American Political Science Review, Page 1 of 15

doi:10.1017/S0003055418000175 © American Political Science Association 2018

What Makes a Good Neighbor? Race, Place, and Norms of PoliticalParticipationALLISON P. ANOLL Vanderbilt University

Social norms are thought to motivate behaviors like political participation, but context should influ-ence both the content and activation of these norms. I show that both race and neighborhood contextmoderate the social value of political participation in the United States. Using original survey data

and a survey experiment, I find that Whites, Blacks, and Latinos not only conceptualize participationdifferently, but also asymmetrically reward those who are politically active, with minority Americansoften providing more social incentives for participation thanWhites. I combine this survey data with geo-graphic demography from the American Community Survey and find that neighborhood characteristicsoutpace individual-level indicators in predicting the social value of political participation. The findingssuggest that scholars of political behavior should consider race, place, and social norms when seeking tounderstand participation in an increasingly diverse America.

INTRODUCTION

Political participation is costly and fraught withuncertain outcomes that often prevent it fromoccurring. Social rewards for engaging, though,

like acceptance, reverence, or even friendship, can off-set these challenges (Salisbury 1969; Gerber, Green,and Larimer 2008; Sinclair 2012). A growing literaturehas established that social norms—or the standards ofgroups that emerge out of social interactions (Cialdiniand Trost 1998)—affect political participation, but fewhave considered how context may alter the content andenforcement of these norms. Instead, most assume awidespread sense of civic duty that incentivizes polit-ical participation similarly across groups. In contrast,I argue that disparate histories in access to the fran-chise and the segregated nature of American commu-nities create race-specific social norms regarding po-litical participation. These group-based norms oftenlead racial minorities to value participation more thanWhites, especially participation that is grassroots innature.Drawing from multiple social science disciplines, I

develop a framework that contextualizes social normsand anticipates racial variations in the social motiva-tions that structure participation. I argue that historical

Allison P. Anoll is an Assistant Professor, Department of PoliticalScience,VanderbiltUniversity,230Appleton Pl.,PMB505,Nashville,TN 37203 (allison.p.anoll@vanderbilt.edu).

I am grateful to Gary Segura, Justin Grimmer, Douglas McAdam,Cindy Kam,Efrén Pérez, Josh Clinton, and Larry Bartels for detailedfeedback on this paper at multiple stages. I am thankful for help-ful comments from anonymous reviewers, editors,Vanderbilt faculty,Stanford faculty and seminar participants, commenters at the APSA2015AnnualMeeting, and scholars at Carlton College,Williams Col-lege, University of Rochester, Florida State University, and HarvardUniversity, where I presented this project. An earlier version of thispaper was circulated under the title, “HowRace and Community Af-fect Norms of Political Action in America.”Data collection was gen-erously supported by Stanford’s Laboratory for the Study of Ameri-can Values, the Office of the Vice Provost for Graduate Education atStanford University, Kyle Dropp, and Gary Segura. Replication filescan be found on the American Political Science Review Dataverse:https://doi.org/10.7910/DVN/SJI4IA

Received:May 31, 2017; revised:December 17, 2017; accepted:March23, 2018.

limitations in access to the franchise have forced mi-nority groups to place additional social weight on un-conventional forms of participation including protestsand political rallies. Furthermore, “the long march”to voting rights characterized most recently by theCivil Rights Movement has created a powerful pro-voting norm among Black Americans. These group-based variations in participatory social norms shouldbe strongest, I argue, in segregated communities wheregroup cohesion allows social information to be easilyshared and where enforcement is most potent. Whenconsidering racial segregation and historical differ-ences in access to democratic governance, it becomesclear that rather than a monolithic civic duty norm,Americans exist in a context-specific framework ofsocial motivations—one that systematically varies byrace and place,and often leads racial minorities tomag-nify the social value of political involvement.I use a mixed-data, mixed-methods approach to test

whether attitudes about political participation and so-cial rewards for political involvement diverge by race.My analyses focus on differences across the threelargest racial groups in the United States: Whites,Blacks, and Latinos.1 Using an original, nationally rep-resentative survey of roughly 2000 respondents, I findthat minority Americans are more likely to value un-conventional political acts thanWhites, believing theseactivities help people in need and improve their com-munities at significantly higher rates. Similarly, BlackAmericans place more symbolic weight on the act ofvoting than do other groups.Next, I use an original survey experiment to show

that beliefs about the value of participation translateinto willingness to socially reward individuals engagedin these acts. I present roughly 1,200 Whites, 1,200Blacks, and 750 Latinos with profiles of hypothetical

1 There is disagreement about whether the category “Latino” moreclosely approximates a racial or ethnic group (Krogstad and Cohn2014;Hochschild,Weaver, and Burch 2012). For the purposes of thispaper, I operationalize groups using three racial categories—self-identified non-Hispanic Whites, non-Hispanic Blacks, and Latinosof any race—but recognize the boundaries of racial groups are con-stantly shifting and that significant within-group variations also likelyexist.

1

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

Allison P. Anoll

community members involved in various political ac-tivities and ask them to judge these potential neighborson likability and respectability. Consistent with the ob-servational results, Blacks and Latinos on average pro-vide more social incentives to be active in grassrootspolitical activities like rallies than doWhites, and BlackAmericans consistently reward voters at comparativelyhigher rates than both groups.I then merge my survey data with information

from the 2009-2013 American Community Surveyto test whether the social value of participating istied to contextual factors. I find that social rewardsare strongly moderated by context. Specifically, BlackAmericans’willingness to reward political involvementsignificantly increases when they live in mostly co-racial neighborhoods. This finding suggests one path-way through which Black Americans have managed toovercome persistent voting costs, engaging at higherrates than expected according to both conventionaland aggregate models of turnout (e.g., Wheaton 2013;Verba, Schlozman, and Brady 1995; Verba and Nie1972). While Latinos living in mostly Latino neighbor-hoods are no more likely to socially reward partici-pators, those living in areas with large populations ofnoncitizens are, suggesting that existing rather than his-torical disenfranchisement may be the primary contex-tual dimension driving norm heterogeneity for Latinos.Together, my results suggest that not only is po-

litical participation social, but that the communitiesin which norms are forged and enforced vary in theUnited States. Many dimensions of community differ-ence likely exist but race remains a fundamental sepa-rator, altering social space, social relationships, and so-cial experience. It is only when we consider race, place,and social norms, focusing on the social contexts inwhich concepts of civic duty are built, that we mightfully understand why some people participate in poli-tics while others do not.

RACE, PLACE, AND SOCIAL NORMS

Scholarship on political participation in the UnitedStates has a long history. Early theorists often arguedthat participation was natural, sown into the very fab-ric of the human spirit (Madison [1787] 2004, 60–9),but twentieth-century work has since established thatinvolvement in politics is in fact teeming with chal-lenges that often prevent it from occurring (Downs1957; Olson 1965; Salisbury 1969). Specifically, partic-ipation takes time and money, among other resources,and political outcomes are often uncertain. Combinedwith the shared nature of the final good,people face thecompelling option to free ride on the efforts of others.Viewed through this lens of collective action, participa-tion becomes anything but inevitable.A long research tradition in social psychology helps

to explain how citizens might overcome these bar-riers to occasionally participate in politics. Scholarshave found over the course of decades that group-based social norms can powerfully shape both attitudesand behaviors—even the seemingly irrational or costly

(Cialdini and Trost 1998; Cialdini, Kallgren, and Reno1991; Crandall 1988; Asch 1955; Sherif 1936). Socialnorms are the informal rules or standards of a groupthat develop through human interactions and help in-dividuals build and maintain relationships. Althoughsome norms simply describe behaviors, injunctivenorms actually prescribe them,defining what is morallyright or valued by a group (Tankard and Paluck 2016).Scholars have shown that group members reward atti-tudes and behaviors that are believed to be prescrip-tively valuable while punishing deviation (Cialdini andTrost 1998). In leveraging social rewards and sanctionslike inclusion, respect, and friendship, group membersshape the behavior of those in their social space by fun-damentally shaping motivation for action.Over time, external norms can become internalized,

transforming into personal norms that guide the ac-tions of individuals even in the absence of observa-tion (Schwartz 1977; Cialdini and Trost 1998; Hogg2003). Through this process, norms also serve a cen-tral role in building and managing self-concept. In theworld of Bentham or Foucault, personal norms mightbe thought of as a cognitive panopticon, a constantmonitoring system based in the principles of the out-side world. But as social psychologist or political scien-tists, we might more directly conceive of them as moralbeliefs or attitudes—individual concepts about what isvaluable, good, and right.Drawing from the work of social psychologists, po-

litical scientists have shown that social influence af-fects political behavior just as it does all other formsof human action (McClendon 2014; Rogers, Fox, andGerber 2013; Panagopoulos 2010; Gerber and Rogers2009). Gerber and colleagues (2008), for instance, per-suasively demonstrate that the threat of social observa-tion increases turnout at rates far surpassing othermes-sages. McKenzie (2004) has found that kinship groupsstrongly influence participatory choices. And Sinclair(2012) demonstrates that political giving to campaignsand candidates is a function of peer networks. Politicalbehaviors, it seems, are fundamentally social behaviors,shaped not only by individual characteristics like re-sources and education (Verba, Schlozman, and Brady1995), but also by social information, the environment,and peer pressure.To date,work on participatory social pressure largely

assumes or asserts the presence of a positively va-lenced participatory norm among Americans. Schol-ars posit a widespread injunctive norm that values po-litical involvement—often, this is called civic duty—and then layer on a measure of observation or pres-sure, demonstrating that with this social componentcomes an increase in political participation. Despitefocusing primarily on White samples (Gerber, Green,and Larimer 2008),2 or only one other racial group ata time (McKenzie 2004), scholars typically interprettheir findings as widely generalizable. Gerber, Green,and Larimer (2008) provide a clear example of this.

2 The authors limit their sample to Republican primary voters. Be-cause most Blacks, and increasingly Latinos, are Democrats, thischoice largely excludes both minority groups from the sample.

2

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

What Makes a Good Neighbor?

In their landmark study on norm enforcement, theauthors write, “while people vary in terms of their will-ingness or eagerness to conform to norms of civic par-ticipation, the norm is widely accepted as an appropri-ate behavioral standard” (Gerber, Green, and Larimer2008, 40).This proposition assumes that citizens across the

country experience their government the sameway,yetthis is decidedly not the case. Rather, since the found-ing of the republic, some groups have been allowed toparticipate in democratic self-governance while othershave not (Takaki 2008; Keyssar 2000). Race, in par-ticular, has served as a salient dividing line in politics,defining inclusion and exclusion, overdetermining lifechances, structuring access to resources, and moderat-ing the reception of new citizens (Haney-Lopez 1997;Katznelson 2005; Dawson 1994; Jiménez 2010). Racialgroups continue to face different constraints in accessto the franchise and experience a fundamentally differ-ent “face-of-the-state” when interacting with bureau-crats and elected officials (Uggen,Shannon,andManza2012; Lerman and Weaver 2014; Butler and Broock-man 2011).

Seeking political voice,Americanminorities throughthe centuries have turned to alternative means of polit-ical involvement, including contentious protest meth-ods and grassroots organizing (Barreto et al. 2009;Hogan 2007;McAdam 1988).With traditional avenuesof participation closed, racial minorities have devel-oped a unique toolset of participatory skills that has,in some cases, effectively changed the course of pol-icy (Gillion 2013; Enos, Kaufman, and Sands 2016).These struggles may lead minority Americans to as-sign additional value to protestmethods,behaviors thathave been uniquely important to the group in the past.This may be especially true for groups who continueto lack access to the franchise like Black communi-ties that are disproportionately affected by felony dis-enfranchisement or Latino communities heavily popu-lated by undocumented residents. Furthermore, groupswho fought for access to the franchise may place ad-ditional symbolic weight on the act of voting that farexceeds its instrumental value, even exceeds the per-ceived value among Whites. As a result, what the “ap-propriate behavioral standard” is for political involve-ment may vary dramatically across communities withdifferent historical relationships to participation, thetools of involvement, and governmental power.This should be especially true in the context of

the segregated social environments that persist in theUnited States today. Despite the work of activists overthe past century to deconstruct American segrega-tion, the nation’s macrostructures like schools, neigh-borhoods, and places of worship remain extensivelysegregated along the dimension of race (Logan andStults 2011;Logan,Stowell, andOakley 2002;Reardon,Yun,andMcNulty Eitle 2000;Dougherty 2003). In fact,since the 1980s, the trend toward racial integration haslargely stagnated or in some cases reversed (Reardonand Owens 2014; Logan 2011).

Segregation encourages the development of group-based norms in two ways. First, segregation struc-

tures the micro-level relationships through which so-cial norms emerge and are perpetuated (McPherson,Smith-Lovin, and Cook 2001; Lawler, Ridgeway, andMarkovsky 1993). If social spaces of racial groups aredistinct, the opportunity for norm divergence is presentand members of socially separate groups may begin todevelop unique norms in response to their experiences.Second, racially homogenous neighborhoods increasethe ease of spreading information, levels of group co-hesion, and willingness to trust and interact with neigh-bors (Oliver 2010; Larson and Lewis 2017; DeSanteand Perry 2016). Trust, cohesion, and the spread ofinformation each increase the enforcement power ofnorms; norms are most potent when individuals careabout the opinions of other group members and wheninformation about deviance or compliance can spreadamong connections.The combination of distinct social spaces and diver-

gent experiences with government create the oppor-tunity for participatory norms to develop differentlyacross racial groups. Rather than producing a mono-lithic civic duty norm, I argue that historical relation-ships to the franchise and the macrostructures of seg-regation create systematic variations in participatorynorms. Specifically, I hypothesize that (1) racial minori-ties including Black Americans and Latinos will placemore social value on contentious political acts thanWhites, reflecting the historical and present-day use-fulness of these nonelectoral techniques for minoritygroups. Furthermore, I expect that (2) neighborhoodcontext will moderate the strength of norms. Specifi-cally, I expect that Blacks living in co-racial neighbor-hoods or Latinos living in areas largely populated bynaturalized or noncitizens will be most likely to incen-tivize nonelectoral political participation with socialrewards.In addition to contentious political acts, I expect that

norms related to the value of voting will also varyacross racial communities. Specifically, I expect that (3)BlackAmericans—a group with both a long and recentstruggle to gain the franchise—will place special sym-bolic significance on the act of voting, rewarding votersmore than Whites and maybe even more than Latinos.While many Latinos have ancestral ties to those whofought for the right to vote,many still lack the franchisedue to citizenship status and, as a result, information orenforcement of voting norms may be weaker. Finally, Ihypothesize that (4) norms of voting will be strongestin Black communities where clarity and enforcementof this unique social norm is easiest.These hypotheses have critical implications for un-

derstanding who participates in the United States andhow mobilization efforts should be designed. First,higher-than-average social rewards for engaging inminority communities may help to explain persis-tent aggregate-level patterns in political participation.Specifically, for decades scholars have found that whencontrolling for individual-level resources like incomeand education, Black Americans often outperformWhites in models of voting (Verba, Schlozman, andBrady 1995; Wolfinger and Rosenstone 1980; Uhlaner,Cain, and Kiewiet 1989). In 2012, this pattern even

3

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

Allison P. Anoll

appeared in the aggregate when Black voter turnoutexceeded all other major racial groups despite persis-tent barriers like voter identification laws and lower-than-average resources (Wheaton 2013). Scholars haveshown that high Black turnout is most pronounced inneighborhoods dense with Black residents even whencontrolling for mobilization efforts (Barber and Imai2014; Fraga 2016). A theory of contextualized socialnorms suggests one explanation for these patterns:powerful pro-voting norms that are especially preva-lent in Black neighborhoods provide Black Americanswith incentives to vote that can counteract other costs.Second, the hypotheses add to a growing literature

that suggests predictors of political attitudes and be-haviors among racial minorities may be different thanfor Whites (Masuoka and Junn 2013; Barreto and Se-gura 2009; Dawson 1994). Rather than simply studyingthe behavior of majority Americans and generalizingthese findings to minorities, I consider the way con-text alters the content and activation of political socialnorms.

THE VALUE OF PARTICIPATION VARIES BYRACE

I have argued that minority groups with distinct histo-ries and ongoing experiences with exclusion frompolit-ical power should see the purpose, value, and meaningof political participation differently than the dominantgroup. To test this theory, I begin by examining indi-vidual attitudes—or personal norms—about the valueof political participation across racial groups. Drawingfrom ethnographic and historical work, I construct aset of questions to measure dimensions of potentialdifference in beliefs about the usefulness of politicalparticipation. Specifically, I examine individualisticversus communitarian approaches to political activity(Takaki 2008;Dawson 2001;Benmayor,Torruellas, andJuarbe 1997; Flores and Benmayor 1997), test whetherparticipation is a way to express the Good Samaritannorm and care for others (Dawson 2001;Dawson 1994;Harris-Lacewell 2004; Darley and Batson 1973), and,finally, ask a more general question about politicalparticipation as tied to community membership. Thesequestions unpack the broader concept of civic duty andconsider whether prescriptive beliefs are tied to polit-ical participation for some groups more than others.Reflecting my focus on both contentious and tra-

ditional forms of political participation, respondentswere asked to evaluate two types of involvement: vot-ing in presidential elections and attending political ral-lies. While voting requires legal enfranchisement andis both state sanctioned and state organized, politi-cal rallies are generally more grassroots in nature andare accessible to excluded political groups. For eachact, respondents indicated whether they believed theact (a) makes their life better, (b) makes their com-munity better, (c) helps people in need, and (d) indi-cates a good community member. These survey ques-tions were fielded as part of an omnibus online studyin August of 2014 to a nationally diverse sample of

roughly 2,000 American adults. I used the online sur-vey platform, Survey Sampling International, whichdraws from a preconstructed panel of respondents.3Table 1 displays the mean score of each question for

the three racial groups.4 Significant differences (p <0.05) between the mean estimate forWhites and eitherBlacks or Latinos are reported with an a in Columns2 and 3, respectively, and were calculated with a two-tailed t-test. Significant differences between the twominority group estimates are indicated with a b inColumn 3.5

Turning to political rallies first, I find that,on average,both minority racial groups value contentious, grass-roots political participation more than Whites. This istrue across all four measures. On average, Blacks andLatinos are significantly more likely to believe that po-litical rallies make their lives better (0.38, p < 0.05;0.40,p< 0.05, respectively), improve their communities(0.36, p< 0.05; 0.53, p< 0.05), and help people in need(0.62, p < 0.05; 0.45, p < 0.05). Furthermore, politicalrally attendance is more likely to be connected to thesocial trait of good community membership for bothminority racial groups (0.31, p < 0.05; 0.33, p < 0.05).

These differences are quite substantive in size: a 0.53shift in evaluating political rally attendance as a wayto make communities better represents the equiva-lent of half a standard deviation change in the mea-sure. The smallest reported difference, 0.31, still rep-resents nearly a third of a standard deviation changein evaluations of political rally attenders as good com-munity members. The results indicate that both BlackAmericans and Latinos, on average, see more valuein contentious, grassroots political participation thando Whites, consistent with Hypothesis 1. On the otherhand, there is no significant difference between Blackand Latino evaluations of political rally attenders onthese measures.Similarly, I find that Black Americans are, on aver-

age, significantly more likely than Whites to see vot-ing in presidential elections as valuable, in line withHypothesis 3. Blacks are significantly more likely than

3 In addition to quota sampling on the front-end, I built and applieda survey weight using the Current Population Survey to ensure rep-resentativeness on the back-end. This weight was constructed usingpopulation estimates on the dimensions of gender, age, education,income, geographic region,marital status, and race.Becausemy anal-ysis focuses on only Black, White, and Latino respondents, and be-cause of somemissing cases onweight dimensions,my final analyticalN is 1,482.4 Table A1 in the online appendix also presents these estimates withcontrols for partisanship, education, income, age, and gender added.The addition of these controls does not change the substantive re-sults. In Table 1, I choose to present means because many have ar-gued interpreting racial coefficients when including a large numberof sociodemographic controls essentializes racial groups and maymiss the conglomeration of social forces that make race what it is(Gelman and Hill 2007; King and Zeng 2006; Sen and Wasow 2016).5 Some may argue a Bonferroni correction is required for these sig-nificance tests because I am completing three t-tests per row.A Bon-ferroni correction demands dividing the conventional 0.05 signifi-cance level by the number of tests completed, producing a new, con-servative test of significance at the 0.0167 level. All significant re-sults presented in Table 1 also hold at this more demanding level ofsignificance.

4

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

What Makes a Good Neighbor?

TABLE 1. Racial Variations in Beliefs About the Value ofPolitical Actions

White Black Latino

Political rallies...make my life better 2.07 2.45a 2.47a

make my community better 2.29 2.65a 2.82a

help people in need 2.25 2.87a 2.70a

Political rally attenders aregood community members 2.73 3.04a 3.06a

Voting in presidential elections...makes my life better 3.34 3.59a 3.21b

makes my community better 3.46 3.68a 3.52helps people in need 3.22 3.65a 3.30b

People who vote aregood community members 3.77 3.76 3.50ab

N 1095 232 155

Notes: Question wording, ”Please indicate how much you believe each of the fol-lowing statements to be true.” Response categories: 1 = Not at all, 2 = Not very,3 = Somewhat, 4 = Very, 5 = Extremely. Columns show weighted means for eachracial group. Significant differences (p < 0.05) determined through a two-tailed t-testbetween White and minority means are indicated with an a in the second and thirdcolumns if present; significant differences (p < 0.05) between Blacks and Latinos areindicated with a b in the third column if present.

Whites to believe that voting has a positive impacton their lives (0.25, p < 0.05) and community (0.22,p < 0.05). The size of the difference grows consider-ably when comparing the two groups on beliefs aboutwhether voting is an effective way to help people inneed. Here, Black Americans are nearly a half a point,or more than one-third of a standard deviation, morelikely than their White counterparts to say that votingin presidential elections is a way to help others (0.43,p < 0.05). Blacks, however, are no more likely thanWhites to see people who vote as good communitymembers.Among Latinos, the results are slightly different.

There is no significant difference between Latinos andWhites in their beliefs about the ability of voting to im-prove their lives or communities. Furthermore, the twogroups are similar in their beliefs about whether vot-ing can help people in need. It is only on the measureof whether voters make for good community membersthat the two groups diverge: Latinos are significantlyless likely than Whites to think voters are good peopleto have as part of their communities (−0.27, p < 0.05).

The results suggest that, on average, Americans ofdifferent racial groups perceive the value of politicalaction differently.That is, rather than amonolithic civicduty norm, racial group membership moderates per-ceptions of political activity and the underlying con-struct of what civic participationmeans andwhy it mat-ters. In particular, Black Americans are more likelythan both Whites and Latinos to see voting as havinga positive impact on their lives and communities andas a way to care for those most in need, consistent withHypothesis 3.Bothminority groups,however, aremorelikely than Whites to value political rallies, especially

their ability to transform their communities for the bet-ter, consistent with Hypothesis 1.This first section of findings focuses on individual-

level prescriptive beliefs about the value of politicalaction and how, on average, these beliefs vary acrossracial groups. Next, I turn to examining whether thesepersonal norms carry with them social expectations, af-fecting the participatory social motivations that groupmembers encounter.

SOCIAL REWARDS FOR PARTICIPATIONVARY BY RACE

Injunctive social norms derive their power as behav-ioral motivators from the social rewards and sanctionscommunity members levy against each other. The de-sire to be liked, respected, and accepted motivatescommunity members to align their behaviors with thevalues of their community (Cialdini and Trost 1998).Defining what is good and right and attaching socialbenefits to these behaviors, injunctive social norms al-ter the incentive structure for action and, in turn, shapethe behavior of individuals and communities. If pre-scriptive beliefs about participation vary across racialgroups as the previous section suggests, do the socialincentives to be active in different types of political be-haviors vary as well?To test for this heterogeneity, I measured differences

in social rewards for political activity by race using anoriginal online experiment that included a large sam-ple of Black, White, and Latino respondents. Draw-ing from a design utilized by Gerber and colleagues(Gerber et al. 2016), I administered a within-subjects

5

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

Allison P. Anoll

TABLE 2. Randomly Assigned Values in Experiment

Name Louis, Anthony, Martin, DavidAge 34, 35, 37, 38Occupation X-ray Technician, High School Teacher, Physical Therapist,

Legal AssistantCommunity Activity Votes in every presidential election, Annually attends political rallies,

Annually attends neighborhood potluck dinners, Annually attendscharitable events

experiment that asked respondents to evaluate fourdifferent hypothetical individuals who are “interestedin moving to your community.”6 Respondents vieweda table of information about each individual on sepa-rate survey pages, which included randomized entriesabout the individual’s name, age, occupation, and com-munity involvement (see Table 2). Embedded in thesetables were two political activity treatments—“votes inevery presidential election” and “annually attends po-litical rallies”—and a nonpolitical control—“annuallyattends neighborhood potlucks.”7 Immediately follow-ing the table, respondents were asked to evaluate theindividual on two measures of social desirability: lik-ability and respectability. Both measures follow fromprevious experimental work (Gerber et al. 2016;Byronand Baldridge 2007; Ridgeway and Correll 2006; Sny-der and Haugen 1994; Hamilton and Fallot 1974) andare designed to estimate the relative social incentivesattached to political actions.This experimental design allows me to examine dif-

ferences in the social desirability of political acts acrossgroups while controlling for potential differences initem functioning. For instance, by including a nonpo-litical control (potluck), I am able to account for anyaverage sociability shifts that exist across groups. Fur-thermore, by including information about age, name,and occupation, I decrease satisficing as the respon-dents are less able to determine the purpose of thestudy and increase external validity by forcing respon-dents to considermultiple characteristics as they woulda real human being.8

6 In contrast to the counterfactual model (Holland 1986; Rubin1974), a within-subject design draws from a tradition of crossoverstudies popular in medical trials (Chow and Liu 1999) where eachrespondent views both the treatment and the control.One importantassumption must hold true for within-subject crossover designs toproperly identify treatment effects: early treatment exposures mustnot affect the results of later ones. To rule out these potential carry-over effects, I examined whether the means of the first appearanceof each treatment were significantly different than the aggregate es-timates. The results, presented in Table A2 of the online appendix,establish that there is no significant difference.7 Although a charitable events treatment was included in this design,for brevity and focus, it is not discussed in this paper.8 Name, age, and occupation values were designed to be function-ally equivalent. Names are gender consistent and were chosen to beracially ambiguous across Black, Latino, and White. Ages roughlyplace someone in middle age. I used the Nakao-Treas Prestige Scoreand Hauser-Warren Socioeconomic Index to choose occupationsconsidered generally equal in terms of socioeconomic status and gen-der distribution (Hauser and Warren 1997). Furthermore, all valuesare randomized to appear once, canceling out any unexpected effects.

The experiment was administered on three omnibussurveys using the online platforms YouGov/Polimetrixand LatinoDecisions between themonths of April andOctober in 2014.9 The first survey, conducted throughYouGov/Polimetrix, yielded a nationally representa-tive sample of 2,000 Americans drawn from an opt-in panel of U.S. adults.10 To supplement the minor-ity sample available in this pool, I joined this datawith two racial minority-specific samples conductedthrough LatinoDecisions andYouGov/Polimetrix.Theresulting sample included 1,254 White respondents,1,229 Black respondents, and 748 Latino respondents.11Figure 1 plots the effect of voting and rally atten-

dance on likability and respectability among Black andLatino respondents compared to Whites. The point es-timates are calculated using a difference-in-differenceframework that provides an assessment of cross-racialheterogeneity:

DiDj = (yT j − yC j|Minority) − (yT j − yC j|White)

Movement away from the zero line represents the dif-ference in the treatment effect for minority respon-dents compared to Whites. Each point estimate is dis-played with a 95% confidence interval around it.12These point estimates test whether minority groups,

on average, evaluate politically active individuals dif-ferently than Whites. I expect that minority Ameri-cans will reward grassroots political behavior like po-litical rallies more than Whites, reflecting Hypothe-sis 1 and the results in the previous section. Further-more, I expect that Black Americans will reward vot-ers more positively than Whites consistent with Hy-pothesis 3. While Black Americans and Latinos arealso largely segregated from each other and have dis-tinct political histories, which may produce variation in

9 Resource constraints required working with other scholars in anomnibus setting to collect data for this study. Thus, rather than a sin-gle survey with equal proportions of all three racial groups, I usedthree separate omnibus studies to construct a dataset that eventuallyyielded a large and representative sample of each group.10 YouGov/Polimetrix uses a combination of sampling and match-ing techniques to approximate the demographic composition of theUnited States.11 To ensure the representativeness of my analysis, I constructed andapplied a weight to the sample using Current Population Surveyestimates.12 Raw mean estimates for the control and each treatment are pro-vided in Figure A1 of the online appendix.

6

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

What Makes a Good Neighbor?

FIGURE 1. Minority-White Differences in Social Rewards by Political Act

evaluations between minority groups as well, I haveless clear hypotheses about the direction of this varia-tion, and so I focus my analyses on the minority-Whitecomparisons.The first panel (a) of Figure 1 shows evaluations of

rally attenders minus the control for Blacks and Lati-nos compared to Whites. Consistent with Hypothesis1, both minority groups evaluate political rally atten-ders as significantly more likable and respectable thantheirWhite counterparts. Specifically, the treatment ef-fect among Blacks compared to Whites is a positive0.37 difference in likability (p < 0.01) and a 0.34 dif-ference in respectability (p < 0.01). Among Latinos,the treatment effect for rally attendance is nearly halfa point more than White respondents (0.44, p < 0.01),the largest effect across groups and treatments. In termsof respectability, the effect of political rally attendanceis 0.13 (p < 0.05) points more among Latinos thanWhites.On bothmeasures, I find that,on average, racialminorities provide more social rewards for participa-tion in political rallies than do White Americans, con-sistent with Hypothesis 1.Panel b of Figure 1 provides estimates for voter

evaluations among minority respondents compared toWhites. Like the observational results presented inthe previous section and consistent with Hypothesis3, Black Americans, on average, socially value votersmore than do their White counterparts. In particular,Blacks compared to Whites think voters are signifi-cantly more likable (0.39, p < 0.01) and respectable(0.20,p< 0.01).Like the observational results, the find-ings for Latinos aremoremixed.Latinos also see votersas more likable than do White respondents, althoughby a smaller margin (0.17,p< 0.01).On the respectabil-ity measure, there is a negative (−0.03) although

insignificant difference between Latino and Whiterespondents.A couple things can be gleaned from these findings.

First, on average, minority Americans socially valuethose who are politically active at higher rates than doWhites. This is especially true for evaluations of polit-ical rally attenders. Both Latinos and Blacks are moregenerous across both social measures than are Whitesin rewarding rally attenders. These results are reflec-tive both of political reality—for instance, protests andrallies composed of largely racial minorities under thebanners of Black Lives Matter and 2006 ImmigrationReform—and with content differences found in theprevious section. Recall that both Blacks and Latinosconnected political rallies withmore positive outcomesand characteristics compared to Whites, including theability of political rallies to increase the quality of com-munities and help people in need. In both content andsocial rewards, these differences may reflect the acces-sible, grassroots nature of rallies, available as a politicaltool for those locked out of electoral settings or whomaintain minority status in a majoritarian democracy.Second, when it comes to voting, Black Americans

provide more social rewards to voters than do Whites.This finding is consistent with a robust literature thatpoints to turnout overperformance among Blacks inmodels of voting in both the aggregate (Wheaton 2013)and after controlling for socioeconomic status (Anoll2014; Verba, Schlozman, and Brady 1995; Wolfingerand Rosenstone 1980). By ascribing socially desirablecharacteristics to those who regularly vote, Blacks mayadd an additional layer of social incentives that encour-ages political action among group members. This set ofincentives could change the calculus of political actionindividuals face: despite having lower resources than

7

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

Allison P. Anoll

theirWhite counterparts and facing higher costs for po-litical activity, Blacks may encounter social incentivesthat prioritize and reward political action at a higherdegree.Latinos, on the other hand, provide either slightly

more or the same amount of social incentives to vote asWhites, and fewer incentives than Blacks. These find-ings mirror the norms content results in the previoussection,which found that Blacks were more likely thanWhites to think voting makes their lives better, theircommunities better, and most importantly, helps peo-ple in need. There were no such differences betweenLatinos and Whites on these dimensions. Thus, whileBlacks may overcome the costs of voting through pro-viding an additional set of incentives in the form of so-cial rewards and social pressure, Latinos who face mo-bilization challenges due to language heterogeneity, el-igibility, and resources do not. Latinos and Whites pro-vide the same level of social incentives for voting—butLatinos face more costly action.These results may helpexplain low Latino voter turnout compared to Whitesand Blacks.Together, these results suggest that the social value

of political actions varies in substantively meaning-ful ways across racial communities. Some have argued,in addition, that further variation likely exists withingroups. For instance, close analyses of voter turnoutshow that much of the unexplained variation acrossrace happens among Blacks at low levels of socioeco-nomic status (Anoll 2014) and a recent study of voterturnout using validated data shows that “overperfor-mance” happens largely among Black women (An-solabehere and Hersh 2013). Recognizing these im-portant dimensions of within-group difference, I pro-vide in Figure 2 difference-in-difference estimates inlikability and respectability evaluations for four dif-ferent subgroups: respondents with low socioeconomicresources (defined as a yearly family income below$30,000 and no more than a high school education, orapproximating the bottom third of the SES distribu-tion in American society), respondents with mediumor high socioeconomic resources (that is, more than$30,000 in yearly family income and at least some col-lege),women, and men.Bars around the estimates rep-resent 95% confidence intervals.While the experimen-tal nature of the design allows us to rule out differ-ences as simply the result of variations in usage ofthe scale across groups, the addition of these covari-ate analyses tests whether racial variation presented inFigure 1 is the product of income, education, or gendercovariation.The results are largely consistent with the aggregate

estimates presented in Figure 1. Generally, across sub-groups and measures, minority Americans continue toevaluate those who are politically active more favor-ably than their White counterparts. This is true de-spite reduced sample size, which expands confidenceintervals.Thus,when controlling for socioeconomic sta-tus and gender—two important predictors of politicalparticipation—Blacks and Latinos often see voters andoverwhelmingly see political rally attenders in a morepositive light than the majority racial group.

Despite this consistency, a few within-group trendsareworth commenting on,considering extant literatureon patterns in political participation. First, among lowSES respondents, the Black-White gap in respectabilityof voters surges up to 0.37 (p < 0.01), while the gapbetween high SES Whites and Blacks lingers at 0.19(p < 0.05). These relative differences mirror turnouttrends across race and resources, which show that lowSES Blacks often turnout at higher rates than low SESWhites, while the two groups look more similar at highlevels of socioeconomic resources. This finding is espe-cially important considering the relative distribution ofeach group, with many more Black Americans fallinginto the bottom tier of socioeconomic resources thanamong Whites.Interesting differences equally consistent with na-

tional trends appear in evaluations across gender.BothBlack women and Black men evaluate voters and po-litical rally attenders more positively than their Whitecounterparts (p < 0.01), but reflecting validated vot-ing estimates (Ansolabehere and Hersh 2013), thestrongest evaluation of voter and rally attender likabil-ity appears among Black women (0.49,p< 0.01; 0.50,p< 0.01, respectively).A similar,although less consistenttrend is present among Latinos.These results suggest that some important hetero-

geneity exists within racial subgroups, as is to be ex-pected. Although race is an important and relevantcategory of distinction in the United States (Omi andWinant 1994; Masuoka and Junn 2013), racial groupsare not homogenous bodies. However, by and large,subgroup differences reflect aggregate differences intheir general interpretation: rather than a single socialincentive structure for political action, Americans ofdifferent racial communities reward individuals differ-ently when engaging in politics. I turn next to examin-ing how segregation and social context may moderatethese relationships.

SOCIAL CONTEXT MODERATESPARTICIPATORY REWARDS

The results so far indicate that individuals’ racial groupmembership affects views of political activity. I haveargued that this heterogeneity is the product of con-textual factors: disparate histories in access to the fran-chise and group segregation creates norm variance.Here, I more directly test whether contextual factorsare related to the social value of political involvement.Context should matter for shaping norm hetero-

geneity in two ways. First, norms are most potent in so-cial spaces where there is group cohesion (Oliver 2010;Larson and Lewis 2017). In these spaces,members caremore about the opinion of those around them and, as aresult, aremorewilling to conform to group norms.Fur-thermore, an increased density in social connectionsand higher levels of trust aid in the flow of informationand norm enforcement. Thus, spaces with large num-bers of group members who share a salient identity arelikely spaces with clearer,more widely enforced norms.

8

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

What Makes a Good Neighbor?

FIGURE 2. Minority-White Subgroup Differences in Political Activity Evaluations

9

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

Allison P. Anoll

However, heterogeneity in participatory normsshould only appear in contexts where group identity isclearly tied to divergent experiences with government.After all, it is these divergent experiences that creategroup-specific norms. Thus, three factors are importantfor the expectation that context shapes participatorynorms: group cohesion, density of group members, anddivergent experiences with government.Considering these factors, I expect that the propor-

tion of co-racial residents in Black respondents’ neigh-borhoods will be positively related to both social re-wards for voting and political rally attendance. In theseneighborhoods, group cohesion is high and divergentnorms regarding political participation observed in theprevious two sections should be intensified. AmongLatinos, my expectations are slightly different. Manyhave argued that in the aggregate, Latinos are a morediverse and less cohesive group (Beltrán 2010; Dávila2012), but that immigration serves as a unifying polit-ical dimension (Jiménez 2010; Segura 2006). Further-more, Latinos as a group are less clearly tied to a his-torical narrative of exclusion (Takaki 2008), but for-eignness continues to moderate experiences with polit-ical voice (Pantoja, Ramirez, and Segura 2001;Michel-son 2003). Thus, while a Latino context may matterless for participatory norm divergence due to both lessgroup cohesion and less consistency in governmentalexperience, a social context filled with group membersjoined by their experiences with immigration maymat-ter quite a lot.To test if contextual variables moderate participa-

tory social rewards, I merge my survey data on par-ticipator likability and respectability with zip code in-formation from the 2009-2013 American CommunitySurvey: 5-Year Data Set (Minnesota Population Cen-ter 2011).13 I create three contextual measures to oper-ationalize my hypotheses: the proportion of residentsin Black and White respondents’ zip codes who areBlack; the proportion of residents in Latino and Whiterespondents’ zip codes who are Latino; and the pro-portion of residents in Latino and White respondents’zip codes who are foreign-born (that is, are either nat-uralized or noncitizens).14 The results in the previoussection show that likability and respectabilitymeasuresare directionally consistent, so for the sake of presen-tational simplicity, I merge these two measures into a

13 Scholars have used everything from respondent-created boundarymaps to government geo-indicators like census tract and county tomeasure social context (DeSante and Perry 2016; Wong et al. 2012;Stolle, Soroka, and Johnston 2008). However, recent work by Velezand Wong (2017) shows that, across an array of measures, zip codemost closely approximates people’s perceptions of the racial compo-sition of their “local” community. In fact, zip code even outperformsrespondent-constructed boundary maps.14 These measures are best understood as capturing the density ofa specific group in a geographic area rather than how segregatedgroups are from each other. I depart from common segregationmeasures here like dissimilarity, isolation, or contact (Logan 2011;Hutchens 2001; Wong 2013) because I am theoretically less con-cerned with how much contact each group has with the other andmore concerned with whether a critical mass of co-racial residentssurrounds the respondent in ways that could communicate a race-specific norm.

mean social value score and standardize the variablefrom zero to one. A value of one represents an evalua-tion of extremely likable and respectable while a zeromeans the participator is seen as not at all likable or re-spectable. Because I am now looking at heterogeneitywithin group rather than across, I examine raw likabil-ity and respectability scores instead of those adjustedto the experimental potluck baseline.15Table 3 presents linear regression results predicting

the social value of voters and political rally attendersby respondent’s race, neighborhood characteristic, in-come, education, and gender.Models 1 and 2 show therelationship between the proportion of a respondent’szip code that is composed of Black residents and thesocial evaluation of political participators amongBlackand White respondents. Models 3 and 4 show paral-lel analyses but for Latinos and Whites, using propor-tion Latino as the main independent variable. Finally,Models 5 and 6 show the relationship between citi-zenship composition of a neighborhood among Latinoand White respondents. The reference category foreach regression is the minority group. To aid in theinterpretation of the interaction between racial groupmembership and social context, Figure 3 plots the rel-ative predicted probabilities for the reference minoritygroup and White respondents, holding other values attheir mean category. Bands represent 80% confidenceintervals.The results from Models 1 and 2 and Models 5 and

6 are striking in their emphasis on community-levelfactors over individual demographics. In fact, for eachdependent variable in these models, the neighborhoodcharacteristics are the largest predictors, trumping in-come and education across the board. Furthermore, inthese models,Whites and the minority group are statis-tically indistinguishable at the intercept; in most cases,divergence only appears in attitudes amongWhites andminorities whenmoderated by the characteristics of re-spondents’ social spaces. Specifically, shifting the pro-portion of a respondent’s zip code that is Black fromzero to one produces a six-percentage-point increasein Black Americans’ evaluation of voters, but has noeffect on White residents. This is the equivalent of alittle more than a third of a point change in respectabil-ity and likability of voters on a seven-point scale. Indi-vidual income and education, on the other hand, havealmost no predictive power.A similar pattern appears in Model 2: changing the

proportion of Black residents in a respondents’ neigh-borhood produces a six-percentage-point change in

15 The analyses in this section are observations and, as a result, somemay question the causal direction of my findings. However, existingwork has established that when it comes to community sorting, peo-ple are quite constrained. Mummolo and Nall (2016), for instance,find that concerns about neighborhood quality and affordability limitsorting on even the most salient political dimension—partisanship.Furthermore, 37% of Americans have never lived anywhere otherthan their hometown and, when people do move, they overwhelm-ingly cite either job opportunities, community quality for children,or returning to family ties, not political considerations (Taylor et al.2008). Thus, I think it is unlikely that most people are sorting intocommunities based on participatory norms.

10

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

What Makes a Good Neighbor?

TABLE 3. Neighborhood Context Moderates the Social Value of Participation

Black & White Latino & White Latino & White

(1) Voter (2) Rally (3) Voter (4) Rally (5) Voter (6) Rally

Intercept 0.76 (0.01)∗ 0.70 (0.01)∗ 0.77 (0.02)∗ 0.74 (0.02)∗ 0.73 (0.02)∗ 0.71 (0.02)∗

White 0.00 (0.01) − 0.02 (0.01) − 0.02 (0.01) − 0.04 (0.02)∗ 0.03 (0.02) − 0.01 (0.02)Income − 0.00 (0.00) − 0.00 (0.00)∗ 0.00 (0.00) − 0.00 (0.00)∗ 0.00 (0.00) − 0.00 (0.00)∗

Education − 0.01 (0.00)∗ − 0.00 (0.00) − 0.01 (0.00)∗ − 0.00 (0.00) − 0.01 (0.00)∗ − 0.00 (0.00)Male − 0.04 (0.01)∗ − 0.01 (0.01)+ − 0.04 (0.01)∗ − 0.02 (0.01)∗ − 0.04 (0.01)∗ − 0.02 (0.01)∗

Prop. Black Zip 0.06 (0.02)∗ 0.06 (0.02)∗

Prop. Black Zip∗White − 0.07 (0.05)+ 0.03 (0.05)Prop. Latino Zip − 0.01 (0.03) − 0.00 (0.03)Prop. Latino Zip∗White 0.08 (0.04)+ − 0.00 (0.05)Prop. Foreign-Born Zip 0.14 (0.05)∗ 0.12 (0.06)∗

Prop. Foreign-Born∗ White − 0.16 (0.07)∗ − 0.18 (0.08)∗

Num. obs. 2463 2463 1981 1981 1981 1981RMSE 0.18 0.19 0.19 0.20 0.19 0.20R2 0.03 0.02 0.02 0.02 0.03 0.03Mean of DV 0.72 0.66 0.71 0.66 0.71 0.66SD of DV 0.18 0.19 0.17 0.19 0.17 0.19

Notes: OLS regression with weights predicting social value of voting and political rally attendance. Social value calculated by averaginglikability and respectability scores and standardizing from 0–1. A value of 1 represents “extremely likable and respectable” and a valueof 0 means “not at all likable and respectable.” Excluded category is minority respondents—either Black or Latino. Standard errors inparentheses, ∗p < 0.05, and +p < 0.10.

Black respondents’ evaluation of rally attenders, al-though in this case,White respondents are similarly af-fected.16 Together, Models 1 and 2 show that, on av-erage, social incentives to participate in both electoraland nonelectoral acts are stronger in Black neighbor-hoods. Because social motivations can help to over-come other costs associatedwith political participation,these results suggest a possible explanation for previ-ous findings, which show that Black Americans livingin mostly Black spaces tend to turn out at higher ratesthan their counterparts living in more diverse contexts(Barber and Imai 2014; Fraga 2016).Models 3 and 4 show a less persuasive case that the

proportion of Latinos in respondents’ social spaces af-fects the social value of political action.Rather, there isno significant relationship between the social value ofparticipators and the proportion of Latinos living in azip code. Instead, the social characteristic that mattersfor creating strong social commitments to participationis the proportion of a neighborhood that is either natu-ralized or noncitizens.Models 5 and 6 show that chang-ing the proportion of foreign-born citizens in Latino re-spondents’ neighborhoods from zero to one producesa fourteen-percentage-point increase in social evalua-tions of voters and a twelve-percentage-point increase

16 Of the four models where social context affects minority evalua-tions, this model is an outlier in that Whites are also positively af-fected. I am hesitant to infer too much from this result, first, becauseit is an outlier, and second, because I do not have strong hypothesesabout howWhites will be affected in these neighborhoods.However,the finding suggests the possibility that strong neighborhood normsabout political participation among Blacks may spill over to Whitesliving in these contexts as well.

in evaluations of rally attenders. This is equivalent tonearly a whole point change in the social evaluationof participators on a one to seven scale.Whites, on theother hand, are largely unaffected by this changing so-cial context. Like the results presented for Black re-spondents, this analysis makes clear the powerful im-pact of community-level attributes but highlights that,for Latinos, the primary contextual variable that mat-ters might be immigration status.Together, these results point to the effect that

community-level attributes have in shaping norma-tive beliefs about political action. By focusing onindividual-level characteristics such as resources or byexamining normative pressure in the aggregate, schol-ars often ignore the very social nature of norms, whichrequires community to develop, spread, be enforced,and, eventually, have behavioral impacts. By placingnormative evaluations in the context of communitiesand groups, it becomes clear that it is not only therace of an individual that creates differences in the so-cial evaluations of political behaviors, but also, maybemore importantly, race interacted with the attributes ofthe communities in which people live.These systematicvariations in the social reward system likely influencethe actions of individuals embedded in communities ofdifferent types, shaping political behavioral outcomesin the aggregate.

DISCUSSION

As the demography of the United States has changedin recent years, a growing number of pundits—and

11

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

Allison P. Anoll

FIGURE 3. Moderating Effect of Neighborhood Context on Social Rewards

the occasional academic (Huntington 2004)—havebemoaned the diversification of the country. Claim-ing racial minorities are lacking in values like patrio-tism, individualism, and work ethic, conservative pun-dits like Glenn Beck and Rush Limbaugh have posited“the total degradation of what used to be Americanculture” (Gardner, Thompson, and Rucker 2010; Lim-baugh 2010). These claims are a modern manifesta-tion of an enduring narrative in America, one whichin its many forms has over time asserted deficienciesamong non-Whites. Built on racialized beliefs aboutfamily values, crime, self-control, and laziness, this nar-rative has continually served as justification for demo-cratic exclusion (Takaki 2008; Haney-Lopez 1997; U.S.Department of Labor 1965).

Contrary to these claims, I find that on many di-mensions of civic participation, racial minorities areeven more committed to the American ideal of self-governance, encouraging and rewarding political in-volvement at higher rates thanWhites.Minority Amer-icans are more likely than Whites to see political ac-tivities as a way to care for their community and help

people in need, and more likely on average to pro-vide social rewards to those active in politics. Thisis especially true for more grassroots political activi-ties that may provide minority Americans with demo-cratic voice when electoral opportunities are restricted.Those living in mostly Black neighborhoods or sur-rounded by large numbers of foreign-born residentsare especially likely to believe that the politically ac-tive make good neighbors. My findings suggest thaton the dimension of political participation, racial mi-norities often express an even stronger commitment tocore democratic behaviors, consistent with other publicopinion work on White versus minority values (Tayloret al. 2012; Citrin et al. 2007).

This commitment likely helps non-Whites overcomethe inevitable costs and challenges involved in par-ticipating in politics. Despite reform attempts, barri-ers to Black and Latino political participation in theform of both unequal resources and institutional con-straints remain a hallmark of the American politi-cal system (Uggen, Shannon, and Manza 2012; Wil-son 2011). Yet, racial minorities are often able to

12

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

What Makes a Good Neighbor?

overcome these barriers to participate at unexpect-edly high rates, sometimes even outperforming Whites(e.g. Wheaton 2013; Verba, Schlozman, and Brady1995;Pantoja,Ramirez,and Segura 2001).Strong socialnorms that value and reward participation may helpexplain these trends, speaking to, for instance, decadesof “overperformance” in Black American political par-ticipation (Barber and Imai 2014; Fraga 2016; Verba,Schlozman,andBrady 1995;Wolfinger andRosenstone1980).

Unlike previous work on participatory social norms,I highlight the group-based nature of norm develop-ment and enforcement. Race continues to moderateAmericans’ experiences with government, and segre-gation emphasizes these differences, providing bothreason and social space for norms to develop differ-ently. Whether studying norms or other forms of po-litical behavior, it is not uncommon for scholars torely on mostlyWhite samples, remove racial minoritiesthrough fixed effects in models, or ignore them entirelywhen interpreting results and generalizing claims (seeBarreto and Segura 2009 and Masouka and Junn 2013for a discussion).But the assumption that political vari-ables affect members of each racial group similarly isnot tenable, especially in situations where social inter-actions are the foundation of a theory.Does social pressure play as important a role in mo-

bilizing minority Americans as it does Whites, consid-ering this heterogeneity in norm development? Morework should consider the relationship between socialmotivations and political involvement in non-Whitesamples but initial evidence suggests the answer tothis question is yes. Social psychologists find normspowerfully motivate behavior across societies and cul-tures but the content of norms and the appropriatereference groups for social pressure change with con-text (Tankard and Paluck 2016; Fryer and Torelli 2010;Paluck and Green 2009). Structural features that con-tinue to impose unequal costs on the participation ofminority Americans, however, likely raise the amountof social rewards required to offset these costs. That is,racial minorities may have to socially motivate politicalinvolvement exponentiallymore thanWhites to get thesame level of group participation.Understanding racial heterogeneity in political at-

titudes and behavior is increasingly important in theUnited States. Recent projections suggest that non-Whites are on track to become the majority of Amer-icans around midcentury (Lofquist et al. 2012; Jonesand Bullock 2012), yet aggregate trends in turnoutamong racial minorities remain notoriously difficult toexplain. My findings suggest that to understand thesenational trends in political involvement, political sci-entists should consider how social incentives to partic-ipate are tied to race and place. While social motiva-tions most certainly matter for overcoming the collec-tive action problem,norms that guide these rewards arenot uniform. Rather, the meaning of civic duty variessystematically across racial communities. Together, so-cial norms and social context emerge as a structure forunderstanding participation in an increasingly diverseAmerica.

SUPPLEMENTARY MATERIAL

To view supplementary material for this article, pleasevisit https://doi.org/10.1017/S0003055418000175.

Replication materials can be found on Dataverse at:https://doi.org/10.7910/DVN/SJI4IA

REFERENCES

Anoll, Allison. 2014. “Racial Differences in the Socioeconomic-Participation Relationship.” Midwest Political Science Associa-tion, Annual Meeting, Chicago, IL. Accessed on April 3–6, 2014.

Ansolabehere, Stephen, and Eitan Hersh. 2013. “Gender,Race,Age,and Voting:AResearch Note.”Politics andGovernance 1 (2): 132–37.

Asch, Solomon E. 1955. “Opinions and Social Pressure.”ScientificAmerican 193: 31–55.

Barber,Michael, and Kosuke Imai. 2014. “Estimating NeighborhoodEffects on Turnout fromGeo-Coded Voter Registration Records.”Princeton, NJ:Mimeo.

Barreto,Matt A., SylviaManzano,Ricardo Ramírez, and Kathy Rim.2009. “Mobilization, Participation, and Solidaridad: Latino Partic-ipation in the 2006 Immigration Protest Rallies.” Urban AffairsReview 44 (5): 736–64.

Barreto, Matt A., and Gary M. Segura. 2009. “Estimating the Ef-fects of Traditional Predictors, Group Specific Forces, and Anti-Black Affect on 2008 Presidential Vote Among Latinos and Non-Hispanic Whites.” Ohio State University Conference on the 2008Election, Columbus, OH.

Beltrán, Cristina. 2010. The Trouble with Unity: Latino Politics andthe Creation of Identity Oxford: Oxford University Press.

Benmayor, Rina, Rosa M. Torruellas, and Ana L. Juarbe. 1997.“Claiming Cultural Citizenship in East Harlem: ‘Si Esto PuedeAyudar a la Comunidad Mía...”’ In Latino Cultural Citizenship:Claiming Identity, Space, and Rights, eds. William V. Flores andRina Benmayor. Boston,MA: Beacon Press, 152–209.

Butler,Daniel, andDavid Broockman.2011. “Do Politicians RaciallyDiscriminate Against Constituents? A Field Experiment on StateLegislators.”American Journal of Political Science 55 (3): 463–477.

Byron, Kristin, and David C. Baldridge. 2007. “E-Mail Recipients’Impressions of Senders’ Likability: The Interactive Effect of Non-verbal Cues andRecipients Personality.”Journal of Business Com-munication 44 (2): 137–60.

Chow, Shein-Chung, and Jen-Pei Liu. 1999. Design and Analysisof Bioavailibility and Bioequivalence Studies, 2nd ed. New York:Marcel Dekker, Inc.

Cialdini, Robert B., Carl A. Kallgren, and Raymond R. Reno. 1991.“A Focus Theory of Normative Conduct: A Theoretical Refine-ment andReevaluation of theRole ofNorms inHumanBehavior.”Advances in Experimental Social Psychology 21: 201–35.

Cialdini,Robert B., andMelanie R.Trost. 1998. “Social Influence, So-cial Norms, Conformity, Compliance.” In Handbook of Psychol-ogy, 4th ed.,eds.DanielGilbert,Susan Fiske,andGardner Lindzey.New York:McGraw-Hill, 151–92.

Citrin, Jack,AmyLerman,MichaelMurakami, andKathryn Pearson.2007. “Testing Huntington: Is Hispanic Immigration a Threat toAmerican Identity?”Perspectives on Politics 5 (1): 31–48.

Crandall,Christian S.1988.“Social Contagion of Binge Eating.”Jour-nal of Personality and Social Psychology 55 (4): 588–98.

Darley, John M., and Daniel Batson. 1973. “’From Jerusalem to Jeri-cho’: A Study of Situational and Dispositional Variables in Help-ing Behavior.” Journal of Personality and Social Psychology 27 (1):100–8.

Dàvila, Arlene. 2012. Latinos, Inc.: The Marketing and Making of aPeople. Berkeley, CA: University of California Press.

Dawson,Michael. 1994.Behind the Mule: Race and Class in African-American Politics. Princeton, NJ: Princeton University Press.

Dawson, Michael. 2001. Black Visions: The Roots of Contempo-rary African-American Political Ideologies. Chicago:University ofChicago Press.

DeSante, Christopher D., and Brittany N. Perry. 2016. “Bridgingthe Gap: How Geographic Context Affects Political Knowledge

13

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

Allison P. Anoll

Among Citizen and Non-Citizen Latinos.” American Politics Re-search 44 (3): 548–77.

Dougherty, Kevin D. 2003. “How Monochromatic Is Church Mem-bership? Racial-Ethnic Diversity in Religious Community.” Soci-ology of Religion 64 (1): 65–85.

Downs, Anthony. 1957. An Economic Theory of Democracy. NewYork: Harper.

Enos, Ryan D., Aaron Russell Kaufman, and Melissa L. Sands. 2016.“Can Violent Protest Change Local Policy Support? Evidencefrom theAftermath of the 1992 LosAngeles Riot.”Urban PoliticalEconomy Conference, Vanderbilt University, March 17-18, 2017.

Flores, William V., and Rina Benymayor. 1997. “Constructing Cul-tural Citizenship.” In Latino Cultural Citizenship: Claiming Iden-tity, Space, and Rights, eds.William V. Flores, and Rina Benmayor.Boston,MA: Beacon Press, 1–23.

Fraga, Bernard. 2016. “Candidates or Districts? Reevaluating theRole of Race on Voter Turnout.” American Journal of PoliticalScience 60 (1): 97–122.

Fryer, Roland G., and Paul Torelli. 2010. “An Empirical Analysis of‘Acting White.’” Journal of Public Economics 94 (5-6): 380–96.

Gardner, Amy, Krissah Thompson, and Philip Rucker. 2010. “Beck,Palin Tell Thousands to ‘Restore America’.”The Washington Post,August 29,2010.http://www.washingtonpost.com/wp-dyn/content/article/2010/08/28/_AR2010082801106_pf.html.

Gelman, Andrew, and Jennifer Hill. 2007. Data Analysis Using Re-gression and Multi-level/Hierarchical Models. New York: Cam-bridge University Press.

Gerber,Alan S.,Donald P.Green, and ChristopherW.Larimer. 2008.“Social Pressure and Voter Turnout: Evidence from a Large-ScaleField Experiment.”American Political Science Review 102 (1): 33–47.

Gerber, Alan S., Gregory A. Huber, David Doherty, andConor Dowling. 2016. “Why People Vote: Estimating the SocialReturns of Voting.” British Journal of Political Science 46 (2):241–64.

Gerber, Alan S., and Todd Rogers. 2009. “Descriptive Social Normsand Motivation to Vote: Everybody’s Voting and So Should You.”The Journal of Politics 71 (1): 178–91.

Gillion, Daniel Q. 2013. The Political Power of Protest: Minority Ac-tivism and Shifts in Public Policy. Cambridge: Cambridge Univer-sity Press.

Hamilton,David L., and Roger D.Fallot. 1974. “Information Salienceas a Weighting Factor in Impression Formation.” Journal of Per-sonality and Social Psychology 30 (4): 444–8.

Haney-Lopez, Ian F. 1997.White by Law: The Legal Construction ofRace. New York: New York University.

Harris-Lacewell, Melissa Victoria. 2004. Barbershops, Bibles, andBET:Everyday Black Talk and the Development of Black PoliticalThought. Princeton, NJ: Princeton University Press.

Hauser, Robert M., and John Robert Warren. 1997. “SocioeconomicIndexes for Occupations: A Review, Update, Critique.” Sociologi-cal Methodology 27 (1): 177–298.

Hochschild, Jennifer L., Vesla M. Weaver, and Traci R. Burch. 2012.Creating a New Racial Order: How Immigration, Multiracialism,Genomics,and the YoungCanRemakeRace inAmerica.Princeton,NJ: Princeton University Press.

Hogan,Wesley C. 2007.Many Minds, One Heart: SNCC’s Dream foraNewAmerica.ChapelHill,NC:TheUniversity ofNorthCarolinaPress.

Hogg,Michael. 2003. “Intergroup Relations.” InHandbook of SocialPsychology, ed. John Delamater.New York:Kluwer/Plenum, 479–501.

Holland, Paul W. 1986. ”Statistics and Causal Inference.” Journal ofthe American Statistical Association 81 (396): 945–60.

Huntington,Samuel P. 2004.WhoAreWe? TheChallenges toAmer-ica?s National Identity. New York: Simon and Schuster.

Hutchens,Robert. 2001.“NumericalMeasures of Segregation:Desir-able Properties and Their Implications.”Mathematical Social Sci-ences 42: 13–29.

Jiménez, Tomás R. 2010.Replenished Ethnicity: Mexican Americans,Immigration, and Identity. Berkeley, CA: University of BerkeleyPress.

Jones,Nicholas A., and Jungmiwha Bullock. 2012. “The Two orMoreRaces Population: 2010.”United States Census Bureau, September,2012. http://www.census.gov/prod/cen2010/briefs/c2010br-13.pdf.

Katznelson, Ira. 2005.When Affirmative Action Was White: An Un-told History of Racial Inequality in Twentieth-Century America.New York:W.W. Norton and Company.

Keyssar, Alexander. 2000. The Right to Vote: The Contested Historyof Democracy in the United States.” New York: Basic Books.

King, Gary, and Langche Zeng. 2006. “The Dangers of ExtremeCounterfactuals.”Political Analysis 14 (2): 131–51.

Krogstad, Jens Manuel, and D’Vera Cohn. 2014. “U.S. CensusLooking at Big Changes In How It Asks about Race andEthnicity.” Pew Research Center. http://www.pewresearch.org/fact-tank/2014/03/14/u-s-census-looking-at-big-changes-in-how-it-asks-about-race-and-ethnicity.

Larson, Jennifer M., and Janet I. Lewis. 2017. “Ethnic Networks.”American Journal of Political Science 6 (2): 350–64.

Lawler, Edward J., Cecilia Ridgeway, and Barry Markovsky. 1993.“Structural Social Psychology and theMicro-Macro Problem.”So-ciological Theory 11 (3): 268–90.

Lerman, Amy, and Vesla Weaver. 2014. Arresting Citizenship: TheDemocratic Consequences of American Crime Control. Chicago:Chicago University Press.

Limbaugh, Rush. 2010. “Transcript: NFL’s ‘Futbol Americano’Telecast is a Symptom of What Ails America.” http://www.rushlimbaugh.com/daily/2010/09/27/nfl_s_futbol_americano_telecast_is_a_symptom_of_what_ails_america.

Lofquist,Daphne, Terry Lugaila,Martin O’Connell, and Sarah Feliz.2012. “Households and Families: 2010.”United States Census Bu-reau. http://www.census.gov/prod/cen2010/briefs/c2010br-14.pdf.

Logan, John R. 2011. “Separate and Unequal: The NeighborhoodGap for Blacks, Hispanics and Asians in Metropolitan America.”Census Brief prepared for Project US2010. http://www.s4.brown.edu/us2010.

Logan, John R., Jacob Stowell, and Deidre Oakley. 2002. “ChoosingSegregation: Racial Imbalance in American Public Schools, 1990-2000.” Lewis Mumford Center for Comparative Urban and Re-gional Research, University of Albany, NY.

Logan, John R., and Brian J. Stults. 2011. “The Persistence of Seg-regation in the Metropolis: New Findings from the 2010 Census.”Census Brief prepared for Project US2010. http://www.s4.brown.edu/us2010.

Madison, James. [1787] 2004. “Federalist No. 10.” In The FederalistPapers. New York: Pocket Books.

Masuoka, Natalie, and Jane Junn. 2013. The Politics of Belonging:Race, Public Opinion, and Immigration. Chicago: Chicago Univer-sity Press.

McAdam,Douglas. 1988.Freedom Summer. New York:Oxford Uni-versity Press.

McClendon, Gwyneth H. 2014. “Social Esteem and Participa-tion in Contentious Politics: A Field Experiment at an LGBTPride Rally.” American Journal of Political Science 58 (2):279–90.

McKenzie, Brian D. 2004. “Religious Social Networks, Indirect Mo-bilization, and African-American Political Participation.”PoliticalResearch Quarterly 57 (4): 621–32.

McPherson, Miller, Lynn Smith-Lovin, and James M. Cook. 2001.“Birds of a Feather: Homophily in Social Networks.” Annual Re-view of Sociology 27: 415–44.

Michelson, Melissa. 2003. “The Corrosive Effect of Acculturation:How Mexican Americans Lose Political Trust.” Social ScienceQuarterly 84 (4): 918–33.

Minnesota Population Center. 2011.National Historical GeographicInformation System:Version 2.0.Database.Minneapolis,MN:Uni-versity of Minnesota. Accessed on March 8, 2015. www.nhgis.org.

Mummolo, Jonathan,andClaytonNall. 2016.“Why PartisansDoNotSort:TheConstraints on Political Segregation.”The Journal of Pol-itics 79 (1): 45–59.

Oliver, J. Eric. 2010. The Paradoxes of Integration: Race, Neighbor-hood, and Civic Life in Multiethnic America. Chicago: ChicagoUniversity Press.

Olson, Mancur. 1965. The Logic of Collective Action: Public Goodsand the Theory of Groups. Cambridge, MA: Harvard UniversityPress.

Omi, Michael, and Howard Winant. 1994. Racial Formation in theUnited States: From the 1960s to the 1990s. New York: Routledge.

Paluck,Elizabeth L.,andDonald P.Green.2009.“Deference,Dissent,and Dispute Resolution: A Field Experiment on a Mass Media

14

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175

What Makes a Good Neighbor?

Intervention in Rwanda.” American Political Science Review 103(4): 622–44.

Panagopoulos, Costas. 2010. “Affect, Social Pressure and ProsocialMotivation: Field Experimental Evidence of the Mobilizing Ef-fects of Pride, Shame and Publicizing Voting Behavior.” PoliticalBehavior 32 (3): 369–86.

Pantoja, Adrian, Ricardo Ramirez, and Gary Segura. 2001. “Citi-zens by Choice,Voters by Necessity: Patterns in Political Mobiliza-tion by Naturalized Latino.” Political Research Quarterly 54 (4):729–50.

Reardon, Sean F., and Ann Owens. 2014. “60 Years After Brown:Trends and Consequences of School Segregation.”Annual Reviewof Sociology 40: 199–218.

Reardon, Sean R., John T. Yun, and Tamela McNulty Eitle. 2000.“The Changing Structure of School Segregation: Measurementand Evidence of Multiracial Metropolitan-Area School Segrega-tion, 1989–1995.”Demography 37 (3): 351–64.

Ridgeway,Cecila,and ShelleyCorrell.2006.“Consensus and theCre-ation of Status Beliefs.” Social Forces 85 (1): 431–53.

Rogers, Todd, Craig R. Fox, and Alan S. Gerber. 2013. “RethinkingWhy People Vote.” In The Behavioral Foundations of Public Pol-icy, ed.Eldar Shafir. Princeton,NJ:Princeton University Press, 91–107.

Rubin, Donald B. 1974. “Estimating Causal Effects of Treatmentsin Randomized and Nonrandomized Studies.” Journal of Educa-tional Psychology 66 (5): 688–701.

Salisbury, Robert H. 1969. “An Exchange Theory of InterestGroups.”Midwest Journal of Political Science 13 (1): 1–32.

Schwartz, Shalom H. 1977. “Normative Influence on Altruism.”In Advances in Experimental Social Psychology, Vol. 10, ed.Leonard Berkowitz. New York: Academic Press, 221–79.

Segura, Gary. 2006. “Symposium Introduction: Immigration and Na-tional Identity.”Perspectives on Politics 4 (2): 227–78.

Sen, Maya, and Omar Wasow. 2016. “Race as a ’Bundle ofSticks’: Designs that Estimate Effects of Seemingly ImmutableCharacteristics.” Annual Review of Political Science 19: 499–522.

Sherif,M. 1936.The Psychology of Social Norms. New York:Harper.Sinclair, Betsy. 2012.The Social Citizen: Peer Networks and PoliticalBehavior. Chicago: University of Chicago Press.

Snyder, Mark, and Julie A. Haugen. 1994. “Why Does BehavioralConfirmation Occur? A Functional Perspective on the Role ofthe Perceiver.” Journal of Experimental Social Psychology 30:218–46.

Stolle, Dietlind, Stuart Soroka, and Richard Johnston. 2008. “WhenDoes Diversity Erode Trust? Neighborhood Diversity, Interper-sonal Trust and the Mediating Effect of Social Interactions.”Polit-ical Studies 56 (1): 57–75.

Takaki, Ronald. 2008.A Different Mirror: A History of MulticulturalAmerica. Boston,MA: Little, Brown and Company.

Tankard, Margaret E., and Elizabeth Levy Paluck. 2016. “Norm Per-ception as a Vehicle for Social Change.” Social Issues and PolicyReview 10 (1): 181–211.

Taylor, Paul, Mark Hugo Lopez, Jessica Martínez, and Gabriel Ve-lasco. 2012. “When Labels Don’t Fit: Hispanics and Their Viewsof Identity.” Pew Charitable Trust: Hispanic Trends Project,April 4, 2012. http://www.pewhispanic.org/2012/04/04/iii-the-american-experience.

Taylor, Paul, Rich Morin, D’Vera Cohn, and Wendy Wang. 2008.“American Mobility: Who Moves? Who Stays Put? Where’sHome?” Pew Research Center. http://www.pewsocialtrends.org/files/2010/10/Movers-and-Stayers.pdf.

Uggen, Christopher, Sarah Shannon, and Jeff Manza. 2012. “State-Level Estimates of FelonDisenfranchisement in theUnited States,2010.”The Sentencing Project.

Uhlaner, Carole J., Bruce E. Cain, and D. Roderick Kiewiet. 1989.“Political Participation of Ethnic Minorities in the 1980s.”PoliticalBehavior 11 (3): 195–231.

U.S. Department of Labor. 1965. “The Negro Family: The Casefor National Action.” Office of Policy Planning and Research,March, 1965. http://www.dol.gov/oasam/programs/history/webid-meynihan.htm.

Velez,Yamil Ricardo, and GraceWong. 2017. “Assessing ContextualMeasurement Strategies.”The Journal of Politics 79 (3): 1084–9.

Verba, Sidney, and Norman H. Nie. 1972. Participation in America:Political Democracy and Social Equality. New York: Harper andRow.

Verba, Sidney, Kay Schlozman, and Henry Brady. 1995. Voice andEquality: Civic Voluntarism in American Politics. Boston, MA:Harvard University Press.

Wheaton,Sarah. 2013. “For First Time onRecord,Black Voting RateOutpaced Rate for Whites in 2012.” New York Times. Accessedon May 8, 2013. http://www.nytimes.com/2013/05/09/us/politics/rate?of?black?voters?surpassed?that?for?whites?in?2012.html?hpr=0.

Wilson, William Julius. 2011. “The Declining Significance of Race:Revisited and Revised.”Daedalus 140 (2): 55–69.

Wolfinger,Raymond E., and Steven J.Rosenstone. 1980.Who Votes?New Haven: Yale University Press.

Wong, David W. S. 2013. “Comparing Traditional and Spatial Segre-gation Measures: A Spatial Scale Perspective.”Urban Geography25 (1): 66–82.

Wong,Cara, Jake Bowers,TarahWilliams,andKatherineDrake Sim-mons. 2012. “Bringing the Person Back In: Boundaries, Percep-tions, and the Measurement of Social Context.” The Journal ofPolitics 74 (4): 1153–70.

15

Dow

nloa

ded

from

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e. V

ande

rbilt

Uni

vers

ity L

ibra

ry, o

n 07

May

201

8 at

14:

04:3

7, s

ubje

ct to

the

Cam

brid

ge C

ore

term

s of

use

, ava

ilabl

e at

htt

ps://

ww

w.c

ambr

idge

.org

/cor

e/te

rms.

htt

ps://

doi.o

rg/1

0.10

17/S

0003

0554

1800

0175