directly Keyrsquos assertionsof over 50 years ago of the continued salience of racein the South the black-white divide in the South iscertain to shape the attitudes and incentives of Latinosin the region and how these new Latino immigrantssituate themselves vis-agrave-vis black Americans has pro-found implications for the social and political fabricof the South

What do these findings mean for Southern poli-tics and the politics of race We must be cautious indrawing broad conclusions about the future politics ofthe South from a one-city study with a small surveysample Yet our findings are suggestive of possible pat-terns that might be exhibited in other parts of theSouth and implicative of future trends

Latino immigrantsrsquo negative views of black Amer-icans most likely brought with them from their homecountries and reinforced rather than reduced byneighborhood interactions with blacks suggest thatthese new Latino immigrants may behave in wayssimilar to the Chinese in Mississippi in the mid-nineteenth century and the Cubans in Miami in themid-twentieth centurymdashidentification with whitesdistancing themselves from blacks and feeling noresponsibility to rectify the continuing inequalities of black Americans Given the increasing number ofLatino immigrants in the South and the possibilitythat over time their numbers might rival or evensurpass black Americans in the region if large por-tions of Latino immigrants maintain negative atti-tudes of black Americans where will this leave blacksWill blacks find that they must not only makedemands on whites for continued progress but alsomount a fight on another front against Latinos Orwill Latino immigrants begin to see themselves as

closer to blacks the longer they reside in the UnitedStates as some scholars suggest Clearly in order toanswer this question we need to do similar work in aseries of Southern locations But based on our resultsin this paper which are by no means definitive but arehighly suggestive we fear the former rather than thelatter is the future of the politics of race in the South

Acknowledgment

This is a revised version of a paper presented at theannual meeting of the Midwest Political Science Asso-ciation Chicago IL April 14ndash18 2004 The surveyreported in this paper was funded by a grant from TheFord Foundation (St Benedict the Black Meets theVirgin of Guadalupe Project Grant 1025-1445) Wethank the Foundation and Dr Melvin Oliver formerVice President of Asset Building and CommunityDevelopment now Dean of Social Science at the University of California Santa Barbara for theirsupport of our research We would also like to thankthe three anonymous reviewers whose commentspushed us to think more critically about this impor-tant topic We are very appreciative of their efforts onour behalf

Manuscript submitted 9 May 2005Manuscript accepted for publication 18 October 2005

References

Allport Gordon W 1954 The Nature of Prejudice CambridgeAddison-Wesley

Amir Yehuda 1969 ldquoContact Hypothesis in Ethnic RelationsrdquoPsychological Bulletin 71 319ndash42

Amir Yehuda 1976 ldquoThe Role of Intergroup Contact in Changeof Prejudice and Ethnic Relationsrdquo In Toward the Eliminationof Racism ed Phyllis A Katz New York Pergamonpp 245ndash308

Assis Claudia and Julian Pecquet 2002 ldquoHispanicsrsquo Search for aBetter Life Pushes Durham into Povertyrdquo The Herald Sun 25September A12

Bonilla-Silva Eduardo 2004 ldquoFrom Bi-Racial to Tri-RacialTowards a New System of Racial Stratification in the USArdquoEthnic and Racial Studies 27 (6) 931ndash50

Ciscel David H Barbara Ellen Smith and Marcela Mendoza2003 ldquoGhosts in the Global Machine New Immigrants and theRedefinition of Workrdquo Journal of Economic Issues 37 (2)333ndash41

Clark Juan M 1991 ldquoThe Social Impact of Cuban Immigrationin Floridardquo In Cuban Exiles in Florida Their Presence and Con-tributions eds Antonio Jorge Jamie Suchlicki and AdolfoLeyva de Varona University of Miami Research Institute forCuban Studies pp 39ndash61

rsquo

Cobb James C and William Stueck eds 2005 Globalization andthe American South Athens University of Georgia Press

Davis Darren W and Ronald E Brown 2002 ldquoThe Antipathy ofBlack Nationalism Behavioral and Attitudinal Implications ofan African American Ideologyrdquo American Journal of PoliticalScience 46 (2) 239ndash53

Dawson Michael C 1994 Behind the Mule Race and Class inAfrican-American Politics Princeton Princeton UniversityPress

de la Cadena Marisol 2001 ldquoReconstructing Race RacismCulture and Mestizaje in Latin Americardquo NACLA Report on theAmericas 34 (6) 16ndash23

de la Fuente Alejandro 1995 ldquoRace and Inequality in Cuba1899ndash1981rdquo Journal of Contemporary History 30 (1) 131ndash69

de la Fuente Alejandro 1998 ldquoRace National Discourse and Politics in Cubardquo Latin American Perspectives 25 (3) 43ndash70

Dulitzky Ariel E 2005 ldquoA Region in Denial Racial Discrimina-tion and Racism in Latin Americardquo In Neither Enemies norFriends Latinos Blacks Afro-Latinos eds Anani Dzidzienyoand Suzanne Oboler New York Palgrave-Macmillan 39ndash59

Duchon Deborah A and Arthur D Murphy 2001 ldquoIntroduc-tion From Patrones and Caciques to Good Ole Boysrdquo In LatinoWorkers in the Contemporary South eds Arthur D MurphyColleen Blanchard and Jennifer A Hill Athens University ofGeorgia Press pp 1ndash9

Durand Jorge Douglas S Massey and Fernando Charvet 2000ldquoThe Changing Geography of Mexican Immigration to theUnited States 1910ndash1996rdquo Social Science Quarterly 81 (1)1ndash16

Ellison Christopher G and Daniel A Powers 1994 ldquoThe ContactHypothesis and Racial Attitudes Among Black AmericansrdquoSocial Science Quarterly 75 (2) 385ndash400

Fossett Mark A and K Jill Kiecolt 1989 ldquoThe Relative Size ofMinority Populations and White Racial Attitudesrdquo SocialScience Quarterly 70 (4) 820ndash35

Geipel John 1997 ldquoBrazilrsquos African Legacyrdquo History Today 47 (8)18ndash24

Gong Lum v Rice 1927 275 US 78

Grenier Guillermo J and Max Castro 2001 ldquoBlacks and Cubansin Miami The Negative Consequences of the Cuban Enclaveon Ethnic Relationsrdquo In Governing American Cities Inter-Ethnic Coalitions Competition and Conflict ed Michael Jones-Correa New York Russell Sage Foundation pp 137ndash57

Griffith David 1993 Jonesrsquos Minimal Low-Wage Labor in theUnited States Albany State University of New York Press

Guimaraes Antonio Sergio 2001 ldquoRace Class and Color BehindBrazilrsquos lsquoRacial Democracyrsquordquo NACLA Report on the Americas 34(6) 38ndash41

Hanchard Michael G 1994 Orpheus and Power The MovimentoNegro of Rio de Janeiro and Satildeo Paulo Brazil 1945ndash1988Princeton Princeton University Press

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2000 ldquo lsquoMakingCarpet by the Milersquo The Emergence of a Mexican ImmigrantCommunity in an Industrial Region of the USHistoric SouthrdquoSocial Science Quarterly 81 (1) 49ndash66

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2001 ldquoA New Desti-nation for an Old Migration Origins Trajectories and LaborMarket Incorporation of Latinos in Dalton Georgiardquo In LatinoWorkers in the Contemporary South eds Arthur D MurphyColleen Blanchard and Jennifer A Hill Athens University ofGeorgia Press pp 127ndash35

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2003 ldquoMexicanImmigrant Communities in the South and Social Capital The

Case of Dalton Georgiardquo Southern Rural Sociology 19 (1)20ndash45

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2005 ldquoAppalachiaMeets Aztlaacuten Mexican Immigration and Intergroup Relationsin Dalton Georgiardquo In New Destinations Mexican Immigrationin the United States eds Rubeacuten Hernaacutendez-Leoacuten and VictorZuntildeiga New York Russell Sage Foundation pp 244ndash73

Hood M W III and Irwin L Morris 1998 ldquoGive Us Your TiredYour Poor But Make Sure They Have a Green Card TheEffects of Documented and Undocumented Migrant Contexton Anglo Opinion Toward Immigrationrdquo Political Behavior 20(1) 1ndash15

Jackman Mary R and Marie Crane 1986 ldquorsquoSome of my BestFriends are Black rsquo Interracial Friendship and WhitesrsquoRacial Attitudesrdquo Public Opinion Quarterly 50 (4) 459ndash86

Kandel William and Emilio A Parrado 2004 ldquoHispanics in theAmerican South and the Transformation of the Poultry Indus-tryrdquo In Hispanic Spaces Latino Places Community and CulturalDiversity in Contemporary America ed Daniel D ArreolaAustin University of Texas Press pp 255ndash76

Kaufmann Karen M 2003 ldquoCracks in the Rainbow Group Commonality as a Basis for Latino and African-AmericanPolitical Coalitionsrdquo Political Research Quarterly 56 (2)199ndash210

Key V O 1949 Southern Politics New York Vintage Books

Kinder Donald R and Tali Mendelberg 1995 ldquoCracks in Ameri-can Apartheid The Political Impact of Prejudice AmongDesegregated Whitesrdquo The Journal of Politics 57 (2) 402ndash24

King Gary Michael Tomz and Jason Wittenberg 2000 ldquoMakingthe Most of Statistical Analyses Improving Interpretation andPresentationrdquo American Journal of Political Science 44 (2)347ndash61

Loewen James W 1971 The Mississippi Chinese Between Blackand White Cambridge Harvard University Press

Massey Douglas Jorge Durand and Nolan J Malone 2002Beyond Smoke and Mirrors Mexican Immigration in an Era ofEconomic Integration New York Russell Sage Foundation

Matthews Donald R and James W Prothro 1967 Negroes and theNew Southern Politics New York Harcourt Brace amp World Inc

McClain Paula D Monique L Lyle Niambi M Carter Gerald FLackey Kendra Davenport Cotton Shayla C NunnallyThomas J Scotto Jeffrey D Grynaviski and J Alan Kendrick2005 ldquoBlack Americans and Latino Immigrants in a SouthernCity Friendly Neighbors or Economic Competitorsrdquo Type-script Duke University

Menchaca Martha 2001 Recovering History Constructing RaceThe Indian Black and White Roots of Mexican AmericansAustin University of Texas Press

Mindiola Jr Tatcho Yolanda Flores Niemann and NestorRodriguez 2002 Black-Brown Relations and StereotypesAustin University of Texas Press

Mohl Raymond A 1990 ldquoOn the Edge Blacks and Hispanics inMetropolitan Miami since 1959rdquo Florida Historical Quarterly69 (1) 37ndash57

Mohl Raymond A 2003 ldquoGlobalization Latinization and theNuevo New Southrdquo Journal of American Ethnic History 22 (4)31ndash66

Moumlrner Magnus ed 1967 Race Mixture in the History of LatinAmerica Boston Little Brown and Company

Morris Alan 1999 ldquoRace Relations and Racism in a RaciallyDiverse Inner City Neighbourhood A Case Study of HillbrowJohannesburgrdquo Journal of Southern African Studies 25 (4)667ndash94

Murphy Arthur D Colleen Blanchard and Jennifer A Hill 2001Latino Workers in the Contemporary South Athens Universityof Georgia Press

Oliver J Eric and Janelle Wong 2003 ldquoInter-group Prejudice inMultiethnic Settingsrdquo American Journal of Political Science 47(4) 567ndash82

Peacock James L Harry L Watson and Carrie R Matthews ed2005 The American South in a Global World Chapel Hill Uni-versity of North Carolina Press

Powers Daniel and Christopher Ellison 1994 ldquoInterracialContact and Black Racial Attitudesrdquo Social Forces 74 (1)205ndash27

Quan Robert Seto 1982 Lotus Among the Magnolias The Missis-sippi Chinese Jackson University Press of Mississippi

Quillian Lincoln 1996 ldquoGroup Threat and Regional Change inAttitudes Towards African Americansrdquo American Journal ofSociology 102 (3) 816ndash60

Rice et al v Gong Lum et al 1925 139 Miss 760 104 So 105

Scott Rebecca J 1985 Slave Emancipation in Cuba The Transitionto Free Labor 1860ndash1899 Princeton Princeton UniversityPress

Seed Patricia 1982 ldquoSocial Dimensions of Race Mexico City1753rdquo The Hispanic American Historical Review 62 (4)569ndash606

Shingles Richard D 1981 ldquoBlack Consciousness and Political Par-ticipation The Missing Linkrdquo American Political Science Review75 (2) 76ndash90

Sigelman Lee and Susan Welch 1993 ldquoThe Contact HypothesisRevisited Black-White Interaction and Positive Racial Atti-tudesrdquo Social Forces 71 (3) 781ndash95

Stephan Walter G and Cookie White Stephan 1985 ldquoIntergroupAnxietyrdquo The Journal of Social Issues 41 (3) 157ndash75

Suro Roberto and Audrey Singer 2002 Latino Growth in Metro-politan America Changing Patterns New Locations Washing-ton Brookings Institution Center on Urban and MetropolitanPolicy and The Pew Hispanic Center

Sweet James H 1997 ldquoThe Iberian Roots of American RacistThoughtrdquo The William and Mary Quarterly 54 (1) 143ndash66

Tate Katherine 1993 From Protest to Politics The New Black Votersin American Elections (enlarged edition) Cambridge HarvardUniversity Press and Russell Sage Foundation

Taylor Marylee C 1998 ldquoHow White Attitudes Vary with theRacial Composition of Local Populations Numbers CountrdquoAmerican Sociological Review 63 (4) 512ndash35

Tomz Michael Jason Wittenberg and Gary King 2003 CLARIFYSoftware for Interpreting and Presenting Statistical ResultsVersion 21 Stanford University University of Wisconsinand Harvard University httpgkingharvardedustatsshtml(August 9 2003)

Torres Cruz C 2000 ldquoEmerging Latino Communities A NewChallenge for the Rural Southrdquo The Southern Rural Develop-ment Center No 12 (August) 1ndash8

US Census Bureau 2002a Census of PopulationhttpfactfindercensusgovservlettQTTable (August 132002)

US Census Bureau 2002b Redistricting File httpcenstatscensusgovdataNC160379000pdf (August 132002)

US Census Bureau 2003 US Interim Projections httpwwwcensusgovusinterimproj (September 25 2003)

Wade Peter 1993 Blackness and Race Mixture The Dynamics ofRacial Identity in Colombia Baltimore The Johns HopkinsUniversity Press

Wade Peter 1997 Race and Ethnicity in Latin America LondonPluto Press

Welch Susan Lee Sigelman Timothy Bledsoe and MichaelCombs 2001 Race and Place Race Relations in an AmericanCity New York Cambridge University Press

Winant Howard 1992 ldquoRethinking Race in Brazilrdquo Journal ofLatin American Studies 24 (1) 173ndash92

Paula D McClain is professor of political scienceDuke University Durham NC 27708-0204 NiambiM Carter is PhD candidate in political science DukeUniversity Durham NC 27708-0204 Victoria MDeFrancesco Soto is PhD candidate in politicalscience Duke University Durham NC 27708-0204Monique L Lyle is PhD candidate in political scienceDuke University Durham NC 27708-0204 Jeffrey DGrynaviski is assistant professor of political scienceUniversity of Chicago Chicago IL 60637 Shayla CNunnally is assistant professor of political science andInstitute of African American Studies University ofConnecticut Storrs CT 06269-1024 Thomas J Scottois assistant professor of political science West VirginiaUniversity Morganstown WV 26506 J Alan Kendrickis assistant professor of political science St Augus-tinersquos College Raleigh NC 27610 Gerald F Lackey isgraduate student in sociology University of NorthCarolina-Chapel Hill Chapel Hill NC 27599-3210Kendra Davenport Cotton is PhD candidate in polit-ical science University of North Carolina-Chapel HillChapel Hill NC 27599-3210

great extent Therefore this means Latino immigrantshave no reference point for black Americans otherthan their own attitudes which might have beenformed prior to their arrival in the United States andwhich do not appear to dissipate with increased inter-action with blacks Yet the finding that these negativeattitudes are modulated by a sense of linked fate sug-gests possibilities for the formation of connections toblack Americans in the absence of the presence of anextant American Latino community

Whatrsquos more these findings are importantbecause the South still contains the largest populationof black Americans in the United States and is con-sidered their ldquoregional homelandrdquo Moreover theSouth has suffered through some of the most politi-cized battles over race relations in recent history Noother section of the country has been as rigidlydefined along the black-white racial divide as has theSouth While we did not test directly Keyrsquos assertionsof over 50 years ago of the continued salience of racein the South the black-white divide in the South iscertain to shape the attitudes and incentives of Latinosin the region and how these new Latino immigrantssituate themselves vis-agrave-vis black Americans has pro-found implications for the social and political fabricof the South

What do these findings mean for Southern poli-tics and the politics of race We must be cautious indrawing broad conclusions about the future politics ofthe South from a one-city study with a small surveysample Yet our findings are suggestive of possible pat-terns that might be exhibited in other parts of theSouth and implicative of future trends

Latino immigrantsrsquo negative views of black Amer-icans most likely brought with them from their homecountries and reinforced rather than reduced byneighborhood interactions with blacks suggest thatthese new Latino immigrants may behave in wayssimilar to the Chinese in Mississippi in the mid-nineteenth century and the Cubans in Miami in themid-twentieth centurymdashidentification with whitesdistancing themselves from blacks and feeling noresponsibility to rectify the continuing inequalities of black Americans Given the increasing number ofLatino immigrants in the South and the possibilitythat over time their numbers might rival or evensurpass black Americans in the region if large por-tions of Latino immigrants maintain negative atti-tudes of black Americans where will this leave blacksWill blacks find that they must not only makedemands on whites for continued progress but alsomount a fight on another front against Latinos Orwill Latino immigrants begin to see themselves as

closer to blacks the longer they reside in the UnitedStates as some scholars suggest Clearly in order toanswer this question we need to do similar work in aseries of Southern locations But based on our resultsin this paper which are by no means definitive but arehighly suggestive we fear the former rather than thelatter is the future of the politics of race in the South

Acknowledgment

This is a revised version of a paper presented at theannual meeting of the Midwest Political Science Asso-ciation Chicago IL April 14ndash18 2004 The surveyreported in this paper was funded by a grant from TheFord Foundation (St Benedict the Black Meets theVirgin of Guadalupe Project Grant 1025-1445) Wethank the Foundation and Dr Melvin Oliver formerVice President of Asset Building and CommunityDevelopment now Dean of Social Science at the University of California Santa Barbara for theirsupport of our research We would also like to thankthe three anonymous reviewers whose commentspushed us to think more critically about this impor-tant topic We are very appreciative of their efforts onour behalf

Manuscript submitted 9 May 2005Manuscript accepted for publication 18 October 2005

References

Allport Gordon W 1954 The Nature of Prejudice CambridgeAddison-Wesley

Amir Yehuda 1969 ldquoContact Hypothesis in Ethnic RelationsrdquoPsychological Bulletin 71 319ndash42

Amir Yehuda 1976 ldquoThe Role of Intergroup Contact in Changeof Prejudice and Ethnic Relationsrdquo In Toward the Eliminationof Racism ed Phyllis A Katz New York Pergamonpp 245ndash308

Assis Claudia and Julian Pecquet 2002 ldquoHispanicsrsquo Search for aBetter Life Pushes Durham into Povertyrdquo The Herald Sun 25September A12

Bonilla-Silva Eduardo 2004 ldquoFrom Bi-Racial to Tri-RacialTowards a New System of Racial Stratification in the USArdquoEthnic and Racial Studies 27 (6) 931ndash50

Ciscel David H Barbara Ellen Smith and Marcela Mendoza2003 ldquoGhosts in the Global Machine New Immigrants and theRedefinition of Workrdquo Journal of Economic Issues 37 (2)333ndash41

Clark Juan M 1991 ldquoThe Social Impact of Cuban Immigrationin Floridardquo In Cuban Exiles in Florida Their Presence and Con-tributions eds Antonio Jorge Jamie Suchlicki and AdolfoLeyva de Varona University of Miami Research Institute forCuban Studies pp 39ndash61

rsquo

Cobb James C and William Stueck eds 2005 Globalization andthe American South Athens University of Georgia Press

Davis Darren W and Ronald E Brown 2002 ldquoThe Antipathy ofBlack Nationalism Behavioral and Attitudinal Implications ofan African American Ideologyrdquo American Journal of PoliticalScience 46 (2) 239ndash53

Dawson Michael C 1994 Behind the Mule Race and Class inAfrican-American Politics Princeton Princeton UniversityPress

de la Cadena Marisol 2001 ldquoReconstructing Race RacismCulture and Mestizaje in Latin Americardquo NACLA Report on theAmericas 34 (6) 16ndash23

de la Fuente Alejandro 1995 ldquoRace and Inequality in Cuba1899ndash1981rdquo Journal of Contemporary History 30 (1) 131ndash69

de la Fuente Alejandro 1998 ldquoRace National Discourse and Politics in Cubardquo Latin American Perspectives 25 (3) 43ndash70

Dulitzky Ariel E 2005 ldquoA Region in Denial Racial Discrimina-tion and Racism in Latin Americardquo In Neither Enemies norFriends Latinos Blacks Afro-Latinos eds Anani Dzidzienyoand Suzanne Oboler New York Palgrave-Macmillan 39ndash59

Duchon Deborah A and Arthur D Murphy 2001 ldquoIntroduc-tion From Patrones and Caciques to Good Ole Boysrdquo In LatinoWorkers in the Contemporary South eds Arthur D MurphyColleen Blanchard and Jennifer A Hill Athens University ofGeorgia Press pp 1ndash9

Durand Jorge Douglas S Massey and Fernando Charvet 2000ldquoThe Changing Geography of Mexican Immigration to theUnited States 1910ndash1996rdquo Social Science Quarterly 81 (1)1ndash16

Ellison Christopher G and Daniel A Powers 1994 ldquoThe ContactHypothesis and Racial Attitudes Among Black AmericansrdquoSocial Science Quarterly 75 (2) 385ndash400

Fossett Mark A and K Jill Kiecolt 1989 ldquoThe Relative Size ofMinority Populations and White Racial Attitudesrdquo SocialScience Quarterly 70 (4) 820ndash35

Geipel John 1997 ldquoBrazilrsquos African Legacyrdquo History Today 47 (8)18ndash24

Gong Lum v Rice 1927 275 US 78

Grenier Guillermo J and Max Castro 2001 ldquoBlacks and Cubansin Miami The Negative Consequences of the Cuban Enclaveon Ethnic Relationsrdquo In Governing American Cities Inter-Ethnic Coalitions Competition and Conflict ed Michael Jones-Correa New York Russell Sage Foundation pp 137ndash57

Griffith David 1993 Jonesrsquos Minimal Low-Wage Labor in theUnited States Albany State University of New York Press

Guimaraes Antonio Sergio 2001 ldquoRace Class and Color BehindBrazilrsquos lsquoRacial Democracyrsquordquo NACLA Report on the Americas 34(6) 38ndash41

Hanchard Michael G 1994 Orpheus and Power The MovimentoNegro of Rio de Janeiro and Satildeo Paulo Brazil 1945ndash1988Princeton Princeton University Press

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2000 ldquo lsquoMakingCarpet by the Milersquo The Emergence of a Mexican ImmigrantCommunity in an Industrial Region of the USHistoric SouthrdquoSocial Science Quarterly 81 (1) 49ndash66

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2001 ldquoA New Desti-nation for an Old Migration Origins Trajectories and LaborMarket Incorporation of Latinos in Dalton Georgiardquo In LatinoWorkers in the Contemporary South eds Arthur D MurphyColleen Blanchard and Jennifer A Hill Athens University ofGeorgia Press pp 127ndash35

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2003 ldquoMexicanImmigrant Communities in the South and Social Capital The

Case of Dalton Georgiardquo Southern Rural Sociology 19 (1)20ndash45

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2005 ldquoAppalachiaMeets Aztlaacuten Mexican Immigration and Intergroup Relationsin Dalton Georgiardquo In New Destinations Mexican Immigrationin the United States eds Rubeacuten Hernaacutendez-Leoacuten and VictorZuntildeiga New York Russell Sage Foundation pp 244ndash73

Hood M W III and Irwin L Morris 1998 ldquoGive Us Your TiredYour Poor But Make Sure They Have a Green Card TheEffects of Documented and Undocumented Migrant Contexton Anglo Opinion Toward Immigrationrdquo Political Behavior 20(1) 1ndash15

Jackman Mary R and Marie Crane 1986 ldquorsquoSome of my BestFriends are Black rsquo Interracial Friendship and WhitesrsquoRacial Attitudesrdquo Public Opinion Quarterly 50 (4) 459ndash86

Kandel William and Emilio A Parrado 2004 ldquoHispanics in theAmerican South and the Transformation of the Poultry Indus-tryrdquo In Hispanic Spaces Latino Places Community and CulturalDiversity in Contemporary America ed Daniel D ArreolaAustin University of Texas Press pp 255ndash76

Kaufmann Karen M 2003 ldquoCracks in the Rainbow Group Commonality as a Basis for Latino and African-AmericanPolitical Coalitionsrdquo Political Research Quarterly 56 (2)199ndash210

Key V O 1949 Southern Politics New York Vintage Books

Kinder Donald R and Tali Mendelberg 1995 ldquoCracks in Ameri-can Apartheid The Political Impact of Prejudice AmongDesegregated Whitesrdquo The Journal of Politics 57 (2) 402ndash24

King Gary Michael Tomz and Jason Wittenberg 2000 ldquoMakingthe Most of Statistical Analyses Improving Interpretation andPresentationrdquo American Journal of Political Science 44 (2)347ndash61

Loewen James W 1971 The Mississippi Chinese Between Blackand White Cambridge Harvard University Press

Massey Douglas Jorge Durand and Nolan J Malone 2002Beyond Smoke and Mirrors Mexican Immigration in an Era ofEconomic Integration New York Russell Sage Foundation

Matthews Donald R and James W Prothro 1967 Negroes and theNew Southern Politics New York Harcourt Brace amp World Inc

McClain Paula D Monique L Lyle Niambi M Carter Gerald FLackey Kendra Davenport Cotton Shayla C NunnallyThomas J Scotto Jeffrey D Grynaviski and J Alan Kendrick2005 ldquoBlack Americans and Latino Immigrants in a SouthernCity Friendly Neighbors or Economic Competitorsrdquo Type-script Duke University

Menchaca Martha 2001 Recovering History Constructing RaceThe Indian Black and White Roots of Mexican AmericansAustin University of Texas Press

Mindiola Jr Tatcho Yolanda Flores Niemann and NestorRodriguez 2002 Black-Brown Relations and StereotypesAustin University of Texas Press

Mohl Raymond A 1990 ldquoOn the Edge Blacks and Hispanics inMetropolitan Miami since 1959rdquo Florida Historical Quarterly69 (1) 37ndash57

Mohl Raymond A 2003 ldquoGlobalization Latinization and theNuevo New Southrdquo Journal of American Ethnic History 22 (4)31ndash66

Moumlrner Magnus ed 1967 Race Mixture in the History of LatinAmerica Boston Little Brown and Company

Morris Alan 1999 ldquoRace Relations and Racism in a RaciallyDiverse Inner City Neighbourhood A Case Study of HillbrowJohannesburgrdquo Journal of Southern African Studies 25 (4)667ndash94

Murphy Arthur D Colleen Blanchard and Jennifer A Hill 2001Latino Workers in the Contemporary South Athens Universityof Georgia Press

Oliver J Eric and Janelle Wong 2003 ldquoInter-group Prejudice inMultiethnic Settingsrdquo American Journal of Political Science 47(4) 567ndash82

Peacock James L Harry L Watson and Carrie R Matthews ed2005 The American South in a Global World Chapel Hill Uni-versity of North Carolina Press

Powers Daniel and Christopher Ellison 1994 ldquoInterracialContact and Black Racial Attitudesrdquo Social Forces 74 (1)205ndash27

Quan Robert Seto 1982 Lotus Among the Magnolias The Missis-sippi Chinese Jackson University Press of Mississippi

Quillian Lincoln 1996 ldquoGroup Threat and Regional Change inAttitudes Towards African Americansrdquo American Journal ofSociology 102 (3) 816ndash60

Rice et al v Gong Lum et al 1925 139 Miss 760 104 So 105

Scott Rebecca J 1985 Slave Emancipation in Cuba The Transitionto Free Labor 1860ndash1899 Princeton Princeton UniversityPress

Seed Patricia 1982 ldquoSocial Dimensions of Race Mexico City1753rdquo The Hispanic American Historical Review 62 (4)569ndash606

Shingles Richard D 1981 ldquoBlack Consciousness and Political Par-ticipation The Missing Linkrdquo American Political Science Review75 (2) 76ndash90

Sigelman Lee and Susan Welch 1993 ldquoThe Contact HypothesisRevisited Black-White Interaction and Positive Racial Atti-tudesrdquo Social Forces 71 (3) 781ndash95

Stephan Walter G and Cookie White Stephan 1985 ldquoIntergroupAnxietyrdquo The Journal of Social Issues 41 (3) 157ndash75

Suro Roberto and Audrey Singer 2002 Latino Growth in Metro-politan America Changing Patterns New Locations Washing-ton Brookings Institution Center on Urban and MetropolitanPolicy and The Pew Hispanic Center

Sweet James H 1997 ldquoThe Iberian Roots of American RacistThoughtrdquo The William and Mary Quarterly 54 (1) 143ndash66

Tate Katherine 1993 From Protest to Politics The New Black Votersin American Elections (enlarged edition) Cambridge HarvardUniversity Press and Russell Sage Foundation

Taylor Marylee C 1998 ldquoHow White Attitudes Vary with theRacial Composition of Local Populations Numbers CountrdquoAmerican Sociological Review 63 (4) 512ndash35

Tomz Michael Jason Wittenberg and Gary King 2003 CLARIFYSoftware for Interpreting and Presenting Statistical ResultsVersion 21 Stanford University University of Wisconsinand Harvard University httpgkingharvardedustatsshtml(August 9 2003)

Torres Cruz C 2000 ldquoEmerging Latino Communities A NewChallenge for the Rural Southrdquo The Southern Rural Develop-ment Center No 12 (August) 1ndash8

US Census Bureau 2002a Census of PopulationhttpfactfindercensusgovservlettQTTable (August 132002)

US Census Bureau 2002b Redistricting File httpcenstatscensusgovdataNC160379000pdf (August 132002)

US Census Bureau 2003 US Interim Projections httpwwwcensusgovusinterimproj (September 25 2003)

Wade Peter 1993 Blackness and Race Mixture The Dynamics ofRacial Identity in Colombia Baltimore The Johns HopkinsUniversity Press

Wade Peter 1997 Race and Ethnicity in Latin America LondonPluto Press

Welch Susan Lee Sigelman Timothy Bledsoe and MichaelCombs 2001 Race and Place Race Relations in an AmericanCity New York Cambridge University Press

Winant Howard 1992 ldquoRethinking Race in Brazilrdquo Journal ofLatin American Studies 24 (1) 173ndash92

Paula D McClain is professor of political scienceDuke University Durham NC 27708-0204 NiambiM Carter is PhD candidate in political science DukeUniversity Durham NC 27708-0204 Victoria MDeFrancesco Soto is PhD candidate in politicalscience Duke University Durham NC 27708-0204Monique L Lyle is PhD candidate in political scienceDuke University Durham NC 27708-0204 Jeffrey DGrynaviski is assistant professor of political scienceUniversity of Chicago Chicago IL 60637 Shayla CNunnally is assistant professor of political science andInstitute of African American Studies University ofConnecticut Storrs CT 06269-1024 Thomas J Scottois assistant professor of political science West VirginiaUniversity Morganstown WV 26506 J Alan Kendrickis assistant professor of political science St Augus-tinersquos College Raleigh NC 27610 Gerald F Lackey isgraduate student in sociology University of NorthCarolina-Chapel Hill Chapel Hill NC 27599-3210Kendra Davenport Cotton is PhD candidate in polit-ical science University of North Carolina-Chapel HillChapel Hill NC 27599-3210

rsquo

Cobb James C and William Stueck eds 2005 Globalization andthe American South Athens University of Georgia Press

Davis Darren W and Ronald E Brown 2002 ldquoThe Antipathy ofBlack Nationalism Behavioral and Attitudinal Implications ofan African American Ideologyrdquo American Journal of PoliticalScience 46 (2) 239ndash53

Dawson Michael C 1994 Behind the Mule Race and Class inAfrican-American Politics Princeton Princeton UniversityPress

de la Cadena Marisol 2001 ldquoReconstructing Race RacismCulture and Mestizaje in Latin Americardquo NACLA Report on theAmericas 34 (6) 16ndash23

de la Fuente Alejandro 1995 ldquoRace and Inequality in Cuba1899ndash1981rdquo Journal of Contemporary History 30 (1) 131ndash69

de la Fuente Alejandro 1998 ldquoRace National Discourse and Politics in Cubardquo Latin American Perspectives 25 (3) 43ndash70

Dulitzky Ariel E 2005 ldquoA Region in Denial Racial Discrimina-tion and Racism in Latin Americardquo In Neither Enemies norFriends Latinos Blacks Afro-Latinos eds Anani Dzidzienyoand Suzanne Oboler New York Palgrave-Macmillan 39ndash59

Duchon Deborah A and Arthur D Murphy 2001 ldquoIntroduc-tion From Patrones and Caciques to Good Ole Boysrdquo In LatinoWorkers in the Contemporary South eds Arthur D MurphyColleen Blanchard and Jennifer A Hill Athens University ofGeorgia Press pp 1ndash9

Durand Jorge Douglas S Massey and Fernando Charvet 2000ldquoThe Changing Geography of Mexican Immigration to theUnited States 1910ndash1996rdquo Social Science Quarterly 81 (1)1ndash16

Ellison Christopher G and Daniel A Powers 1994 ldquoThe ContactHypothesis and Racial Attitudes Among Black AmericansrdquoSocial Science Quarterly 75 (2) 385ndash400

Fossett Mark A and K Jill Kiecolt 1989 ldquoThe Relative Size ofMinority Populations and White Racial Attitudesrdquo SocialScience Quarterly 70 (4) 820ndash35

Geipel John 1997 ldquoBrazilrsquos African Legacyrdquo History Today 47 (8)18ndash24

Gong Lum v Rice 1927 275 US 78

Grenier Guillermo J and Max Castro 2001 ldquoBlacks and Cubansin Miami The Negative Consequences of the Cuban Enclaveon Ethnic Relationsrdquo In Governing American Cities Inter-Ethnic Coalitions Competition and Conflict ed Michael Jones-Correa New York Russell Sage Foundation pp 137ndash57

Griffith David 1993 Jonesrsquos Minimal Low-Wage Labor in theUnited States Albany State University of New York Press

Guimaraes Antonio Sergio 2001 ldquoRace Class and Color BehindBrazilrsquos lsquoRacial Democracyrsquordquo NACLA Report on the Americas 34(6) 38ndash41

Hanchard Michael G 1994 Orpheus and Power The MovimentoNegro of Rio de Janeiro and Satildeo Paulo Brazil 1945ndash1988Princeton Princeton University Press

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2000 ldquo lsquoMakingCarpet by the Milersquo The Emergence of a Mexican ImmigrantCommunity in an Industrial Region of the USHistoric SouthrdquoSocial Science Quarterly 81 (1) 49ndash66

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2001 ldquoA New Desti-nation for an Old Migration Origins Trajectories and LaborMarket Incorporation of Latinos in Dalton Georgiardquo In LatinoWorkers in the Contemporary South eds Arthur D MurphyColleen Blanchard and Jennifer A Hill Athens University ofGeorgia Press pp 127ndash35

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2003 ldquoMexicanImmigrant Communities in the South and Social Capital The

Case of Dalton Georgiardquo Southern Rural Sociology 19 (1)20ndash45

Hernaacutendez-Leoacuten Rubeacuten and Victor Zuntildeiga 2005 ldquoAppalachiaMeets Aztlaacuten Mexican Immigration and Intergroup Relationsin Dalton Georgiardquo In New Destinations Mexican Immigrationin the United States eds Rubeacuten Hernaacutendez-Leoacuten and VictorZuntildeiga New York Russell Sage Foundation pp 244ndash73

Hood M W III and Irwin L Morris 1998 ldquoGive Us Your TiredYour Poor But Make Sure They Have a Green Card TheEffects of Documented and Undocumented Migrant Contexton Anglo Opinion Toward Immigrationrdquo Political Behavior 20(1) 1ndash15

Jackman Mary R and Marie Crane 1986 ldquorsquoSome of my BestFriends are Black rsquo Interracial Friendship and WhitesrsquoRacial Attitudesrdquo Public Opinion Quarterly 50 (4) 459ndash86

Kandel William and Emilio A Parrado 2004 ldquoHispanics in theAmerican South and the Transformation of the Poultry Indus-tryrdquo In Hispanic Spaces Latino Places Community and CulturalDiversity in Contemporary America ed Daniel D ArreolaAustin University of Texas Press pp 255ndash76

Kaufmann Karen M 2003 ldquoCracks in the Rainbow Group Commonality as a Basis for Latino and African-AmericanPolitical Coalitionsrdquo Political Research Quarterly 56 (2)199ndash210

Key V O 1949 Southern Politics New York Vintage Books

Kinder Donald R and Tali Mendelberg 1995 ldquoCracks in Ameri-can Apartheid The Political Impact of Prejudice AmongDesegregated Whitesrdquo The Journal of Politics 57 (2) 402ndash24

King Gary Michael Tomz and Jason Wittenberg 2000 ldquoMakingthe Most of Statistical Analyses Improving Interpretation andPresentationrdquo American Journal of Political Science 44 (2)347ndash61

Loewen James W 1971 The Mississippi Chinese Between Blackand White Cambridge Harvard University Press

Massey Douglas Jorge Durand and Nolan J Malone 2002Beyond Smoke and Mirrors Mexican Immigration in an Era ofEconomic Integration New York Russell Sage Foundation

Matthews Donald R and James W Prothro 1967 Negroes and theNew Southern Politics New York Harcourt Brace amp World Inc

McClain Paula D Monique L Lyle Niambi M Carter Gerald FLackey Kendra Davenport Cotton Shayla C NunnallyThomas J Scotto Jeffrey D Grynaviski and J Alan Kendrick2005 ldquoBlack Americans and Latino Immigrants in a SouthernCity Friendly Neighbors or Economic Competitorsrdquo Type-script Duke University

Menchaca Martha 2001 Recovering History Constructing RaceThe Indian Black and White Roots of Mexican AmericansAustin University of Texas Press

Mindiola Jr Tatcho Yolanda Flores Niemann and NestorRodriguez 2002 Black-Brown Relations and StereotypesAustin University of Texas Press

Mohl Raymond A 1990 ldquoOn the Edge Blacks and Hispanics inMetropolitan Miami since 1959rdquo Florida Historical Quarterly69 (1) 37ndash57

Mohl Raymond A 2003 ldquoGlobalization Latinization and theNuevo New Southrdquo Journal of American Ethnic History 22 (4)31ndash66

Moumlrner Magnus ed 1967 Race Mixture in the History of LatinAmerica Boston Little Brown and Company

Morris Alan 1999 ldquoRace Relations and Racism in a RaciallyDiverse Inner City Neighbourhood A Case Study of HillbrowJohannesburgrdquo Journal of Southern African Studies 25 (4)667ndash94

Murphy Arthur D Colleen Blanchard and Jennifer A Hill 2001Latino Workers in the Contemporary South Athens Universityof Georgia Press

Oliver J Eric and Janelle Wong 2003 ldquoInter-group Prejudice inMultiethnic Settingsrdquo American Journal of Political Science 47(4) 567ndash82

Peacock James L Harry L Watson and Carrie R Matthews ed2005 The American South in a Global World Chapel Hill Uni-versity of North Carolina Press

Powers Daniel and Christopher Ellison 1994 ldquoInterracialContact and Black Racial Attitudesrdquo Social Forces 74 (1)205ndash27

Quan Robert Seto 1982 Lotus Among the Magnolias The Missis-sippi Chinese Jackson University Press of Mississippi

Quillian Lincoln 1996 ldquoGroup Threat and Regional Change inAttitudes Towards African Americansrdquo American Journal ofSociology 102 (3) 816ndash60

Rice et al v Gong Lum et al 1925 139 Miss 760 104 So 105

Scott Rebecca J 1985 Slave Emancipation in Cuba The Transitionto Free Labor 1860ndash1899 Princeton Princeton UniversityPress

Seed Patricia 1982 ldquoSocial Dimensions of Race Mexico City1753rdquo The Hispanic American Historical Review 62 (4)569ndash606

Shingles Richard D 1981 ldquoBlack Consciousness and Political Par-ticipation The Missing Linkrdquo American Political Science Review75 (2) 76ndash90

Sigelman Lee and Susan Welch 1993 ldquoThe Contact HypothesisRevisited Black-White Interaction and Positive Racial Atti-tudesrdquo Social Forces 71 (3) 781ndash95

Stephan Walter G and Cookie White Stephan 1985 ldquoIntergroupAnxietyrdquo The Journal of Social Issues 41 (3) 157ndash75

Suro Roberto and Audrey Singer 2002 Latino Growth in Metro-politan America Changing Patterns New Locations Washing-ton Brookings Institution Center on Urban and MetropolitanPolicy and The Pew Hispanic Center

Sweet James H 1997 ldquoThe Iberian Roots of American RacistThoughtrdquo The William and Mary Quarterly 54 (1) 143ndash66

Tate Katherine 1993 From Protest to Politics The New Black Votersin American Elections (enlarged edition) Cambridge HarvardUniversity Press and Russell Sage Foundation

Taylor Marylee C 1998 ldquoHow White Attitudes Vary with theRacial Composition of Local Populations Numbers CountrdquoAmerican Sociological Review 63 (4) 512ndash35

Tomz Michael Jason Wittenberg and Gary King 2003 CLARIFYSoftware for Interpreting and Presenting Statistical ResultsVersion 21 Stanford University University of Wisconsinand Harvard University httpgkingharvardedustatsshtml(August 9 2003)

Torres Cruz C 2000 ldquoEmerging Latino Communities A NewChallenge for the Rural Southrdquo The Southern Rural Develop-ment Center No 12 (August) 1ndash8

US Census Bureau 2002a Census of PopulationhttpfactfindercensusgovservlettQTTable (August 132002)

US Census Bureau 2002b Redistricting File httpcenstatscensusgovdataNC160379000pdf (August 132002)

US Census Bureau 2003 US Interim Projections httpwwwcensusgovusinterimproj (September 25 2003)

Wade Peter 1993 Blackness and Race Mixture The Dynamics ofRacial Identity in Colombia Baltimore The Johns HopkinsUniversity Press

Wade Peter 1997 Race and Ethnicity in Latin America LondonPluto Press

Welch Susan Lee Sigelman Timothy Bledsoe and MichaelCombs 2001 Race and Place Race Relations in an AmericanCity New York Cambridge University Press

Winant Howard 1992 ldquoRethinking Race in Brazilrdquo Journal ofLatin American Studies 24 (1) 173ndash92

Paula D McClain is professor of political scienceDuke University Durham NC 27708-0204 NiambiM Carter is PhD candidate in political science DukeUniversity Durham NC 27708-0204 Victoria MDeFrancesco Soto is PhD candidate in politicalscience Duke University Durham NC 27708-0204Monique L Lyle is PhD candidate in political scienceDuke University Durham NC 27708-0204 Jeffrey DGrynaviski is assistant professor of political scienceUniversity of Chicago Chicago IL 60637 Shayla CNunnally is assistant professor of political science andInstitute of African American Studies University ofConnecticut Storrs CT 06269-1024 Thomas J Scottois assistant professor of political science West VirginiaUniversity Morganstown WV 26506 J Alan Kendrickis assistant professor of political science St Augus-tinersquos College Raleigh NC 27610 Gerald F Lackey isgraduate student in sociology University of NorthCarolina-Chapel Hill Chapel Hill NC 27599-3210Kendra Davenport Cotton is PhD candidate in polit-ical science University of North Carolina-Chapel HillChapel Hill NC 27599-3210

Murphy Arthur D Colleen Blanchard and Jennifer A Hill 2001Latino Workers in the Contemporary South Athens Universityof Georgia Press

Oliver J Eric and Janelle Wong 2003 ldquoInter-group Prejudice inMultiethnic Settingsrdquo American Journal of Political Science 47(4) 567ndash82

Peacock James L Harry L Watson and Carrie R Matthews ed2005 The American South in a Global World Chapel Hill Uni-versity of North Carolina Press

Powers Daniel and Christopher Ellison 1994 ldquoInterracialContact and Black Racial Attitudesrdquo Social Forces 74 (1)205ndash27

Quan Robert Seto 1982 Lotus Among the Magnolias The Missis-sippi Chinese Jackson University Press of Mississippi

Quillian Lincoln 1996 ldquoGroup Threat and Regional Change inAttitudes Towards African Americansrdquo American Journal ofSociology 102 (3) 816ndash60

Rice et al v Gong Lum et al 1925 139 Miss 760 104 So 105

Scott Rebecca J 1985 Slave Emancipation in Cuba The Transitionto Free Labor 1860ndash1899 Princeton Princeton UniversityPress

Seed Patricia 1982 ldquoSocial Dimensions of Race Mexico City1753rdquo The Hispanic American Historical Review 62 (4)569ndash606

Shingles Richard D 1981 ldquoBlack Consciousness and Political Par-ticipation The Missing Linkrdquo American Political Science Review75 (2) 76ndash90

Sigelman Lee and Susan Welch 1993 ldquoThe Contact HypothesisRevisited Black-White Interaction and Positive Racial Atti-tudesrdquo Social Forces 71 (3) 781ndash95

Stephan Walter G and Cookie White Stephan 1985 ldquoIntergroupAnxietyrdquo The Journal of Social Issues 41 (3) 157ndash75

Suro Roberto and Audrey Singer 2002 Latino Growth in Metro-politan America Changing Patterns New Locations Washing-ton Brookings Institution Center on Urban and MetropolitanPolicy and The Pew Hispanic Center

Sweet James H 1997 ldquoThe Iberian Roots of American RacistThoughtrdquo The William and Mary Quarterly 54 (1) 143ndash66

Tate Katherine 1993 From Protest to Politics The New Black Votersin American Elections (enlarged edition) Cambridge HarvardUniversity Press and Russell Sage Foundation

Taylor Marylee C 1998 ldquoHow White Attitudes Vary with theRacial Composition of Local Populations Numbers CountrdquoAmerican Sociological Review 63 (4) 512ndash35

Tomz Michael Jason Wittenberg and Gary King 2003 CLARIFYSoftware for Interpreting and Presenting Statistical ResultsVersion 21 Stanford University University of Wisconsinand Harvard University httpgkingharvardedustatsshtml(August 9 2003)

Torres Cruz C 2000 ldquoEmerging Latino Communities A NewChallenge for the Rural Southrdquo The Southern Rural Develop-ment Center No 12 (August) 1ndash8

US Census Bureau 2002a Census of PopulationhttpfactfindercensusgovservlettQTTable (August 132002)

US Census Bureau 2002b Redistricting File httpcenstatscensusgovdataNC160379000pdf (August 132002)

US Census Bureau 2003 US Interim Projections httpwwwcensusgovusinterimproj (September 25 2003)

Wade Peter 1993 Blackness and Race Mixture The Dynamics ofRacial Identity in Colombia Baltimore The Johns HopkinsUniversity Press

Wade Peter 1997 Race and Ethnicity in Latin America LondonPluto Press

Welch Susan Lee Sigelman Timothy Bledsoe and MichaelCombs 2001 Race and Place Race Relations in an AmericanCity New York Cambridge University Press

Winant Howard 1992 ldquoRethinking Race in Brazilrdquo Journal ofLatin American Studies 24 (1) 173ndash92

Paula D McClain is professor of political scienceDuke University Durham NC 27708-0204 NiambiM Carter is PhD candidate in political science DukeUniversity Durham NC 27708-0204 Victoria MDeFrancesco Soto is PhD candidate in politicalscience Duke University Durham NC 27708-0204Monique L Lyle is PhD candidate in political scienceDuke University Durham NC 27708-0204 Jeffrey DGrynaviski is assistant professor of political scienceUniversity of Chicago Chicago IL 60637 Shayla CNunnally is assistant professor of political science andInstitute of African American Studies University ofConnecticut Storrs CT 06269-1024 Thomas J Scottois assistant professor of political science West VirginiaUniversity Morganstown WV 26506 J Alan Kendrickis assistant professor of political science St Augus-tinersquos College Raleigh NC 27610 Gerald F Lackey isgraduate student in sociology University of NorthCarolina-Chapel Hill Chapel Hill NC 27599-3210Kendra Davenport Cotton is PhD candidate in polit-ical science University of North Carolina-Chapel HillChapel Hill NC 27599-3210