what triggers public opposition to immigration? anxiety ...pscourses.ucsd.edu/ps100da/brader what...

What Triggers Public Opposition to Immigration?Anxiety, Group Cues, and Immigration Threat

Ted Brader University of MichiganNicholas A. Valentino The University of Texas at AustinElizabeth Suhay University of Michigan

We examine whether and how elite discourse shapes mass opinion and action on immigration policy. One popular butuntested suspicion is that reactions to news about the costs of immigration depend upon who the immigrants are. Weconfirm this suspicion in a nationally representative experiment: news about the costs of immigration boosts white oppositionfar more when Latino immigrants, rather than European immigrants, are featured. We find these group cues influenceopinion and political action by triggering emotions—in particular, anxiety—not simply by changing beliefs about theseverity of the immigration problem. A second experiment replicates these findings but also confirms their sensitivity to thestereotypic consistency of group cues and their context. While these results echo recent insights about the power of anxiety,they also suggest the public is susceptible to error and manipulation when group cues trigger anxiety independently of theactual threat posed by the group.

Immigration surged onto the national agenda follow-ing the 2004 election, as politicians wrangled overreforms on what is perceived to be a growing prob-

lem for the United States (U.S.). Public concern followed,with 10% of Americans by 2006 naming it the most im-portant problem facing the country, the highest level in 20years of polling by Pew Research Center. It became a majorissue in the 2008 presidential election. Republican angerat John McCain for his support of the Comprehensive Im-migration Reform Act contributed to the near collapse ofthe early front-runner’s campaign in 2007. A few monthslater, Republicans in the lead-off Iowa caucuses picked im-migration as the country’s most important problem morethan any other issue (including war, terrorism, and theeconomy).1 Debates about immigration have flared at in-

Ted Brader is associate professor of political science, University of Michigan, Center for Political Studies, 4242 ISR, 426 Thompson Street,Ann Arbor, MI 48106-1248 ([email protected]). Nicholas A. Valentino is Mike Hogg professor of community affairs, University of Texasat Austin, Department of Government, 1 University Station A1800, Austin, TX 78712-0119 ([email protected]). Elizabeth Suhayis a doctoral candidate in political science, University of Michigan, Department of Political Science, 5700 Haven Hall, Ann Arbor, MI48109-1045 ([email protected]).

Data collection for this article was made possible by Time-sharing Experiments for the Social Sciences (TESS) and grants from the HowardR. Marsh Center for the Study of Journalistic Performance and the Horace H. Rackham School of Graduate Studies. Antoine Banks, AndreaBenjamin, and Eric Groenendyk provided excellent research assistance. For helpful comments on this research at various stages, we thankJake Bowers, Nancy Burns, Jamie Druckman, Zachary Elkins, Stanley Feldman, Rob Franzese, Marty Gilens, Don Green, Greg Huber,Leonie Huddy, Don Kinder, John Lapinski, Skip Lupia, George Marcus, David Mayhew, Tali Mendelberg, Joanne Miller, Ken Scheve, JohnSweeney, Cara Wong, and anonymous reviewers, as well as audiences and discussants at MPSA and APSA annual meetings, DartmouthCollege, the University of Michigan, New York University, Princeton University, and Yale University.

1Unlike periodic open-ended polling by Pew on the most important problem, entrance and exit polls offer voters a choice among a smallnumber of issues. Pew data are publicly available on their website (http://people-press.org), while marginal results for exit polls can befound on the sites of most national news outlets (e.g., see http://politics.nytimes.com/election-guide/2008/index.html).

tervals throughout U.S. history (Tichenor 2002). Currentepisodes reflect mounting pressures from heavy immi-gration and an expanding Latino electorate. West Europealso has experienced a rising tide of migrants, spurringbitter debates over how to deal with the newcomers anda growth in electoral position taking (Fetzer 2000; Sni-derman, Hagendoorn, and Prior 2004; Sniderman et al.2000). Even though the issue has cut across partisan divi-sions in the past, activists and politicians increasingly aretaking the debate to the American public.

Public debates about immigration, like those in otherdomains, often suggest the interests, values, or lifestylesof citizens are in harm’s way. Elite discourse tends toemphasize adverse consequences for jobs, taxes, crime,schools, cultural norms, or social harmony (Simon and

American Journal of Political Science, Vol. 52, No. 4, October 2008, Pp. 959–978

C©2008, Midwest Political Science Association ISSN 0092-5853

959

960 TED BRADER, NICHOLAS A. VALENTINO, AND ELIZABETH SUHAY

Alexander 1993). It is not surprising, therefore, that op-ponents of immigration always outnumber supporters.Nonetheless, public opinion on the issue can be volatile,with opposition spiking to nearly two-thirds in some yearsand dropping to as little as one-third in others (Lapinskiet al. 1997; Simon and Lynch 1999). Why do Americansfind the “threat” of immigration more or less compelling?As with many issues, Americans tend to be poorly in-formed and uncertain about immigration (Scheve andSlaughter 2001) and much of what they learn comesthrough the mass media. We suspect that the way journal-ists and politicians portray immigration plays a significantrole in activating (or assuaging) opposition.

Our aim is to ascertain which features of public debateaffect popular support for immigration. We begin withan assumption about the essential nature of immigrationas a political issue: it generates conflict between groups—citizens versus noncitizens, English speakers versus for-eign language speakers, whites versus nonwhites (King2000). Discourse and opinion on immigration is “groupcentric” (Nelson and Kinder 1996). Immigration attitudesseem rooted substantially in group identity and prejudice(Citrin et al. 1997; Kinder 2003). In contrast to this lineof work, some studies stress the role of economic factorsin explaining significant individual and cross-nationalvariation in immigration opinion (Quillian 1995; Scheveand Slaughter 2001). Whether prior studies emphasize“symbolic” group factors or “real” costs, they say littleabout how or why opposition to immigration changes,especially in the short term. Sniderman and colleagues(2004), however, show that political attitudes toward im-migration are shaped by “situational triggers” as well aspredisposing factors.

Can ethnic or racial group cues act as situational trig-gers that change how the American public responds to im-migration? Many observers regard as commonplace thenotion that support for immigration depends on who theimmigrants are. Many opponents of immigration claim,however, that opposition is based on the consequencesof immigration, not the identity of immigrants. More-over, prominent recent scholarship has been either silenton the question or found that opinion in some Euro-pean countries does not depend on the race or ethnicityof immigrants (Scheve and Slaughter 2001; Sniderman,Hagendoorn, and Prior 2004; Sniderman et al. 2000). Webelieve immigrant identity matters in the United States.In particular, we suspect that news about the potentialthreat of immigration will have a larger impact on whites’attitudes when the immigrants being discussed are non-white. Testing this suspicion is our starting point, becauseresearchers have yet to establish that group cues affect howindividuals respond to information about immigration.

Ascertaining the existence of such effects, however, doesnot explain why they occur. Our primary goal is to un-derstand why the race or ethnicity of immigrants, aboveand beyond arguments about the consequences of immi-gration, drives opinion and behavior.

If the race or ethnicity of the salient immigrant groupinfluences opinion, there are at least two potential expla-nations. First, group cues might alter public perceptionsabout the severity of the problem. Negative stereotypesabout Latinos in particular might boost concerns aboutcultural assimilation, consumption of scarce public re-sources, crime, and so on. Thus when the news mediahighlight Latinos in discussions about immigration, whitecitizens may come to believe that immigrants pose an evengreater problem than if white Europeans were featured.This strengthening of the perception that immigrationhas a negative impact might then spur greater opposi-tion.

Although perceived harm is a plausible mediator ofgroup cue effects, we suspect a different mechanism is atwork. Racial or ethnic cues may trigger emotional reac-tions, such as anxiety, which may cause changes in opinionand behavior independently of changes in beliefs about theseverity of the immigration problem. This argument differsdramatically from traditional research on political threats,which tends to highlight the impact of cognitively basedperceptions (Miller and Krosnick 2004). Recent researchsuggests that emotions play a central role in both politicaljudgment and behavior (Marcus, Neuman, and MacKuen2000). When anxiety is triggered by threatening cues orconditions, it can facilitate opinion change and motivateinformation seeking and political action. Though priorscholarship posits the emotional power of groups (Kinderand Sanders 1996), ours is the first to test the emotionalimpact of group cues on political attitudes and behavior(cf. Hutchings et al. 2006).

Although our predictions are consistent with re-cent work on anxiety, their normative tone differs. Sev-eral studies stress the potential benefits to learning andrational decision making that follow from anxiety in-creasing openness to new information and decreasingreliance on predispositions (Brader 2006; Marcus, Neu-man, and MacKuen 2000; Valentino et al. 2008). Theseeffects, however, also make people more susceptible tomanipulation if the information at hand is skewed oruntrue. In addition, if anxiety is triggered independentlyof the information provided by stereotypic group cues,the public may react more strongly than the facts war-rant in some situations and not strongly enough inothers.

In this article, we examine how key aspects ofelite discourse influence opposition to immigration. We

ANXIETY, GROUP CUES, AND OPPOSITION TO IMMIGRATION 961

expect that public opinion about immigration depends atleast in part on what citizens read or hear in the news. Weexpect that news stimulates greater opposition when ithighlights costs rather than benefits of immigration. Butwe also think that such reactions are tied to the immi-grant groups made salient in the news. Group cues imbuethe discussion of costs and benefits with emotional sig-nificance. In other words, group images cause changesin attitudes and behavior by triggering an emotional re-action, rather than by simply changing beliefs about theseverity of the problem.

Public Discourse and Opinionon Immigration

Our interest in immigration opinion is driven in part bythe perennial importance of the issue. Debate over im-migration has heated up in recent years, with concernsabout its economic and cultural impact contributing to abacklash against globalization (Cohen 2001; Huntington2004; Scheve and Slaughter 2001). After the 9/11 attacks,the issue prompted new concerns about national security.As politicians and activists bring these concerns furtherinto the public spotlight, their appeals are countered byan increasingly influential and vocal Hispanic minority.In the 1990s, Governor Pete Wilson rode to reelectionin California championing Proposition 187, which de-nied government services to immigrants. This “victory,”however, may have damaged Republican Party prospectsin the nation’s most populous state by alienating His-panic citizens. George W. Bush took the opposite tack byproposing to liberalize immigration policy and receivedrecord levels of support from Hispanics in the 2004 elec-tion. Beginning in 2006, local and national efforts to ex-pand criminal sanctions against illegal immigrants andtheir employers led to massive protests. At the same time,conservative Republicans became increasingly outragedby congressional attempts to pass reform legislation thatwould balance border control with pathways to legal sta-tus for illegal immigrants.

Immigration’s status on the national agenda haswaxed and waned throughout history (Tichenor 2002).Although immigrant nationalities have changed, therehave been striking similarities in the central issues andcharacter of the debate across time. Though rising levelsof immigration carry both costs and benefits, the empha-sis of public discourse tends to be on the former (Simonand Alexander 1993). Politicians and activists have al-ways voiced concerns about the economic and culturalthreat immigration poses to Americans (Cohen 2001).

News coverage from 1995 to 2005 was twice as likely tostress the costs of immigration as the benefits (Pande2006).

Anti-immigration rhetoric also often has group over-tones and, explicitly or implicitly, makes a distinction be-tween stigmatized ethnic or racial groups, such as Hispan-ics, and “good” immigrants, such as those from Canada,Ireland, or Poland (Huntington 2004; King 2000). Again,from 1995 to 2005, 84% of immigration stories in majorpapers mentioned specific groups (Pande 2006). Twice asmany stories referred to Hispanics as the next most men-tioned group (East Asians), and more stories spotlightedHispanic migrants than migrants from all other regionscombined.

Popular attitudes often seem to mirror these patternsin elite discourse. Public opinion about immigration runsheavily toward opposition (Simon and Lynch 1999), witha plurality of Americans preferring to reduce the numberof immigrants or hold it constant. The share of the pub-lic favoring increases in immigration is consistently tiny,usually hovering around 10%. Surveys suggest oppositionis tied to both “realistic” economic concerns and “sym-bolic” cultural concerns, but especially the latter (Citrinet al. 1997; Kinder 2003; Scheve and Slaughter 2001; Sni-derman et al. 2000). Although research has turned upnumerous correlates of opinion, we know little about theeffects of political discourse. There is evidence, however,that opposition to immigration (in the Netherlands) is re-sponsive to situational triggers such as the relative salienceof national identity or of economic versus cultural threats(Sniderman, Hagendoorn, and Prior 2004). Finally, wealso know significantly less about what motivates citizensto act on their attitudes.

In order to understand how public discourse affectspopular opposition to immigration, we examine threecharacteristics of that discourse: emphasis on costs andbenefits, ethnic identity cues, and portrayal of immigrantsas low- or high-skilled workers. Prior research indicatesthat negative consequences and group cues are salient inU.S. news coverage of immigration. References to the oc-cupation or skill level of immigrants are not rare, buttypically less pronounced than the other two factors.Nonetheless, many economic models view skill level ascritical to workers’ perceptions or experience of the harmthey face from migrants (Scheve and Slaughter 2001). Inaddition, ethnic identity cues, such as pictures of His-panic immigrants, may signal more than ethnicity. Todaythe popular image of immigrants in the American mind isbased on the stereotype of low-skilled Hispanic laborers.Perhaps reactions to Hispanic immigrants are primar-ily based on the economic skills and needs they are per-ceived as bringing, rather than on their ethnicity per se. In


order to isolate better the effect of ethnic cues, we try todisentangle them from skill level.

The Mediating Role of Beliefsand Emotions

Our second major goal in this article is to examine thepsychological mechanism underlying the effects of groupcues in immigration discourse. Why does the salience ofparticular ethnic or racial groups in elite debate affectpublic opinion? We already have considerable evidencethat such cues influence opinion in other policy domainsby priming attitudes toward the group in question. Briefreferences to groups that benefit from a policy as wellas subtle visual cues in news or political ads can causegroup attitudes to play a larger role in opinion formation(Gilens 1999; Mendelberg 2001; Nelson and Kinder 1996;Valentino, Hutchings, and White 2002). Group primingmay work differently, however, in the explicitly group-based context of immigration, a possibility we exploreelsewhere (Brader and Valentino 2005). If group primingis most evident when cues are subtle (Mendelberg 2001;Valentino, Hutchings, and White 2002), then any effectof the relatively explicit cues in immigration discoursemay function by directly triggering emotions or changingperceptions of threat.

We consider the extent to which immigrant groupcues change political attitudes and behavior by alteringbeliefs about consequences, triggering emotions, or both.One classic route to persuasion is by changing the be-liefs or perceptions about the world that form a basisfor attitudes (Eagly and Chaiken 1993). Belief changedoes not require exposure to direct arguments. Ethnic orracial group cues may carry implications for how cit-izens perceive the nature or severity of any harm (orbenefit) from immigration. For example, migrants fromnon-English-speaking countries might suggest culturaltensions, whereas migrants from poor countries mightsuggest economic competition and a strain on public cof-fers. A citizen’s policy preference may depend on suchfactual perceptions, on what she believes to be true aboutthe policy and the problem it is meant to address. Thus,if group cues change beliefs about the severity of the im-migration “problem,” opinion change may follow.

A long line of research also suggests that elicitingemotions can motivate changes in attention, opinion,and behavior (Eagly and Chaiken 1993). For example,Hot Cognition theory contends that almost any object—social groups, politicians, policy issues—triggers auto-matic positive or negative affect based on a person’s past

evaluations of the object. This initial emotional reac-tion biases subsequent information processing in a di-rection consistent with the initial reaction (Lodge andTaber 2005). Another approach, Affective Intelligencetheory, distinguishes between two basic emotional sys-tems (Marcus, Neuman, and MacKuen 2000). The “dis-position system,” which generates enthusiasm based onfeedback about a person’s progress toward her goals, reg-ulates interest and reliance on habit. The “surveillancesystem” generates anxiety or fear in response to poten-tial threats. Anxiety in turn prompts information seek-ing and a greater focus on contemporary information,thereby increasing the likelihood of learning and per-suasion (Valentino et al. 2008). Because anxiety focusesattention on available information, the direction of per-suasion depends on the implications of that information(e.g., whether it suggests immigration is harmful or ben-eficial). Finally, since emotions are motivational impulses(Lodge et al. 2006), they also make it more likely thatindividuals will take political action (Brader 2005, 2006;Valentino et al. 2006). Might outgroup cues in news aboutimmigration trigger anxiety? We think so. Although polit-ical scientists have only just begun to research “emotionaltriggers” in politics, there is a good deal of work in psy-chology demonstrating the capacity of groups to evokeemotions (Mackie and Smith 2003).

Of course, belief change and emotional responsesare not mutually exclusive. Both may mediate the effectsof group cues on immigration opinion. After all, beliefsabout the severity of immigration problems and emo-tions about that same topic are often strongly correlated,and each may affect opinion formation (Isbell and Ottati2002). We expect that, overall, elite debate about immi-gration influences perceptions about the severity and na-ture of the immigrant “threat” and elicits anxiety or otheremotional reactions. Nonetheless, prior scholarship onthreat has often looked either at the role of perceptions ofharm or anxiety, but rarely both. Where both have beenconsidered, scholars have offered differing assessments,such as that perceived threat and anxiety are closely re-lated elements in the same causal chain (Marcus et al.1995, 2005), that perceived threat matters while anxietydoes not (Miller and Krosnick 2004), or that perceivedthreat and anxiety are strongly correlated but have dis-tinct effects (Huddy et al. 2005).

We expect that cues to the identity of immigrants willprincipally influence public opinion and political actionby triggering anxiety or other emotions. Ethnic and racialgroups are potent, affect-laden symbols in American so-ciety (Mackie and Smith 2003; Sears 2001). Moreover,in the ordinary course of politics, citizens often arrive atjudgments without devoting much attention or effort to


learning, remembering, and sifting through details. Theyare guided instead by “gut” reactions and habits of mind(Lodge and Taber 2005; Marcus, Neuman, and MacKuen2000). Under typical circumstances, therefore, we expectthe emotional impact of group cues to shape opinion andbehavior, even though it is plausible for such cues to af-fect opinion by changing perceptions about the nature orseverity of a threat. To the extent group cues trigger anxi-ety about immigration, we expect changes in opinion andaction consistent with the information at hand. Likewise,if anxiety makes opinion change and action more likely,then we expect anxiety to mediate these observed effectsof outgroup cues.

Our expectations depart in two ways from the posi-tive normative implications of anxiety stressed in Affec-tive Intelligence theory. First, while the theory predictsopenness to new information when anxiety is aroused,people should be open to available arguments, not neces-sarily valid ones.When people focus on the information athand, they are also more vulnerable to the ways in whichthat information can be skewed or manipulated.This maylead to undesirable outcomes. Second, we argue that trig-gers for anxiety may, but need not, correspond to realisticor valid threats. Just as campaign ads evoke emotionswith staged imagery often unrelated to the political mes-sage (Brader 2006), news reports may trigger emotionsnot only with the “hard” information they provide, butalso with the way in which the story is packaged. Thus,group cues in immigration news may trigger anxiety in-dependently of actual information about the severity ofthe threat. Although fear of outgroups is not necessarilyirrational, it may cause individuals to react more stronglythan “facts” warrant as well as to treat equivalent situa-tions unequally (i.e., based solely on the salient group).Such distortions in public opinion can in turn provideincentives, or justifications, for officials to enact biased oroverreaching public policies.

Hypotheses

Our main theoretical goals then are (1) to examine theimpact of group cues in immigration discourse on per-ceptions of threat and emotions and (2) to test whetherchanges in those perceptions, emotions, or both mediatethe impact of discourse on opinion and behavior. The the-oretical discussion yields three main hypotheses. First, theimpact of information about the consequences of immi-gration on attitudes and behavior will depend on whichimmigrant groups are salient. We expect greater opposi-tion when news emphasizing the harm from immigration

contains images of Latino migrants than when it portrayswhite European migrants.

Next we will explore why, if at all, such an ethno-racial effect occurs. To do so, we will examine the impactof news on beliefs and emotions about immigration. Weknow from decades of research that individuals who areexposed to a one-sided message from a trusted source aremore likely to accept the message content as true (Eaglyand Chaiken 1993). Therefore, few would be surprisedif a credible news report that stressed the costs of immi-gration increased the belief that immigration is harmful.Given their highly symbolic nature (Sears 2001), groupcues should be potent triggers of emotion. Specifically,outgroup cues should be a potent source of anxiety innews about immigration. In one sense, all immigrants aremembers of an outgroup (i.e., “foreigners”). But the his-tory of immigration makes clear that some immigrants,those ethnically or racially distinct from the native major-ity, are regarded as more alien than others (King 2000). Weexpect white Americans to respond with stronger, morenegative emotions when elite debate focuses on Latino im-migrants. Our second hypothesis, therefore, is that whileany news emphasizing the costs of immigration shouldincrease the perception that immigration is harmful toAmericans, it will be more likely to trigger anxiety whenthe story calls attention to Latino, as opposed to whiteEuropean, immigrants.

Opinions and behavior in turn may change in re-sponse to shifting perceptions of harm, emotional reac-tions, both, or neither. We expect most people to updatetheir opinions and take new action only when their rou-tine is interrupted by the experience of anxiety (Brader2006; Marcus, Neuman, and MacKuen 2000). Thus, ourthird hypothesis is that opinion change, information seek-ing, and political action should be more likely to the extentnews about immigration arouses anxiety; in other words,anxiety should mediate these effects. According to Affec-tive Intelligence theory (Marcus, Neuman, and MacKuen2000), anxiety should directly stimulate a search for infor-mation relevant to the threat and motivate political actionaimed at removing the threat. Because anxiety facilitatespersuasion indirectly, however, by shifting attention toavailable information, the direction of opinion changeshould be consistent with that information. For example,if Latino cues increase anxiety, as our second hypothesissuggests, then we expect those anxious citizens to becomemore opposed to immigration after reading threateningnews about the issue. If those anxious citizens, however,encountered reassuring news, then they should becomemore supportive (or less opposed) to immigration.

Finally, on one point we are less certain. Will we infact observe increased anxiety if stigmatized outgroup


cues accompany otherwise reassuring information aboutimmigration (e.g., that it has beneficial consequences forAmericans or that immigrants bring more skills/assetsthan expected)? Given the automatic nature of emotionaltriggers (Damasio 2000), we expect that such group cues,once encoded with such negative affect, will initiate ananxious response in all situations. However, this auto-matic reaction is also revised and updated continuouslyon the basis of new information (Lazarus 1991; LeDoux1996), so that anxiety may dissipate quickly, perhaps evenbefore an individual is consciously aware of the feeling, ifthe trigger is accompanied by reassuring signals. This per-spective does not leave us with a clear prediction aboutwhether anxiety will change in response to Latino cuesembedded in more reassuring news about immigration.However, research on political tolerance suggests thatanxiety is more likely when disliked political groups areportrayed in a threatening, rather than reassuring, con-text (Marcus et al. 1995). A second possibility is that theemotional “trigger” is not the ethno-racial cue itself, butrather a stereotypic image of that group. Recent researchfinds that cognitive processes such as priming (Mendel-berg 2001; Valentino, Hutchings, and White 2002) andemotional processes (Mackie and Smith 2003) can de-pend on stereotype consistency. Thus, news inconsistentwith stereotypes about Latinos may elicit less anxiety thanstereotype-consistent stories. If this is the case, we shouldexpect greater anxiety when Latino immigrants are por-trayed as low-skilled workers with potentially harmfulconsequences for American society than when they areportrayed as high-skilled or potentially beneficial. Thecompeting hypothesis, in this case, is that Latino immi-grants will trigger greater anxiety regardless of how theyare portrayed.

To test the hypotheses, we conducted two experi-ments in the summer and fall of 2003. Experiments al-low us to distinguish the causal influence of informa-tion about costs and benefits of immigration, informa-tion about immigrant skill level, and cues of immigrantethnicity. The data also shed light on the psychologicalmechanism. Perceptions of harm and anxiety are oftenhighly correlated, and self-reports of each are affected bypredispositions. In an experiment, however, we can fo-cus on changes in perceptions or emotions in response tomanipulations of news content and, in so doing, discernbetter which are mediating any observed effects.

Experiment 1

We embedded an experiment in a nationally represen-tative survey of 354 white, non-Latino adults. The sur-

vey was conducted via WebTV by Knowledge Networks,which maintains a large, randomly contacted respondentpool by offering free Internet access in exchange for occa-sional participation in surveys. The company randomlyselects participants from its pool for each survey.2 Subjectsfrom 46 states took part in ours. The median interviewwas 16 minutes. The median subject was 46 years old withsome college and a $45,000 household income. Fifty-twopercent of the subjects were women. Thirty-one percentidentified as Republicans, and 28% as Democrats.

Prior to the experiment, subjects were told they wouldtake part in “a survey regarding a number of current socialissues.” The survey first measures political predispositionsand economic outlook. Immediately prior to the stimulus,subjects were told that we wanted their reactions to a storythat “had been in the news lately.” After reading, subjectsanswered questions about immigrants and immigration.An appendix provides wording for all questions usedhere.

The web-based platform allows us to deliver stim-uli matching those in actual news coverage. The studyemployed a 2 × 2 design with a control group. We ma-nipulated the ethnic cue by altering the picture and nameof an immigrant (white European versus Latino) featuredin a mock New York Times report about a governors’ con-ference on immigration.3 We also manipulated the toneof the story, focusing either on the positive consequencesof immigration for the nation (e.g., strengthening theeconomy, increasing tax revenues, enriching Americanculture) or the negative consequences (e.g., driving downwages, consuming public resources, undermining Amer-ican values). Tone was also manipulated by portrayal ofgovernors as either glad or concerned about immigra-tion and citizens who had had either positive or negativeinteractions with immigrants.4 Every story stated thatimmigration to the United States is increasing and willcontinue to do so.5

Our goal in manipulating group salience was to testthe impact of the ethnic identity cue itself. We strove

2According to information provided by Knowledge Networks, par-ticipation rate in their panel at the time was 56%. The completionrate among panelists contacted for our survey was 77%.

3The complete text of the stimulus materials is available uponrequest.

4Subjects in the control group read an article on cell phone useand car crashes. In our analyses, we determine effects by runninganalyses of variance (ANOVA) on the 2 × 2 factorial design of thetreatments; the control group is not included in those analyses.We provide control group means in the figures as a reference for“baseline” levels of opinion or behavior.

5Cell sizes: 72 (control), 73 (neg. Lat.), 72 (neg. Eur.), 69 (pos. Lat.),68 (pos. Eur.).


to vary the cue in a way that held constant any newinformation about the nature and severity of immigra-tion’s impact. Each article contained a picture of a “recentimmigrant” who appears to be either Latino or Euro-pean. A caption read: “[Jose Sanchez/Nikolai Vandin-sky] is one of thousands of new immigrants who ar-rived in the U.S. during the first half of this year.” Tomaximize control, a graphic artist altered the images sothat only the heads differed (i.e., bodies, dress, and back-ground were identical). We selected two faces from 40photos based on ratings by an independent panel of eightnaı̈ve judges. Judges rated how much each male appearedethnically European or Hispanic, as well as wealthy, at-tractive, friendly, and law abiding. The faces chosen forour manipulation were maximally distinct on the di-mension of ethnicity, but statistically indistinct on othertraits.6

The article included a quote (identical across con-ditions) from the men on their feelings about comingto the United States and identified their national ori-gin as Mexico or Russia. We named specific countries toconform with standard news practice and, more impor-tantly, to maintain control over what subjects called tomind. Prior research has shown that when informationis omitted, subjects tend to “fill in” the details (Gilliamand Iyengar 2000). For the Latino immigrant, we ex-pected whites to assume Mexican nationality even if itwere not mentioned. We were less certain about the as-sumptions respondents would make about a fair-skinnedimmigrant who appears to be of European descent. Wechose Russia for several reasons. First, most Americanshave heard of it; picking a country many do not know(e.g., Slovenia) would be equivalent to naming no coun-try at all. Second, we wanted to hold the fact of lan-guage differences constant. Third, we ruled out countrieslinked to earlier waves of immigrants with which manywhite Americans identify, such as Ireland or Italy. Fi-nally, working-class Russians have been emigrating tothe United States in fairly large numbers but are notbranded with a racialized stigma (Rumbaut and Portes2001). To the extent Russian nationality has negative as-sociations, ours is a conservative choice that may dampeneffects.

6The mean rating for “how ethnically Hispanic” (1–5 scale) the manappears is 3.9 for the Latino photo and 1.0 for the European photo.This difference is highly significant (t = 8.40, p < .001). The meanrating for “ethnically European” is 1.4 for the Latino photo and 4.4for the European photo. This difference is also highly significant(t = 9.72, p < .001). On the four other trait dimensions, thedifference in mean ratings did not exceed 0.4 points nor approachstatistical significance.

The Impact of News on Opinion,Information Seeking, and Action

Our first hypothesis is that news emphasizing costs shouldboost opposition to immigration more powerfully whenit is accompanied by images of stigmatized groups thanwhen it is accompanied by racial ingroup members. Totest this hypothesis, we examine effects on both opin-ion and political behavior, starting with the former. Wetapped immigration preferences via a standard questionabout whether immigration to the United States shouldbe increased, decreased, or kept the same. We also askeda more specific policy question: whether English shouldbe the official language of the country. Figure 1 displaysthe mean opinion across conditions.

There is a main effect on overall preferences for thenews story’s emphasis on consequences (Mdiff = .42, F =11.92, p < .001) and an interaction effect between em-phasis and ethnic cues (F = 3.24, p < .073).7 Respondentsin the negative (i.e., immigration is costly) Latino con-dition expressed the most opposition to immigration,while those in the positive Latino condition expressed themost support. Although emphasis on costs versus bene-fits works similarly for the European and Latino versions,the negative effect is much larger for stories with Latinocues, consistent with our first hypothesis. The impact ofthe threatening story is over two times stronger whenLatino cues are present.8 This effect is substantively large,shifting opinion 15% of the scale. The interactive patternof support for English-only laws is similar. The negativeLatino story boosts opposition by 7% of the scale, andthe interaction is marginally significant (F = 3.80, p <

.052).Next we examine the impact of our news stories on

political behavior. Our goal was to improve upon typicalsurvey self-reports of willingness to engage in activity. Ouritems were designed so that subjects would believe theywere requesting information from real organizations ortaking action on the issue. First, we asked if they would likemore information about immigration from a variety ofsources, including nonpartisan research centers, the U.S.government, pro-immigrant groups, and anti-immigrantgroups. This question taps information-seeking behavior

7All reports of statistical significance are based on two-tailed testsand controls for education, age, income, and sex. Despite random-ization, the distribution across cells is a bit uneven (e.g., subjectsin the positive European cell are slightly better educated than thosein other cells (t = 2.13, p < .03)).

8If the effect of the news frame is analyzed separately for bothversions, the boost in opposition falls short of significance whenpaired with the European immigrant cue (t = 1.49, p < .139) andis strongly significant when paired with the Latino cue (t = 3.49,p < .001).


FIGURE 1 Impact of News Frame and EthnicCue on Immigration PolicyPreferences

Note: Each graph displays the mean policy opinion of the fourtreatment conditions. The “baseline” control group mean valuesfor each panel are 2.65 (top) and 2.44 (bottom).

by leading subjects to believe they would be contacted bythe group(s). Second, subjects could request that an emailmessage be sent on their behalf to members of Congressindicating their support for a law reducing the numberof new immigrants. This item measures political mobi-lization by invoking real contacting of elected officials.Subjects were told the email would contain their name,city, and opinion on the issue.

In all, 28% of respondents requested informationand half of these requested it from more than one source.Despite the diversity of sources, the decision to requestmaterials from any source was positively correlated with

FIGURE 2 Impact of News Frame and EthnicCue on Information Seeking andPolitical Action

Note: The “baseline” control group mean values for each panel are0.51 (top), 0.14 (middle), and 0.31 (bottom).

the decision to do so from others. We therefore count thenumber of sources requested as a measure of general in-formation seeking. Figure 2 (top panel) displays the meannumber of requests by respondents in each condition. Thepattern only partially matches expectations. The slope forLatino cues is similar to what we have seen previously:respondents request nearly three times as much informa-tion on average when the news is negative. The slope runs


the other direction, though insignificantly so, for Euro-pean cues. Thus the interaction effect is significant (F =4.05, p < .045). Surprisingly, however, the European cueproduces higher demand for information compared tothe Latino cue, regardless of the frame (Mdiff = .31, F =8.44, p < .004).

In an effort to clarify this unanticipated main effect,we compare the treatments to the global control group,in which subjects did not read a story about immigration.The mean number of requests in the control group is 0.51,suggesting the negative Latino cell is closest to “baseline”level of requests. The greatest departure from this baselineis the positive Latino cell (t = 2.63, p < .009). It appears,therefore, that pairing Latino immigrant cues with a pos-itive message about immigration actually suppresses thedemand for information. While the overall pattern of re-sults for Latino cues is consistent with our expectations,their relation to the results for European cues is surpris-ing. The latter appear to push information seeking abovethe baseline. For now, we are left to speculate that the Eu-ropean immigrant cue is sufficiently novel that it piquesthe respondent’s curiosity, particularly when benign con-sequences are emphasized.9

We see the same pattern for information requestsfrom anti-immigration groups. The negative Latino storyboosts the proportion who request information from .01to .10, while the negative European story again suppressesrequests from .20 to .13. The interaction effect is signifi-cant (F=4.97, p< .027). Once again, information seekingis dramatically suppressed from the control group base-line of 0.14 when Latino cues are paired with positivenews (t = 2.81, p < .006). The main effect for ethniccue is also significant (Mdiff = .11, F = 10.08, p < .002),with European cues producing higher levels of interestcompared to Latino cues regardless of the frame. In sum,the results for information seeking do not conform asneatly to expectations as other results. The European cueproduces unexpectedly high information demand. How-ever, the pattern for information seeking when Latinocues are present matches that for opinion change: neg-ative Latino cues result in higher demand than positiveones.

Finally, we examine whether the news story affectsa realistic measure of political action. The third panel inFigure 2 displays the proportion of respondents in eachcondition choosing to send a message to Congress sup-porting a reduction in immigration to the United States.

9This possibility merits future investigation. People may seek infor-mation because they need it, like it, or find it novel or surprising. Wealso checked whether information seeking, especially in the posi-tive frame or European cue stories, might be tied to enthusiasm,but it is not.

The now-familiar interaction reemerges (F = 5.38, p <

.021), with the negative Latino story mobilizing oppo-sition to immigration. The effect is substantial as well:over 45% of respondents in the negative Latino conditionsend an anti-immigration message, 19 percentage pointshigher than the positive Latino condition and 14 pointshigher than the control group. In contrast, the tone ofthe story has no effect on contacting when the Europeanimmigrant is pictured.

Beliefs about Potential Harmversus Emotional Reactions

So far our results confirm the common speculation thatresponses to immigration news depend on the immigrantgroup made salient. The next important theoretical ques-tion is why this happens. Do people believe Latino im-migration is worse than immigration from other groups?Or is the effect grounded in more visceral (automatic),emotional reactions? Our second hypothesis states that,while news about the costs and benefits of immigrationmay affect its perceived impact, Latino cues should triggeranxiety. Shortly after reading the story, subjects reportedtheir views on the likelihood that immigration will havea “negative financial impact on many Americans” and/ora “negative impact on the way of life in many Ameri-can communities.” Summed responses form a perceivedharm scale (� = .82). Figure 3 (top) shows the effect ofthe news manipulations. The main effect of news empha-sis on perceived harm from immigration is large (13%of the scale) and significant (F = 12.83, p < .001). Con-sistent with prior research (Holbrook et al. 2001), theframing effect is asymmetric: negative information pro-duces a sizeable increase over baseline levels (Mcontrol =3.59), while perceived harm in the benefits frame is nearlyidentical to that in the control group. We observe similarresults for perceptions of economic and cultural harmwhen analyzed separately.

What about emotions? Subjects completed a post-test battery of six questions asking how they “feel” aboutincreased immigration—anxious, proud, angry, hopeful,worried, excited. Individuals often report a mix of fearand anger that seems to constitute the sort of anxietyresponse discussed by Marcus, Neuman, and MacKuen(2000). But recent work also draws attention to the factthat anger can emerge as a distinct response from anx-iety, with distinct effects (Huddy, Feldman, and Cass-ese 2007). We checked for differentiation in the self-reports and effects of anxiety and anger, but found theyemerge as part of a common factor and perform similarlywhen analyzed separately. Therefore, we add responses for


FIGURE 3 Impact of News Frame and EthnicCue on Perceived Harm fromImmigration and EmotionalReactions to Immigration

Note: The “baseline” control group mean values for each panelare 3.59 (top), 3.63 (middle), and 1.64 (bottom).

anxious, worried, and angry into a single scale of anxiety(� = .85).10 We examine enthusiasm responses as well, by

10This is consistent with most work on Affective Intelligence the-ory (Marcus et al. 2000). However, our battery is not well suitedto detecting anger as distinct from anxiety because it only containsone anger term (Marcus et al. 2006). Thus, one could reinterpretthe scales as negative and positive affect, rather than as anxiety andenthusiasm specifically. We did not design the study to test com-

combining the proud and hopeful items into a scale (� =.78).11

The middle and lower panels in Figure 3 show theresults for emotions. A main effect emerges for the cost-benefit emphasis of the news on both anxiety (Mdiff =1.16, F = 10.53, p < .001) and enthusiasm (Mdiff = .45,F = 4.07, p < .045). Consistent with the second hypoth-esis, however, respondents report much higher levels ofanxiety about immigration when an emphasis on costs ispaired with the image of a Latino immigrant than whenit is paired with a European immigrant (F = 6.68, p <

.009).12 When the immigrant is Latino, the negative storyboosts anxiety by over 1.8 points on a 0–9 scale (i.e.,20%) compared to the positive story (t = 4.16, p<.001).When the immigrant is European, the negative news storyincreases anxiety by only one-sixth that amount, a statis-tically insignificant difference (t = 1.04, p < .298). Incontrast, in the third panel of the figure, there is no inter-action effect on enthusiasm (F = 0.90, p < .344). Theseresults suggest that white Americans don’t necessarily be-come anxious when reminded about immigration and itsharmful potential, nor when reminded that some immi-grants are nonwhite. They experience substantially greateranxiety when negative consequences of immigration arepaired with a stigmatized outgroup.13 Furthermore, en-thusiastic reactions do not depend at all on the ethniccues paired with news about the issue.

Do Emotions or Beliefs Mediate the Effectsof Outgroup Cues on Opinion

and Behavior?

We have found so far that bad news containing Latinocues strongly drives opposition. We also found that eth-nic cues do not change beliefs about the severity of theimmigration problem. Anxiety, however, is powerfullytriggered by the same kind of news that moves opin-ion the most. Is anxiety mediating the effect of racial-ized news coverage on opinion about immigration? Are

peting accounts of emotional structure, so factor analysis cannotdefinitively adjudicate between one-, two-, or multidimensionalmodels.

11In principal factors analysis, “excited” did not load as stronglyon the enthusiasm factor and seemed to tap general arousal. Wedropped it, though its inclusion does not change our findings.

12The effect is identical for anger (F = 5.59, p < .019) and anxiety(F = 5.52, p < .020).

13The difference in results for beliefs and emotions cannot be ex-plained by a ceiling effect for perceived harm (the scale is 0 to 6) norby the distribution of responses (the standard deviation accountsfor a nearly identical proportion of each scale, differing by less thanone percent).


the more anxious citizens the same ones whose opinionschange and who are motivated to take action? We expectthe effects of threatening news are mediated by negativeemotions.

For mediation to occur, the explanatory variable mustaffect both the dependent variable and the proposed me-diating variable, and the effect of the explanatory variableon the dependent variable must diminish when the me-diating variable is controlled. Table 1 displays results forfour dependent variables. Similar evidence of emotionalmediation occurs for all the dependent variables exam-ined in this study. The first column of estimates showsthe impact of the experimental manipulations on anxi-ety. In the rest of the table, the first of two columns foreach outcome variable shows the unmediated effects ofthe experimental manipulations. Our primary interest isin the second of the two columns for each outcome vari-able, when anxiety is added to the equation. In every case,anxiety about immigration influences opinions and ac-tion. More importantly, there is a sizeable drop in theinteraction term when anxiety is in the model. In orderto determine our confidence in these results, we employthe standard Sobel test for mediation (Baron and Kenny1986; MacKinnon et al. 2002). As Table 1 indicates, the in-teraction effect is mediated significantly by anxiety for allfour variables. The conjunction of Latino cues and nega-tive news about immigration influenced levels of anxiety.Anxiety then caused shifts in policy attitudes, informationseeking, and political action.

Belief about the severity of the immigration problemdoes not mediate the interactive effect of ethnic cues andnews emphasis. This is apparent from the fact that theconditions for mediation are not met (i.e., the explana-tory variable does not cause changes in the mediator).Nonetheless, we confirm this by repeating the mediationanalyses, substituting perceived harm for anxiety. Table 2reminds us there is a main effect of news emphasis onperceived harm, but no interaction effect. Thus, whileperceived harm strongly correlates with opinions and ac-tions, controlling for it does not significantly diminishthe interaction effect. In each case, the interaction ef-fect remains significant and the Sobel test for mediationdoes not approach significance. Although respondentsperceived a threat from immigration, these perceptionsdid not mediate effects found in this study.14 A change inbeliefs about the harm from immigration was not suffi-cient to change opinions or behavior.

14Similarly, enthusiasm neither meets the conditions for mediationnor passes a Sobel test.

Alternative Explanations for the EmotionalSignificance of Group Cues

Consistent with our expectations, we find that stigmatizedgroup cues—in this case, Latinos—imbue news aboutimmigration with greater emotional significance. This inturn increases the impact of the news story on attitudesand behavior. Up until now, we have assumed on the basisof prior research that Latino cues generate stronger emo-tions because of negative affect stemming from ethnocen-trism (Kinder 2003) or racial prejudice (Sears 2001). ButLatino cues may hold emotional significance for some rea-son other than the racial or ethnic stigma attached to thatgroup by white Americans. Perhaps the emotional signif-icance stems instead from the size or economic positionof the group, since Latinos constitute the largest currentpool of immigrants to the United States and are associ-ated with a poorer, less skilled workforce. The questionof why white Americans react so negatively to a poten-tial threat from immigration when Latino (rather thanEuropean) immigrants are salient holds clear substantiveimplications.

An alternative explanation for our results begins withthe claim that European and Latino cues trigger distinctbeliefs about the impact of immigration. Specifically, re-sponses may be tied to beliefs about the sheer size of eachgroup in the current pool of immigrants and/or about therelative skill level of each group. In the first case, the beliefthat Latinos make up a larger share of recent immigrantsto the United States might lead citizens to take the po-tential threat from immigration more seriously when it islinked to Latinos (Quillian 1995). In the second case, thestereotype that Latino immigrants are poor and unskilledmight lead citizens to worry about the greater costs to so-ciety associated with an influx of less skilled labor (Scheveand Slaughter 2001).

We can use data from our initial experiment to ad-dress the possibility that reactions are proportional to thesize of the immigrant group. First, perceptions of threatshould not have increased equally in reaction to bothEuropean and Latino cues, as they did in the first exper-iment, if respondents linked the impact of immigrantsto the size of the group. Nonetheless, the survey askedrespondents for their “best guess” as to the percentageof recent immigrants who have come from Latin Amer-ica and from Europe. The mean guesses were 55% fromLatin America and 20% from Europe, though individualguesses were highly variable.15 If the ability of Latino cuesto modulate levels of anxiety stems from the perception

15There is no right answer. But, if one looks at foreign-born U.S.residents in 2000, 51% were from Latin America and 14.3% werefrom Europe (U.S. Census Bureau 2001).


TA

BL

E1

An

xiet

yM

edia

tes

the

Inte

ract

ion

Eff

ecto

fEth

nic

Cu

esan

dC

ost-

Ben

efit

Em

ph

asis

onA

nti

-Im

mig

rati

onA

ttit

ud

esan

dB

ehav

ior

(Exp

erim

ent1

)

Eff

ecto

nM

edia

tor

Eff

ects

onP

olit

ical

Att

itu

des

and

Beh

avio

r

An

xiet

yab

out

Dec

reas

eSu

pp

ortE

ngl

ish

Req

ues

tAn

ti-I

mm

igra

tion

Sen

dA

nti

-Im

mig

rati

onIm

mig

rati

onIm

mig

rati

onO

nly

Info

rmat

ion

Mes

sage

toC

ongr

ess

Em

phas

is(c

osts

)0.

020.

230.

20–0

.14

–0.1

9–0

.34

–0.4

6–0

.16

–0.2

1(0

.05)

(0.1

9)(0

.19)

(0.2

0)(0

.21)

(0.2

7)(0

.29)

(0.2

3)(0

.24)

Cu

es(L

atin

o)–0

.06

–0.1

9–0

.12

–0.3

4∗–0

.29

–1.4

6∗∗∗

–1.3

9∗∗∗

–0.3

7–0

.28

(0.0

5)(0

.19)

(0.1

9)(0

.20)

(0.2

0)(0

.44)

(0.4

4)(0

.23)

(0.2

4)C

ues

×E

mph

asis

0.18

∗∗∗

0.49

∗0.

200.

52∗

0.34

1.25

∗∗0.

98∗

0.76

∗∗0.

50(0

.07)

(0.2

7)(0

.28)

(0.2

8)(0

.29)

(0.5

3)(0

.54)

(0.3

2)(0

.34)

An

xiet

yab

out

2.25

∗∗∗

1.38

∗∗∗

1.75

∗∗∗

1.76

∗∗∗

Imm

igra

tion

(0.2

7)(0

.28)

(0.4

4)(0

.31)

Sobe

lTes

t(z

)2.

45∗∗

∗2.

28∗∗

2.16

∗∗2.

34∗∗

R2

(adj

ust

ed)

0.18

%P

redi

ctC

orre

ctly

46.1

57.1

56.8

56.1

89.2

93.2

69.6

71.1

%Im

prov

ed7.

718

.70.

0–0

.70.

64.

63.

34.

8N

282

280

280

280

280

278

278

280

280

Not

e:C

oeff

icie

nts

(an

dst

anda

rder

rors

)ar

ees

tim

ates

from

ordi

nar

yle

ast-

squ

ares

regr

essi

on(f

orA

nxi

ety)

,m

axim

um

likel

ihoo

des

tim

atio

nof

orde

red

prob

itm

odel

s(f

orD

ecre

ase

Imm

igra

tion

and

Supp

ort

En

glis

hO

nly

),an

dm

axim

um

likel

ihoo

des

tim

atio

nof

prob

itm

odel

s(f

orR

equ

est

An

ti-I

mm

igra

tion

Info

rmat

ion

and

Sen

dA

nti

-Im

mig

rati

onM

essa

geto

Con

gres

s).A

llm

odel

sin

clu

deco

ntr

ols

for

edu

cati

on,i

nco

me,

age,

and

gen

der.

∗ p<

.10,

∗∗p

<.0

5,∗∗

∗ p<

.01

(tw

o-ta

iled

).


TA

BL

E2

Per

ceiv

edH

arm

Doe

sN

otM

edia

teth

eIn

tera

ctio

nE

ffec

tofE

thn

icC

ues

and

Cos

t-B

enef

itE

mp

has

ison

An

ti-I

mm

igra

tion

Att

itu

des

and

Beh

avio

r(E

xper

imen

t1)

Eff

ecto

nM

edia

tor

Eff

ects

onP

olit

ical

Att

itu

des

and

Beh

avio

r

Per

ceiv

edH

arm

Dec

reas

eSu

pp

ortE

ngl

ish

Req

ues

tAn

ti-I

mm

igra

tion

Sen

dA

nti

-Im

mig

rati

onfr

omIm

mig

rati

onIm

mig

rati

onO

nly

Info

rmat

ion

Mes

sage

toC

ongr

ess

Em

phas

is(c

osts

)0.

09∗

0.23

0.02

–0.1

4–0

.30

–0.3

4–0

.46

–0.1

6–0

.39

(0.0

5)(0

.19)

(0.2

0)(0

.20)

(0.2

1)(0

.27)

(0.2

8)(0

.23)

(0.2

5)C

ues

(Lat

ino)

–0.0

5–0

.19

–0.1

5–0

.35∗

–0.3

0–1

.46∗∗

∗–1

.42∗∗

∗–0

.37

–0.2

8(0

.05)

(0.1

9)(0

.19)

(0.2

0)(0

.20)

(0.4

4)(0

.46)

(0.2

3)(0

.26)

Cu

es×

Em

phas

is0.

060.

49∗

0.48

∗0.

53∗

0.49

∗1.

25∗∗

1.15

∗∗0.

76∗∗

0.68

∗∗

(0.0

7)(0

.27)

(0.2

8)(0

.28)

(0.2

9)(0

.53)

(0.5

4)(0

.32)

(0.3

5)Pe

rcei

ved

Har

m2.

44∗∗

∗1.

35∗∗

∗1.

61∗∗

∗2.

43∗∗

∗

from

Imm

igra

tion

(0.2

7)(0

.26)

(0.4

8)(0

.36)

Sobe

lTes

t(z

)0.

850.

850.

830.

85R

2(a

dju

sted

)0.

11%

Pre

dict

Cor

rect

ly46

.156

.757

.356

.389

.289

.669

.674

.2%

Impr

oved

7.7

18.3

0.5

–0.5

0.6

1.0

3.3

7.9

N28

128

028

027

927

927

827

828

028

0

Not

e:C

oeff

icie

nts

(an

dst

anda

rder

rors

)ar

ees

tim

ates

from

ordi

nar

yle

ast-

squ

ares

regr

essi

on(f

orPe

rcei

ved

Har

m),

max

imu

mlik

elih

ood

esti

mat

ion

ofor

dere

dpr

obit

mod

els

(for

Dec

reas

eIm

mig

rati

onan

dSu

ppor

tE

ngl

ish

On

ly),

and

max

imu

mlik

elih

ood

esti

mat

ion

ofpr

obit

mod

els

(for

Req

ues

tA

nti

-Im

mig

rati

onIn

form

atio

nan

dSe

nd

An

ti-I

mm

igra

tion

Mes

sage

toC

ongr

ess)

.All

mod

els

incl

ude

con

trol

sfo

red

uca

tion

,in

com

e,ag

e,an

dge

nde

r.∗ p

<.1

0,∗∗

p<

.05,

∗∗∗ p

<.0

1(t

wo-

taile

d).


that Latino immigrants are far more numerous than Eu-ropean immigrants, then this perception should eitherincrease along with anxiety or moderate the interactioneffect that gives rise to anxiety. To test these possibilities,we examine the impact of the experimental manipula-tions on the perceived gap between Latino and Europeanimmigration levels (i.e., Percent Latino − Percent Euro-pean). We find no evidence that the interaction whichgenerates anxiety also affects perceived levels of immi-gration (F = 0.48, p < .489). Instead, Latino cues widenthe perceived gap by nearly seven percentage points (F =3.88, p < .050), and the gap is largest, 42%, for the posi-tive Latino condition. Similarly, the interaction effect onanxiety is not moderated by the perceived gap in immi-gration levels. In sum, beliefs about the size of immigrantgroups are influenced by group cues, but this does notexplain changes in anxiety, opinion, or behavior.

Another explanation suggests the interaction effect isdriven by assumptions about the skill level of Latino ver-sus (white) European immigrants. If people assume thatLatino immigrants are less skilled than Europeans, thenthis may explain their anxiety. In this case, low-skilledimmigrants should trigger anxiety regardless of ethnicity.This contrasts with the notion that the effect is driven byrace or ethnicity alone. Of course, these factors are notmutually exclusive. Skill levels may be an important aspectof the negative stereotype. If so, we would expect an inter-action between ethnicity and skill level, such that only thedepiction of low-skilled Latino immigrants would boostanxiety. We performed a second experiment to explorethis possibility.

Experiment 2

We recruited 267 subjects from the local area, includingduring a festival that attracts hundreds of thousands ofvisitors from across the country. They were offered $20for taking a “public opinion survey” about current eventsand completed a self-administered questionnaire at one ofa dozen computer stations. For comparison with the firstexperiment, we limit our analysis to the 220 non-Latinowhite subjects. Fifty-eight percent of these subjects werewomen. The median subject was 45 years old with a bach-elor’s degree and a household income of roughly $52,000.Twenty-four percent of subjects identified as Republicansand 45% identified as Democrats.

Subjects first answered questions regarding predis-positions and political knowledge. They then read thefollowing message: “For this study, we have collected alarge number of articles that have appeared recently in

newspapers around the country. The computer is goingto select one of these newspaper articles at random for youto read. When you are finished with the article, we willask you some questions about it. Please read carefully.”After reading the article, they answered more questionsand then were fully debriefed before leaving.

The experiment employed a 2 × 2 design with acontrol group. Again a mock New York Times article wasused.16 All of the stories included the same basic infor-mation as the negative frame from the first study. As be-fore, we manipulated the salient ethnicity using a photo.The caption identified the immigrant as Jose Sanchezor Nicholas Van Dyke, the latter reflecting a slight alter-ation from the first experiment—a Dutch instead of Rus-sian surname.17 The photos featured the identical facesfrom the first experiment. The second factor involvedthe implication that immigrant workers are either high-or low-skilled. The first condition noted concerns abouthigh-skilled job competitors and quoted high-skilled na-tive workers who had witnessed colleagues losing jobsto immigrants. The second condition included similarcontent, except that the relevant natives and immigrantswere low-skilled. The subtitle also made a reference to“high-skilled” or “low-skilled” immigrants. All subjectswere told that immigration is increasing and that it willhave negative consequences. We focus, in other words,on the threatening half of the design from the firstexperiment.18

The effects of the second experiment were more sub-tle than the first and harder to detect due to the smallersample size. Manipulations of the news story did not sig-nificantly change policy preferences or requests for infor-mation, but they did affect the probability of sending ananti-immigration message to Congress. We focus on thisinstance of political action, since our goal is to see whichthreatening story subjects found most compelling andwhether the same psychological mechanism is at workas in the first study. Figure 4 shows the results for per-ceived harm, anxiety, and sending a message to Congress.In the third panel, we see an interaction between ethnic-ity and skill level emerges for sending a message (F =3.77, p < .054). When the immigrant story featured

16Subjects in the control group read an article about fast-foodchains.

17The switch in nationality offers an opportunity to see how sen-sitive effects are to the European nationality with which Latinosare contrasted. The Dutch were part of more “traditional” waves ofimmigrants to the United Sates. The cue is also more subtle in thisstudy, as neither the immigrant nor his nationality are mentionedin the body of the story. The photo, however, remains prominent.

18Cell sizes: 42 (control), 47 (low sk. Lat.), 44 (low sk. Eur.), 44(high sk. Lat.), 43 (high sk. Eur).


FIGURE 4 Impact of Skill Level Emphasis andEthnic Cue on Perceived Harm fromImmigration, Anxiety aboutImmigration, and Political Action

Note: The “baseline” control group mean values for each panel are2.26 (top), 1.31 (middle), and 0.19 (bottom).

a Latino, the proportion of subjects who requested ananti-immigration message be sent to their congressper-son increased from .11 in the high skill condition to .24in the low skill condition. When the European cue waspresent, opposition was actually higher (.21) after read-ing about high-skilled rather than low-skilled immigrants(.16), though this difference is not significant.

Is the interaction effect on behavior driven by achange in beliefs or emotions? Figure 4 tells a clear story.The middle panel shows a large interaction between eth-nic cue and skill level on anxiety (F = 7.94, p < .005).The low-skilled, Latino immigrant cue boosts anxiety by.77 points on a 0–6 scale, compared to the high-skilledLatino cue. There is a tiny drop in anxiety from the high-skilled to the low-skilled condition when the immigrantis European. No interaction occurs for perceived harm,displayed in the top panel of Figure 4 (F = 0.23, p < .633).There is a modest counterintuitive main effect of ethnic-ity, however: immigration was more likely seen as havinga negative impact when European, not Latino, cues werepresent (Mdiff = .38, F = 3.68, p < .057). Despite this, onlythe low-skilled Latino cue triggered anxiety.19 Consistentwith expectations, anxiety and not perceptions again me-diate the impact of the news on political action (see Table3: Mediated Model 1).

One possible objection to our findings about the me-diating role of anxiety is that the measures of perceivedharm ask about the likely impact on Americans in general.Perhaps anxiety captures the subject’s belief that immi-gration is likely to affect her personally. Fortunately, wecan address this concern with data from the second ex-periment. Subjects were also asked how likely it was thatimmigration would have a “negative financial impact onyou or your family” and/or have a “negative impact on theway of life in your community.” We combined responsesto form a measure of perceptions of personal harm (� =.72). Manipulations of the news article produced neithermain nor interactive effects on perceived personal harm,and this holds for both the combined scale and the in-dividual items (p > .25 for all factors). As a result, themediating role attributed to anxiety cannot be explainedby changes in beliefs about personal harm any more thanby changes in beliefs about national harm (see Table 3:Mediated Models 2 and 3).

The second experiment also supports our principalhypotheses: immigration opinion depends on the immi-grant group made salient, and the effects of these groupcues on political behavior are mediated by anxiety. How-ever, it also supports the corollary prediction that out-group cues generate greater anxiety when they are con-sistent with stereotypes. The power of Latino group cuesto trigger anxiety does not stem simply from ethnicity, orfrom the implied skill level of immigrants, but rather froma combination of these factors. News about immigration

19The fact that European cues increase perceived harm seems tiedto education and income. As education decreases, for example, thedifference drops until perceived harm is higher for Latino cues.However, the interaction effect for anxiety persists regardless ofeducation or income.


TA

BL

E3

An

xiet

yb

utN

otP

erce

ived

Har

mM

edia

tes

the

Inte

ract

ion

Eff

ecto

fEth

nic

Cu

esan

dSk

ill-

Lev

elE

mp

has

ison

An

ti-I

mm

igra

tion

Beh

avio

r(E

xper

imen

t2)

Eff

ects

onSe

nd

ing

anE

ffec

tson

Pot

enti

alM

edia

tors

An

ti-I

mm

igra

tion

Mes

sage

toC

ongr

ess

An

xiet

yP

erce

ived

Per

ceiv

edab

out

Nat

ion

alH

arm

Per

son

alH

arm

Un

med

iate

dM

edia

ted

Med

iate

dM

edia

ted

Imm

igra

tion

ofIm

mig

rati

onof

Imm

igra

tion

Mod

elM

odel

1M

odel

2M

odel

3

Skill

sE

mph

asis

(low

)–0

.06

0.01

0.04

–0.5

8–0

.54

–0.6

1–0

.67∗

(0.0

5)(0

.06)

(0.0

6)(0

.37)

(0.3

8)(0

.38)

(0.3

8)E

thn

icC

ues

(Lat

ino)

–0.0

5–0

.11

0.01

–0.5

6–0

.48

–0.4

7–0

.67∗

(0.0

5)(0

.06)

(0.0

6)(0

.36)

(0.3

7)(0

.38)

(0.3

8)C

ues

×Sk

ills

Em

phas

is0.

20∗∗

∗0.

04–0

.04

1.01

∗∗0.

711.

08∗∗

1.17

∗∗

(0.0

7)(0

.09)

(0.0

8)(0

.50)

(0.5

3)(0

.52)

(0.5

3)A

nxi

ety

abou

tIm

mig

rati

on1.

86∗∗

∗

(0.5

5)Pe

rcei

ved

Nat

ion

alH

arm

ofIm

mig

rati

on1.

28∗∗

∗

(0.4

8)Pe

rcei

ved

Pers

onal

Har

mof

Imm

igra

tion

1.69

∗∗∗

(0.5

1)So

belT

est

(z)

2.16

∗∗0.

43–0

.49

R2

(adj

ust

ed)

0.15

0.10

0.04

%P

redi

ctC

orre

ctly

81.3

81.3

81.3

78.8

%Im

prov

ed1.

71.

71.

7–0

.8N

162

160

160

160

160

160

160

Not

e:C

oeff

icie

nts

(an

dst

anda

rder

rors

)ar

ees

tim

ates

from

ordi

nar

yle

ast-

squ

ares

regr

essi

on(f

orA

nxi

ety

and

Perc

eive

dH

arm

),an

dm

axim

um

likel

ihoo

des

tim

atio

nof

prob

itm

odel

s(f

orSe

ndi

ng

anA

nti

-Im

mig

rati

onM

essa

ge).

All

mod

els

incl

ude

con

trol

sfo

red

uca

tion

,in

com

e,ag

e,ge

nde

r,an

did

eolo

gy.

∗ p<

.10,

∗∗p

<.0

5,∗∗

∗ p<

.01

(tw

o-ta

iled

).


increases anxiety and provokes greater opposition amongwhite Americans when it highlights low-skilled Latinomigrants and emphasizes the negative consequences ofimmigration. In the presence of reassuring reports in-consistent with the stereotype, beliefs may change, butanxiety, attitudes, and behavior do not.

Conclusion

We find substantial support for two key propositions: (1)group cues in immigration discourse can elicit anxietyand (2) changes in anxiety, not perceived threat, medi-ate the impact of these cues on opinion and politicalbehavior. Conventional wisdom suggests that oppositionto immigration grows with the severity of the harm im-migrants pose. Another popular but untested suspicionis that such opposition depends on the identity of salientimmigrant groups. Our results show that, while news em-phasizing the costs of immigration boosts the perceptionthat immigration is harmful, ethnic cues strongly condi-tion emotional reactions to this news. Stigmatized out-groups, in this case Latino immigrants, trigger negativeemotions when costs are emphasized. This emotional trig-ger matters. Even when citizens in our study perceived theharmful consequences of immigration identically, hostileattitudes and actions flared only when group cues elicitedanxiety.

Previous scholarship has shown how group attitudesare correlated with immigration opinion, but no study todate has shown so clearly that racial cues powerfully al-ter immigration opinion. More importantly, perhaps, weshow how group cues seem to have this effect. Two expla-nations are possible: (1) group cues change beliefs aboutthe severity of the immigration problem; or (2) group cuestrigger emotions, and these emotions (not beliefs aboutseverity) drive opposition. In our study, perceptions of thelikely harm from immigration were quite responsive tonews emphasis on costs or benefits, but perceptions alonewere not sufficient to move opinions or provoke politicalaction. This finding held regardless of whether the threatwas perceived as economic or cultural in nature, or aspersonal or collective in scope. In contrast, group cuesand the emotions they triggered proved critical to chang-ing attitudes and behavior. Citizens felt more threatenedby Latino immigration, not European immigration, andthis feeling triggered opposition to immigration and mul-tilingual laws, prompted requests for information, and ledpeople to send anti-immigration messages to Congress.Although group cues also changed beliefs on occasion(e.g., about the relative size of immigrant groups), those

changes did not account for effects on policy opinions orpolitical behavior.20

Our results also have broader implications for re-search on the roles of both groups and emotions in pol-itics. For example, the findings bring new questions intoview for the study of group politics. First, how does theemotional impact of group cues fit with their role in ac-tivating prejudice? We know subtle group cues can affectopinion by increasing the salience of group-related dis-positions (Mendelberg 2001; Nelson and Kinder 1996).However, scholars have found that anxiety-eliciting stim-uli tend to decrease the salience of predispositions injudgment (Brader 2005; Marcus, Neuman, and MacKuen2000), while group priming seems to occur via heightenedaccessibility of group schemas in memory (Valentino,Hutchings, and White 2002). Second, how does individ-ual variation in predispositions such as symbolic racism(Sears 1988) moderate emotional reactions to cues suchas the ones in our immigration stories (cf. Marcus et al.1995)? While at first blush one might guess that thosehigh in racial animus might react more negatively to thecue and therefore be even more powerfully affected bythe story, it is not necessarily so. If racial animus is aproduct of previous, recurring negative emotions towarda group, additional exposure may neither boost the emo-tion above that already elevated baseline, nor have anyadded impact on opinions or behavior. More research onthe emotional impact of groups promises to further ourunderstanding of the socialization and activation of racialpredispositions.

In particular, future work needs to clarify how groupcues trigger emotions. Psychologists have begun to studyemotional reactions to groups (Mackie and Smith 2003)but rarely examine the ways groups get invoked or im-plicated subtly in public discourse. In the present studies,anxiety from Latino cues embedded in the news seemedto evaporate when those same cues were accompaniedby unexpectedly reassuring information about the bene-fits of immigration or the higher skill status of migrants.Why? We speculate that anxiety arousal may hinge onstereotypic depiction of the group. When stereotypes areundermined, the emotional impact wanes. When detailsare simply absent, people “fill in” the missing pieces with

20At some level, “beliefs” about harm and “feeling threatened”must be related. But how? Offering a complete answer to thatquestion would require access to more intrusive types of data. Atthe nonconscious level, we know situations can trigger anxiety onlywhen the brain perceives potential harm; yet emotions often occurwithout conscious awareness and thus without triggering changesin beliefs that would be detected in a self-report (Damasio 2000).Emotions and beliefs likely exert reciprocal influence, especiallyby the time they manifest themselves consciously (Lazarus 1991;LeDoux 1996).


negative stereotypes (Gilliam and Iyengar 2000), and theemotional impact remains. An alternative explanation isthat Latino cues did unconditionally elicit anxiety, butwe did not detect it because the available informationwas immediately reassuring. In this more complicated ac-count, the lack of a corresponding positive shift in opin-ions might be explained by the asymmetric impact ofnegative and positive information (Holbrook et al. 2001).

What do our findings have to say about theories ofemotion and politics? Although our study was not de-signed to adjudicate among competing theories of emo-tion, our results are broadly consistent with tenets of atleast two leading approaches. In keeping with the expec-tations of Hot Cognition theory (Lodge and Taber 2005),we find that affective processes, rather than consciousbeliefs, guide political reasoning. All three types of out-comes we examined—policy opinions, information seek-ing, and political action—moved in a direction consistentwith the negative affect that was triggered (i.e., in an anti-immigration direction). In line with Affective Intelligencetheory (Marcus, Neuman, and MacKuen 2000), we findasymmetric effects for positive and negative emotionsand, specifically, that anxiety facilitates opinion change inthe direction of currently available information. Similarly,we find that anxiety is positively associated with threat-relevant information seeking. Both theories predict thatemotions provide an “essential motivational thrust” tocarry out one’s goals (i.e., act on one’s attitudes; Lodgeet al. 2006, 30; Marcus, Neuman, and MacKuen 2000),and we indeed find that anxiety-eliciting news not onlychanged opinions, but also mobilized opposition into po-litical action.21

We also contribute to research on emotion and poli-tics in significant ways. While recent work helps to es-tablish the causal role of anxiety (Brader 2006), ourstudy verifies that anxiety actually mediates the effects ofthreatening political stimuli. Moreover, we demonstratethe impact of anxiety on actual behavior, rather thanself-reports. Our findings underscore the importance ofemotion and yet depart from recent research emphasiz-

21Although there is considerable overlap in how these two theoriesconceptualize emotion, they do posit some distinct mechanismsand effects. First, while each emphasizes a single basic negativeand positive emotion, Affective Intelligence theory predicts a thirdemotion, anger, will sometimes emerge with consequences distinctfrom anxiety (Marcus et al. 2006). Second, while Hot Cognitiontheory expects both positive and negative affect to prime emotion-ally congruent considerations (Lodge and Taber 2005), AffectiveIntelligence theory predicts that enthusiasm and anxiety triggerdistinct modes of judgment in which either reliance on habit or ef-fortful processing of current considerations holds sway (Marcus etal. 2000). Third, whereas Hot Cognition theory predicts that nega-tive affect is a source of bias, Affective Intelligence theory predictsthat anxiety promotes less biased information seeking.

ing its beneficial effects. One implication of the presentstudy is that citizens’ responses to new information maybe held hostage to the past, in the form of anxiety inducedby outgroup images that may or may not hold contem-porary relevance. Emotions serve an adaptive functionin alerting us to important changes in our environment(Lazarus 1991), but this hard-wiring can leave us overlyresponsive to irrelevant symbols and too weakly respon-sive to what has changed. Our results also illustrate theway in which the outcome of the anxiety-stimulated judg-ment process is conditional on the information available.Structural biases in public discourse that lead some con-sequences, characteristics, or groups to be emphasizedover others may be picked up and amplified by anxiouscitizens. Candidates or interest groups can exploit vulner-abilities for political gain, but journalists also may be partyto influencing the public in this manner, driven solely bynews imperatives or their own stereotypes (Gilens 1999).

In order to shed light on the mechanisms underly-ing dynamics in public opinion, the present study followsearlier work by using experiments to “mimic elite dis-course” (Kinder and Sanders 1996). Our findings haveserious implications for public debates over policy issues.Political rhetoric about public policies is often laden withreferences to religious, national, and racial groups. Whencues about stigmatized outgroups are available, as they of-ten are, political elites can imbue information about theissue with emotional significance for the audience. To un-derstand the impact of elites on mass opinion, one mustattend to the power of feeling, not just seeing, threats.

AppendixQuestion Wording

EMOTIONS. “Now, moving on, we would like to knowhow you feel about increased immigration. The follow-ing questions will ask you how you feel when you thinkabout the high levels of immigration to this country.How [anxious (that is, uneasy)/proud/angry/hopeful/worried/excited] does it make you feel?” (Very, some-what, a little, or not at all?)

BELIEFS. Perception of National Harm: “In youropinion, how likely is it that immigration will have anegative financial impact on many Americans?” “Nowsetting aside any economic concerns, in your opinion,how likely is it that immigration will have a negativeimpact on the way of life in many American communi-ties?” Perception of Personal Harm: “How likely is it thatimmigration will have a negative financial impact onyou or your family?” “Again setting aside any economic


concerns, how likely is it that immigration will have anegative impact on the way of life in your community?”

OPINIONS. Immigration: “Do you think the numberof immigrants from foreign countries who are permittedto come to the United States to live should be increased alot, increased a little, left the same as it is now, decreased alittle, or decreased a lot?” English Only: “Do you favor a lawmaking English the official language of the United States,meaning government business would be conducted inEnglish only, or do you oppose such a law?”

BEHAVIOR. Information Seeking: “There are manyorganizations that put out information on immigration.Would you be interested in receiving more informa-tion about this issue from any of the following types ofgroups?” Contacting: “There are currently several pro-posed laws before Congress that would substantially de-crease the number of legal immigrants that come to thiscountry. Would you like us to send an email message tomembers of Congress indicating either your support oropposition to these reforms?”

References

Baron, Reuben M., and David A. Kenny. 1986. “The Moderator-Mediator Variable Distinction in Social Psychological Re-search.” Journal of Personality and Social Psychology 51(6):1173–82.

Brader, Ted. 2005. “Striking a Responsive Chord: How PoliticalAds Motivate and Persuade Voters by Appealing to Emo-tions.” American Journal of Political Science 49(2): 388–405.

Brader, Ted. 2006. Campaigning for Hearts and Minds. Chicago:University of Chicago Press.

Brader, Ted, and Nicholas A. Valentino. 2005. “Emotions, Iden-tities, and the Activation of Stereotypes in Response to Im-migration Threats.” Presented at the annual meeting of theInternational Society of Political Psychology, Toronto.

Citrin, Jack, Donald Green, Christopher Muste, and Cara Wong.1997. “Public Opinion toward Immigration Reform: TheRole of Economic Motivations.” Journal of Politics 59(3):858–81.

Cohen, Edward. 2001. Politics of Globalization in the UnitedStates. Washington, DC: Georgetown.

Damasio, Antonio R. 2000. “A Second Chance for Emotion.” InCognitive Neuroscience of Emotion, ed. Richard D. Lane andLynn Nadel. New York: Oxford University Press, 12–23.

Eagly, Alice, and Shelly Chaiken. 1993. The Psychology of Atti-tudes. Fort Worth, TX: Harcourt.

Fetzer, Joel S. 2000. Public Attitudes toward Immigration in theUnited States, France, and Germany. New York: CambridgeUniversity Press.

Gilens, Martin. 1999. Why Americans Hate Welfare. Chicago:University of Chicago Press.

Gilliam, Frank D., Jr., and Shanto Iyengar. 2000. “Prime Sus-pects: The Influence of Television News on the ViewingPublic.” American Journal of Political Science 44(3): 560–73.

Holbrook, Allyson L., Jon A. Krosnick, Penny S. Visser, WendiL. Gardner, and John T. Cacioppo. 2001. “Attitudes towardPresidential Candidates and Political Parties: Initial Opti-mism, Inertial First Impressions, and a Focus on Flaws.”American Journal of Political Science 45(4): 930–50.

Huddy, Leonie, Stanley Feldman, and Erin Cassese. 2007. “Onthe Distinct Political Effects of Anxiety and Anger.” In TheAffect Effect, ed. W. Russell Neuman et al. Chicago: Universityof Chicago Press, 202–30.

Huddy, Leonie, Stanley Feldman, Charles Taber, and Gallya La-hav. 2005. “Threat, Anxiety, and Support for AntiterrorismPolicies.” American Journal of Political Science 49(3): 593–608.

Huntington, Samuel P. 2004. “The Hispanic Challenge.” ForeignPolicy (March/April): 30–45.

Hutchings, Vincent L., Nicholas A. Valentino, Tasha Philpot,and Ismail White. 2006. “Racial Cues in Campaign News.”In Feeling Politics, ed. David Redlawsk. New York: PalgraveMacmillan, 165–86.

Isbell, Linda, and Victor Ottati. 2002. “The Emotional Voter.”In The Social Psychology of Politics, ed. Victor Ottati. NewYork: Kluwer, 55–74.

Kinder, Donald R. 2003. “Belief Systems after Converse.” InElectoral Democracy, ed. Michael MacKuen and George Ra-binowitz. Ann Arbor: University of Michigan Press, 13–47.

Kinder, Donald R., and Lynn Sanders. 1996. Divided by Color:Racial Politics and Democratic Ideals. Chicago: University ofChicago Press.

King, Desmond. 2000. Making Americans. Cambridge, MA:Harvard University Press.

Lapinski, John S., Pia Peltola, Greg Shaw, and Alan Yang. 1997.“The Polls—Trends: Immigrants and Immigration.” PublicOpinion Quarterly 61(2): 356–83.

Lazarus, Richard S. 1991. Emotion and Adaptation. New York:Oxford University Press.

LeDoux, Joseph. 1996. The Emotional Brain. New York: Simonand Schuster.

Lodge, Milton, and Charles Taber. 2005. “The Automaticityof Affect for Political Leaders, Groups, and Issues: An Ex-perimental Test of the Hot Cognition Hypothesis.” PoliticalPsychology 26(3): 455–82.

Lodge, Milton, Charles Taber, and Christopher Weber. 2006.“First Steps toward a Dual-Process Accessibility Model ofPolitical Beliefs, Attitudes, and Behavior.” In Feeling Politics,ed. David Redlawsk. New York: Palgrave Macmilllan, 11–30.

Mackie, Diane, and Eliot Smith, eds. 2003. From Prejudice toIntergroup Emotions. New York: Psychology Press.

MacKinnon, David, Chondra Lockwood, Jeanne Hoffman,Stephen West, and Virgil Sheets. 2002. “A Comparison ofMethods to Test Mediation and Other Intervening VariableEffects.” Psychological Methods 7(1): 83–104.

Marcus, George E., Michael MacKuen, Jennifer Wolak, and LukeKeele. 2006. “The Measure and Mismeasure of Emotion.”In Feeling Politics, ed. David Redlawsk. New York: PalgraveMacmillan, 31–45.


Marcus, George E., W. Russell Neuman, and Michael MacKuen.2000. Affective Intelligence and Political Judgment. Chicago:University of Chicago Press.

Marcus, George E., John L. Sullivan, Elizabeth Theiss-Morse,and Daniel Stevens. 2005. “The Emotional Foundationsof Political Cognition.” Political Psychology 26(6): 949–63.

Marcus, George E., John L. Sullivan, Elizabeth Theiss-Morse,and Sandra L. Wood. 1995. With Malice toward Some. Cam-bridge: Cambridge University Press.

Mendelberg, Tali. 2001. The Race Card. Princeton, NJ: PrincetonUniversity Press.

Miller, Joanne M., and Jon A. Krosnick. 2004. “Threat as a Mo-tivator of Political Activism: A Field Experiment.” PoliticalPsychology 25(4): 507–23.

Nelson, Thomas E., and Donald R. Kinder. 1996. “Issue Framesand Group-Centrism in American Public Opinion.” Journalof Politics 58(4): 1055–78.

Pande, Kamla. 2006. “The Effect of September 11, 2001 onMedia Discourse and Public Opinion toward Immigration.”Senior honors thesis. University of Michigan.

Quillian, Lincoln. 1995. “Prejudice as a Response to Per-ceived Group Threat: Population Composition and Anti-Immigrant and Racial Prejudice in Europe.” American Soci-ological Review 60(4): 586–611.

Rumbaut, Ruben, and Alejandro Portes. 2001. “Ethnogenesis:Coming of Age in Immigrant America.” In Ethnicities, ed.Ruben G. Rumbaut and Alejandro Portes. Berkeley: Univer-sity of California Press, 1–19.

Scheve, Kenneth F., and Matthew J. Slaughter. 2001. Globaliza-tion and the Perceptions of American Workers. Washington,DC: Institute for International Economics.

Sears, David O. 1988. “Symbolic Racism.” In EliminatingRacism, ed. Phyllis A. Katz and Dalmas A. Taylor. New York:Plenum, 53–84.

Sears, David O. 2001. “The Role of Affect in Symbolic Poli-tics.” Citizens and Politics, ed. James Kuklinski. New York:Cambridge University Press, 14–40.

Simon, Rita, and Susan Alexander. 1993. The Ambivalent Wel-come. Westport, CT: Praeger.

Simon, Rita, and James Lynch. 1999. “A Comparative Assess-ment of Public Opinion toward Immigrants and Immigra-tion Practices.” International Migration Review 33(2): 455–67.

Sniderman, Paul, Louk Hagendoorn, and Markus Prior. 2004.“Predispositional Factors and Situational Triggers: Exclu-sionary Reactions to Immigrant Minorities.” American Po-litical Science Review 98(1): 35–50.

Sniderman, Paul M., Pierangelo Peri, Rui J.P. de Figueiredo,Jr., and Thomas Piazza. 2000. The Outsider: Prejudice andPolitics in Italy. Princeton, NJ: Princeton University Press.

Tichenor, Daniel J. 2002. Dividing Lines. Princeton, NJ: Prince-ton University Press.

U.S. Census Bureau. 2001. Profile of the Foreign-Born Populationin the United States: 2000. Current Population Reports, P23–206.

Valentino, Nicholas A., Vincent L. Hutchings, Antoine J. Banks,and Anne K. Davis. 2008. “Is a Worried Citizen a Good Citi-zen? Emotions, Political Information Seeking, and Learningvia the Internet.” Political Psychology 29(2): 247–73.

Valentino, Nicholas A., Vincent L. Hutchings, Krysha Gre-gorowicz, Eric Groenendyk, and Ted Brader. 2006. “ElectionNight’s Alright for Fighting: The Role of Anger versusAnxiety in Political Participation.” Presented at the an-nual meeting of the American Political Science Association,Philadelphia.

Valentino, Nicholas A., Vincent L. Hutchings, and Ismail White.2002. “Cues That Matter: How Political Ads Prime RacialAttitudes during Campaigns.” American Political Science Re-view 96(1): 75–90.

what triggers public opposition to immigration? anxiety ...pscourses.ucsd.edu/ps100da/brader what...

Documents