ArticlePDF Available

Visibility and Votes: A Spatial Analysis of Anti-Immigrant Voting in Sweden

Authors:

Abstract and Figures

The mechanisms by which negative attitudes toward immigrants become votes for anti-immigrant parties are not fully understood. Yet, voting for political parties with anti-immigrant platforms is arguably the most common expression of these sentiments in Europe. I use anti-immigrant attitudes as a starting point and hypothesize that superficial intergroup contact, or immigrant ‘visibility’, brings these attitudes to the fore as politically salient. A spatial analysis of electoral data from each polling station in Sweden for the 2010 parliamentary election (n= 5,688) provides support for the hypothesis. Much of the variance in district-level voting can be accounted for by the percent of non-western residents in adjacent neighborhoods. The findings suggest that the probability of anti-immigrant attitudes translating into votes increases in neighborhoods where residents are likely to have fleeting contact with immigrants and I test this further with a city-level case study. I collected observational data on the visibility of non-westerners in a mid-size Swedish city and find that votes for the Sweden Democrats are above the national average where immigrants are most visible. Furthermore, the effect of non-western residents on anti-immigrant voting is most pronounced in regions without histories of significant non-western immigration, suggesting that the negative effects of superficial contact diminish over time.
Content may be subject to copyright.
Visibility and votes: A spatial
analysis of anti-immigrant
voting in Sweden
Sarah Valdez
Carlos IIIJuan March Institute, Universidad Carlos III, C/ Madrid, 135 Edificio 18, Oficina 18.02.E07, 29803 Getafe,
Madrid, Spain. Email: svaldez@march.es
Abstract
The mechanisms by which negative attitudes toward immigrants become votes for
anti-immigrant parties are not fully understood. Yet, voting for political parties with
anti-immigrant platforms is arguably the most common expression of these senti-
ments in Europe. I use anti-immigrant attitudes as a starting point and hypothesize
that superficial intergroup contact, or immigrant ‘visibility’, brings these attitudes to
the fore as politically salient. A spatial analysis of electoral data from each polling
station in Sweden for the 2010 parliamentary election (n= 5,688) provides support
for the hypothesis. Much of the variance in district-level voting can be accounted for
by the percent of non-western residents in adjacent neighborhoods. The findings
suggest that the probability of anti-immigrant attitudes translating into votes increases
in neighborhoods where residents are likely to have fleeting contact with immigrants
and I test this further with a city-level case study. I collected observational data on the
visibility of non-westerners in a mid-size Swedish city and find that votes for the
Sweden Democrats are above the national average where immigrants are most visible.
Furthermore, the effect of non-western residents on anti-immigrant voting is most
pronounced in regions without histories of significant non-western immigration, sug-
gesting that the negative effects of superficial contact diminish over time.
Keywords: immigration, voting, contact, attitudes
1. Introduction
At first glance, an article about voting might seem a bit out of place in a special issue
on attitudes toward immigration. But voting for political parties with explicitly anti-
immigrant platforms is arguably the most common, visible expression of anti-immigrant
attitudes in Europe today and its prevalence is increasing in many countries. Marine Le Pen,
leader of France’s National Front, received 17.9 percent of votes in the first round of the
2012 presidential election—a record number for the party. In the 2011 national election,
MIGRATION STUDIES VOLUME 2 NUMBER 2 2014 162–188 162
doi:10.1093/migration/mnu029
Advance Access publication on 23 March 2014
!The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please email:
journals.permissions@oup.com
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
the Swiss People’s Party surpassed all other parties in Switzerland, with over 25 percent of
the vote. Anti-immigrant parties are growing even in some countries where they have not
been popular before. The British National Party received nearly 50,000 votes in the 2010
Westminster election, despite its first-past-the-post rules, and in the same year the Sweden
Democrats won parliamentary representation for the first time. Group threat is commonly
used to explain both negative attitudes toward immigrants and anti-immigrant party suc-
cess. Europeans may feel that non-western immigrants threaten their economic well-being
or national culture. It is not surprising, then, that anti-immigrant parties have become
successful as migration from developing countries has increased and, more recently, as the
global recession has spread.
While group threat undoubtedly plays a role in the formation of anti-immigrant
attitudes, and attitudes are a necessary precursor to anti-immigrant voting, the relationship
between attitudes and votes is not straightforward. Negative attitudes are strikingly
high and consistent across Europe (Ceobanu and Escandell 2010), while party success is
variable across countries. For instance, the western European countries with the lowest
levels of opposition to immigration—the Nordic countries, Finland, Switzerland, and
Luxembourg—all have nationally successful anti-immigrant parties, while several coun-
tries with less favorable attitudes do not—Portugal, Germany, and Spain, for example
(Sides and Citrin 2007). And while Lubbers et al. (2002) find that national-level variation
in anti-immigrant voting is impacted by public opinion toward immigration, characteris-
tics of the parties themselves have an even greater effect. It is not surprising that attitudes
are not perfect predictors of behavior, as this is a common finding in the social sciences, but
it does highlight the need to further investigate the conditions under which negative atti-
tudes toward immigrants manifest as votes.
One possible mechanism by which anti-immigrant attitudes turn to votes is through the
increased salience of immigration as political issue. Even if political preferences remain
stable, their reordering can lead to a change in voting behavior if a new dimension not
represented by one’s current party becomes relevant. Immigration may increase in salience
via superficial, but routine, interpersonal contact between ethnic majority group members
and immigrant group members which results in immigrants becoming more ‘visible’ in
their communities. Dating back to Allport (1954), it has been hypothesized that fleeting
intergroup contact may heighten negative sentiments toward out-group members (though
meaningful intergroup contact diminishes them). But contact has also been shown to
influence political behavior. In Dutch cities, segregation increased anti-immigrant party
votes and the effect was greatest in more tolerant cities (van der Waal et al. 2013). Since
segregation decreases the overall likelihood of meaningful intergroup contact, this is con-
sistent with expectations of the contact hypothesis. Similar results were found in Britain.
The probability of anti-immigrant party membership increased in ethnically heterogeneous
cities only if those cities were also segregated. In Sweden, living next to, but not within, an
ethnically heterogeneous neighborhood increases anti-immigrant voting. Drawing upon
distinctions between meaningful and superficial contact, Rydgren and Ruth attribute these
findings ‘to the discrepancy between living close to the “imagined other” but not the
“experienced other”’ (2013: 13). Furthermore, in line with expectations regarding the
positive effects of meaningful contact, party membership in Britain was negatively corre-
lated with heterogeneity within a neighborhood, where sustained meaningful contact is
VISIBILITY AND VOTES 163
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
likely to develop (Biggs and Knauss 2012). Ceteris paribus, routine yet superficial contact
with members of an immigrant out-group should increase the probability of casting a vote
for an anti-immigrant party among members of the ethnic majority group with pre-exist-
ing negative attitudes.
Unlike other articles in this issue, I do not attempt to explain the emergence of anti-
immigrant attitudes, nor do I use them as an explanatory variable. I begin with the assump-
tion that they are pre-existing in the population and are a necessary but insufficient cause for
anti-immigrant voting, then examine one condition under which they may become polit-
ically salient—through superficial contact with ethnic ‘others’. This article builds upon
existing literature and tests the relationship between neighborhood ethnic heterogeneity
and anti-immigrant vote in the 2010 parliamentary election in Sweden. It adds to
Rydgren and Ruth’s (2013) findings by testing the effect of a broader set of neighborhoods
using spatial analysis rather than OLS. This method allows me to examine the effects of more
distant neighborhoods, rather than only including those that share a border, and it also
enables me to detect regional variation in the effects of ethnic heterogeneity on voting.
I find that most variation in anti-immigrant party voting can be accounted for by the
ethnic heterogeneity of the five to ten nearest neighborhoods and that the effect is weakest
in regions with the longest histories of immigration. This suggests that the negative impacts
of contact diminish over time, perhaps as meaningful intergroup relationships develop over
a broader set of the population. Additionally, I find that neighborhood-level heterogeneity
is a better predictor of anti-immigrant vote than municipal-level measures of heterogeneity
or group threat, which suggests that contact with one’s neighbors is more relevant than the
overall proportion of immigrants in a city when explaining anti-immigrant political be-
havior. I couple these quantitative analyses with observational fieldwork in a mid-sized
Swedish city, thereby examining the effect of residential segregation on intergroup contact
in shared public spaces.
2. Literature review
Prejudice toward immigrants is widespread across Europe (Zick et al. 2008) and is com-
monly attributed to perceived threats. The group threat hypothesis posits that a minority
group poses a threat to the majority’s social position as they grow in number (Williams
1964) or begin to compete for resources (Blalock 1957). Ethnic majorities may fear that
immigrants will compete for low-skill jobs, undermine local wage structures, or require
government-provided social assistance. Non-western immigrants in particular may have
foreign customs, religions, and languages that threaten dominant national cultures.
Therefore, negative sentiments toward immigrants will form when members of the
ethnic majority group feel economically or culturally threatened. In response, anti-immi-
grant parties arise and promote policies that have been described as xenophobic and
racist—often seeking to reduce immigration rates, cut social benefits to newcomers, pro-
mote assimilation, or increase aid to developing countries so as to reduce the push factors
of immigration.
Economic threat models work well at the individual level, though the evidence is mixed
for larger units of analysis. At the individual level, education is negatively correlated with
164 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
anti-immigrant attitudes across national contexts (see for example, Coenders and
Scheepers 2003). Negative sentiment toward immigrants is also correlated with low
labor force status (Scheepers et al. 2002; Gorodzeisky and Semyonov 2009) and unemploy-
ment (Semyonov et al. 2008). The unemployed (Lubbers et al. 2002), males (Givens 2005),
and the working and middle classes (Ivarsflaten 2005; Norris 2005) are overrepresented
among far right voters. Economic threats are less effective at explaining national trends and
cross-national variation. There is mixed evidence on whether unemployment rates or
economic growth rates affect country-level anti-immigrant attitudes (Quillian 1995;
Arzheimer and Carter 2006). And research does not consistently find the expected correl-
ations between unemployment rates and anti-immigrant votes at the municipal, regional,
or national level (Givens 2005; Lubbers and Scheepers 2001; Arzheimer and Carter 2006;
Knigge 1998).
There is much less evidence on the ways that cultural threats affect anti-immigrant
attitudes and voting behavior (perhaps because cultural variables are more difficult to
measure). However, there is evidence that they may be more important than economic
threats in predicting anti-immigrant attitudes (Sides and Citrin 2007). Schneider (2008)
finds that the percent of non-western immigrants in a population affects anti-immigrant
attitudes more than the percent of low-educated immigrants. This suggests that Europeans
are more threatened by newcomers who are culturally distant than by those who may
require welfare benefits or compete for low-skill jobs.
Much like economic threat, there is mixed evidence regarding the aggregate effects of
cultural threat. The number of immigrants in a country is correlated with perceived threat
at the national level (Lubbers and Scheepers 2001), but effects on voting are less clear. Some
studies find a correlation between overall levels of immigration and far right voting (Swank
and Betz 2003; Van der Brug et al. 2005), while others do not (Norris 2005). These
discrepancies may arise because a greater number of immigrants may cause majority
group members to feel higher levels of intergroup threat, but a greater number of immi-
grants also creates more opportunities for intergroup friendships to develop, which should
reduce feelings of cultural threat. Schneider (2008) finds a curvilinear relationship between
the percent of non-western immigrants and perceived threat among Europeans.
Group threat, either economic or cultural, is certainly integral to the formation of anti-
immigrant attitudes, but does not consistently explain voting behavior. Anti-immigrant
attitudes, therefore, are a necessary but not sufficient condition for anti-immigrant voting.
The same attitudes simply have different effects in different contexts and the conditions
under which anti-immigrant attitudes become politically salient are discussed in the fol-
lowing sections.
2.1 Contact
One reason that attitudes are not reliable predictors of voting behavior is because the
salience of beliefs are more important than their content; after all, an individual is acting
on only a small subset of his beliefs during any one election. Even if an individual has a
negative view of immigrants, he is unlikely to choose a party based on that attitude unless it
is more important to him than other political issues such as the economy, the environment,
or social welfare spending, for example. Intergroup contact can serve to heighten the
VISIBILITY AND VOTES 165
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
salience of anti-immigrant attitudes because, just as meaningful contact with immigrants
can reduce anti-immigrant attitudes, superficial contact exacerbates them and likely brings
them to the fore as politically relevant.
The contact hypothesis in its classic form predicts that if members of groups interact in
situations where they have equal status, intergroup prejudice will be reduced, and the effect
will be enhanced if the institutional environment fosters the perception of common inter-
ests (Allport 1954; Williams 1964; for reviews, see Brewer and Miller 1996; Durrheim and
Dixon 2005). In order for this type of contact to reduce intergroup conflict, these inter-
actions must transcend the individuals involved so that the new positive attitude will be
generalized toward all members of the out-group (Hewstone and Brown 1986). In other
words, a prejudiced individual must perceive the out-group member as an average repre-
sentative of her group, rather than as an exception, although the same positive effects can be
achieved if enough intimacy develops that individual traits become more salient than group
identities (Miller and Brewer 1984). Alternatively, meaningful contact is successful in
reducing intergroup conflict if it promotes a larger, common group identity which both
groups share (Gaertner and Dovidio 2000; Gaertner et al. 1993, 1996).
Recent meta-studies show that most research conducted over the past few decades con-
firm the expectations of the classic contact hypothesis (Brown and Hewstone 2005;
Pettigrew et al. 2011); however, recent advances have also been made. For instance,
Allport originally hypothesized that contact reduced prejudice by increasing knowledge
of the out-group, but more recent work suggests that the effects are more physiological than
cognitive. Contact appears to reduce anxiety and fear of interactions with the out-group
(Page-Gould et al. 2008). And once prejudice toward an out-group has been reduced
through contact, it may generalize to other out-groups with whom there was no contact
(Pettigrew 2009, Schmid et al. 2012). Meaningful contact combined with living in an
ethnically integrated community also has positive effects on trust and altruism (Uslaner
2012). Furthermore, extended contact—having in-group friends who have out-group
friends—can also reduce negative sentiments (Christ et al. 2010; Dovidio et al. 2011;
Pettigrew and Tropp 2008). Therefore, one need not have direct contact with an out-
group member so long as one is closely connected to someone who does. The effects of
contact, then, diffuse through direct links in social networks, though these effects are not as
strong as those produced through direct contact (Turner et al. 2007)
Despite the potential for friendships to reduce negative attitudes toward out-groups,
most daily interaction between members of different social groups is fleeting or superficial;
therefore, the majority of contact does not reduce prejudice. In fact, intergroup contact has
been shown to lead to resegregation and avoidance (Alexander and Tredoux 2010; Taylor
and Moghaddam 1994) or anxiety (Islam and Hewstone 1993; Stephan and Stephan 1985)
that decreases tolerance (Greenland and Brown 1999; Stephan and Stephan 2000). Most
intergroup contact is structured by some degree of informal segregation on a micro level
that shapes everyday interactions (Durrheim and Dixon 2005; Duncan and Duncan 1955;
Goldberg 1998; Massey and Denton 1988; Schnell and Yoav 2001). For this reason, one
would expect that typical, day-to-day interactions between group members are not suffi-
cient to alter attitudes. In fact, it is more likely that ordinary contact will serve to reinforce,
or even amplify, pre-existing prejudices because group members are likely to meet under
suboptimal conditions.
166 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
These patterns of daily interaction can have meaningful effects on people’s attitudes
toward out-group members. Rote activities such as buying groceries, walking down the
sidewalk, or sitting at a cafe can become the basis upon which out-group prejudices become
politically salient. Therefore, when assessing the impact of contact on anti-immigrant at-
titudes and voting behavior, it is important to distinguish meaningful from superficial
contact. Meaningful contact from the perspective of the contact hypothesis is the type
that allows individuals to look past group differences and overcome prejudice. There is
evidence in the European context that friendships between ethnic minority and majority
group members meet this criterion (Schneider 2008; McLaren 2003). Superficial contact,
on the other hand, is the type that sustains or exacerbates prejudice. So, increased visibility
of immigrants should serve to maintain anti-immigrant attitudes among the proportion of
the population holding them, and research on social segregation suggests that this fleeting,
routine contact between individuals comprises the bulk of intergroup contact in most
public settings.
2.2 Salience
I hypothesize that superficial contact not only sustains or aggravates anti-immigrant atti-
tudes; it also brings them to the fore as a politically salient issue. This can help explain the
disjoint between attitudes and voting behavior because one will only vote according to a
belief if it is a politically salient one. Unlike attitudes, which are not easily changed, salience
changes relatively frequently and unpredictably (Ivarsflaten 2005). Eatwell (1994) claims
that simply attributing the rise of anti-immigrant parties in Europe to attitudinal changes
among voters oversimplifies the complex relationship between belief and action, and
changes in ranking of political preferences is a plausible alternative. Even if attitudes and
beliefs are constant, voting behavior will change if saliency shifts. List and Dietrich (2011)
show that partisan change can be derived entirely from a shift in motivationally salient
dimensions, even if the individual has no change in belief and receives no new information.
This is consistent with explanations of the partisan shift that occurred in the United States
when the Civil Rights and Voting Acts made race a politically salient issue (Miller and
Schofield 2003).
Immigration can come to the fore as a politically salient issue through contact with
ethnic minorities. If an individual commonly encounters immigrants in daily life, immi-
gration may come to mind in the polling booth because he considers it relevant at the
moment. The accessibility of an issue in one’s memory serves as an important heuristic in a
field of complex political information (Huckfeldt et al. 1999) and there is evidence from
survey data that issue proximity trumps self-interest in explaining political preferences
(Sears and Lau 1983). Accessibility models have been applied in political science to explain
policy preferences and presidential choice in the United States (Aldrich et al. 1989; Jacobs
and Shapiro 1994; Krosnick 1988) and Canada (Johnston et al. 1992), as well as beliefs on
rights and liberties (Chong 1993). This ‘top of head’ effect determines salience because ‘the
more recently a consideration has been called to mind or thought about, the less time it
takes to retrieve that consideration or related considerations from memory and bring them
to the top of the head for use’ (Zaller 1992). These findings indicate that political prefer-
ences remain fairly stable while their order of importance is transient.
VISIBILITY AND VOTES 167
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
The visibility of immigrants should especially bring immigration to the fore as a political
issue when newcomers are members of an ethnic group distinct from the ethnic majority.
Wimmer (2008) posits that the visibility of ethnic markers makes ethnic boundaries pol-
itically salient, and salience is increased when boundaries are resistant to change or there is
inequality between groups. Furthermore, negative contact with a member of an ethnic out-
group has been shown to increase the salience of ethnicity more than positive contact does
(Paolini et al. 2010). So while fleeting, everyday contact with ethnic minority group mem-
bers is unlikely to change one’s attitude toward immigrants, it is likely to make one’s pre-
existing attitudes politically salient.
1
Of course, daily interactions in one’s neighborhood
are not the only way that immigration can become a salient political issue, but it is certainly
one worth investigating. For the proportion of the population holding negative attitudes
toward immigrants, superficial contact should increase the likelihood that they will vote for
a party with an anti-immigrant platform.
3. Variables
I use data from the 2010 parliamentary election in Sweden because this was the first election
in which the anti-immigrant Sweden Democrats gained national representation, which
helps me control for many explanatory variables. For instance, in a breakthrough election,
a party does not have a long-standing mobilizable base larger than the electoral threshold, a
roll-call record, or a history of compromise with other parties that can explain their success.
Furthermore, a single country study allows me to control for national-level variables that
may contribute to party success—immigration rates, unemployment rates, economic
growth, electoral rules, and political landscape, for example.
The Swedish case is appropriate for studying the effects of contact because Sweden is
marked by significant residential segregation; hence, there is much variation in expected
levels of intergroup contact. There are large differences between cities, with 12 municipa-
lities having foreign-born residents that comprise approximately 20 percent to 40 percent
of the total population, while 88 municipalities have fewer than 6.5 percent, and eighteen
have fewer than 4.1 percent (Statistics Sweden 2006). But there is also significant variation
to be found within cities, with neighborhoods ranging from zero to over 90 percent first-
and second-generation non-western immigrants (Statistics Sweden 2010). These segrega-
tion patterns are partially attributable to the state’s liberal refugee resettlement polices
(refugees are not ‘assigned’ to municipalities) and stock of large, isolated housing projects
commonly referred to as ‘immigrant ghettos’ among Swedes.
I measure variables at three levels in order to test for the effects of threat and contact on
anti-immigrant voting. The smallest units are neighborhoods (n= 5,688), which are nested
in municipalities (n=290). Municipalities, in turn, are nested within labor markets (n=68).
3.1 Neighborhood level
To test the proposed relationship between contact and voting, I analyze vote results from
the 2010 parliamentary election in Sweden. The dependent variable of interest is vote for an
anti-immigrant party. In Sweden, the only current national party whose primary platform
168 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
centers on restricting immigration is the Sweden Democrats (Sverigedemokraterna), a party
formed in 1988 that won parliamentary seats for the first time in 2010 when they garnered
5.7 percent of the vote. I measure votes cast for the Sweden Democrats as a proportion of
total eligible votes
2
(n= 4,665,460) at each polling station in the country.
The primary causal variable of interest, non-western residents, is the proportion of the
population not born in Nordic or EU-27 countries
3
(Statistics Sweden 2010). In Sweden,
this group is comprised mainly of refugees and asylum seekers who arrived in Sweden
beginning in the early 1980s, guest workers from the 1960s and 1970s, family members who
arrived under family reunification laws, and their children born in Sweden. I include the
estimate of immigrants’ children born in Sweden, defined as individuals born in Sweden to
two parents born outside of Sweden,
4
because according to the tenets of the contact hy-
pothesis, contact with any person perceived to be a member of an out-group should have an
effect, regardless of their country of birth.
I measure both of these variables—vote and the proportion of non-western residents—at
the electoral district level. Each electoral district (n= 5,688) contains one polling station,
such as a school or community center, and the median population is 1,659 with values
ranging from 122 to 4,512 inhabitants. This unit of analysis is better than city or regional
level measures to test the contact hypothesis since electoral districts approximate neigh-
borhoods, which are an important source of daily, routine intergroup contact for most
individuals. Official district names correspond to neighborhood names in many instances,
which suggests they coincide with geographic borders that residents consider meaningful.
3.2 Municipal and labor market level
I include several variables necessary to test for the effects of economic threat at the muni-
cipal level (n=290). Education is the percent of the population with three or more years of
post-secondary education.
5
Income is the median disposable income of the working age
population in the year preceding the election, measured in deciles. Using a measure of
disposable income rather than salary allows me to account for differences in tax rates, home
ownership, and capital gains income across municipalities. Unemployment is the percent of
the working age population who registered as unemployed at any time during the year prior
to the election.
6
This is a more inclusive measure than employment because employment
data were collected at only one point during the year, but unemployment includes anyone
who experienced a period of joblessness during the previous year. Non-western income and
non-western unemployment are income and unemployment measures of the refugee popu-
lation reported as ratios to the municipal average.
7
This is the only employment data
available that allows me to distinguish immigrants from the general population, but it
should be noted that it is a conservative estimate of ‘immigrant’ since it only includes
refugees—it excludes their children born in Sweden, non-western labor migrants who
arrived in Sweden mainly during the 1970s and 1980s, and immigrants who arrive for
reasons of family reunification.
To test the effects of economic threat on those who compete most directly with immi-
grants for jobs or wages, I include percent blue collar workers, defined as the percent
working age population employed in agriculture, forestry, fishing, manufacturing,
mining, energy, environmental activities, or construction, as identified by the Swedish
VISIBILITY AND VOTES 169
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
Standard Industrial Classification (SNI 2007) scheme in accordance with its EU counter-
part (EU NACE). I also included a measure of public sector employees because this group is
more protected than private sector employees in the labor market and, therefore, faces less
competition from immigrants. This is measured as the percent of the working age popu-
lation employed in government administration, state enterprise, primary municipal ad-
ministration, state-owned enterprises and organizations, municipally owned enterprises
and organizations, or other public institutions.
The best available proxy for cultural threat at the municipal level is the proportion
of non-western residents in the city because non-westerners are more culturally distant
than their Nordic or EU counterparts in Sweden. I also include several municipal-level
variables that should impact patterns of intergroup contact. I include population density
the number of residents per square kilometer in the municipality—which should affect the
probability of intergroup contact. It does not allow me, however, to distinguish be-
tween meaningful and superficial contact. I also include a measure of the publicly
owned
8
housing stock in each municipality built between 1966 and 1975 because this
was the era of the Million Programme in Sweden. This government program added over
one million new housing units in response to housing shortages during post World War II
urbanization. The housing projects were built on the outskirts of cities and contained
necessary neighborhood amenities such as shopping centers, restaurants, schools, libraries,
and social service offices. Furthermore, because they were often constructed according to
cutting-edge urban planning ideas of the time, a circular road with just one or two en-
trances often surrounds them. This increases the pedestrian-friendliness inside their
boundaries, but also serves to create a sense of geographic isolation from areas beyond
the perimeter. Today, many of these communities have a large proportion of non-western
residents and are referred to colloquially as ‘immigrant ghettos’.
9
This variable, MP-era
housing, offers a proxy of the degree to which the non-western resident population is
segregated, both residentially and socially during patterns of everyday movement, since
these housing units were designed to be self-contained communities on the outskirts of
cities.
In addition to municipal-level measures, I also include economic threat variables—
income, education, unemployment, blue collar, public sector, non-western income, and non-
western unemployment— measured at the level of labor market (n=68). The labor market is
a geographic unit larger than a municipality identified by Statistics Sweden as a closed unit
in terms of labor supply and demand, based on characteristics such as commuting patterns.
This unit of analysis should provide a more accurate measure of economic competition
than municipalities.
4. Methods and results
I test for economic threat and intergroup contact at the municipal and labor market levels,
then focus on the neighborhood level to test for the effects of contact on a smaller scale.
Because my variables are nested, in my first analysis I use a mixed model to estimate the
effects of electoral district, municipal, and labor market variables on anti-immigrant party
vote. I use the xtmixed command in Stata because it allows me to estimate a fixed effects
170 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
model (Snijders and Bosker 1999). This is preferable to a random effects model because my
data are a census rather than a sample.
10
Next, I focus on neighborhood-level contact by testing for spatial autocorrelation and
spatial heterogeneity in the data. Spatial autocorrelation is a test of whether a variable is
correlated with itself across space, which would violate the assumption of independence
necessary for OLS regression. In this analysis, it indicates the extent to which neighbor-
hoods of high (or low) anti-immigrant vote are clustered. Because I find significant spatial
autocorrelation, I run geographically weighted regression models that identify spatial het-
erogeneity in the data. Spatial heterogeneity exists when the effect of a variable changes by
location: in essence, an independent variable works differently in different places. In this
case, geographically weighted regressions are particularly useful as an exploratory tool
because they identify where the presence of immigrants has a greater or lesser effect on
voting.
Lastly, I draw upon observational fieldwork conducted in Sweden during 2008 and 2009
to provide insight into the ways that residential segregation impacts intergroup contact in
public spaces. I measure the visibility of immigrants in a mid-sized Swedish city with a
Million Programme housing project. I compare the visibility of ethnic minorities in various
parts of the city with voting results at corresponding polling stations.
4.1 Multi-level models
I begin with estimates of the effect of municipal characteristics (level 2) on electoral district
votes (level 1). Model 1 is empty and indicates that with no explanatory variables included,
approximately 62 percent of the variance in anti-immigrant party vote can be explained at
the municipal level (Table 1). Model 2 assesses the effects of classic economic threat vari-
ables. Non-western income has the largest effect on vote and is positively correlated with
support for the Sweden Democrats. This may indicate that Swedes feel they compete with
immigrants for jobs, but do not feel the threat that immigrants will undermine local wage
structures. Furthermore, the finding that non-western unemployment is negatively corre-
lated with party support indicates that Swedes are not politically motivated by the eco-
nomic threat of providing social benefits to immigrants. Education and public sector
employment decrease anti-immigrant vote consistent with threat models, and unemploy-
ment is positively correlated with vote as expected. However, inconsistent with economic
threat models, neither income nor blue collar employment is statistically significant.
11
In Model 3 I add measures that impact intergroup contact—population density and non-
western residents at both the municipal and neighborhood levels. Non-western residents at
the district level have a small but significant effect on anti-immigrant party vote, but there is
no effect at the municipal level. This indicates that voters are sensitive to ethnic hetero-
geneity at the neighborhood level and that, in terms of the presence of non-western resi-
dents, neighborhood characteristics are more relevant than municipal ones in determining
voting behavior. Population density of the municipality has no effect on vote, which indi-
cates that if contact impacts voting behavior, then intergroup contact is determined by
factors other than simply the compactness of the city; perhaps segregation patterns or
features of the urban landscape such as transportation networks matter.
VISIBILITY AND VOTES 171
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
Therefore, in Model 4 I add a measure of the proportion of public housing built in the
municipality during the Million Programme era. The effect of this variable is significant and
negative, which is consistent with my hypothesis that segregation reduces visibility and,
hence, anti-immigrant voting. The addition of this variable also makes non-western
Table 1. Multilevel models
a
of anti-immigrant party vote in Sweden
(1) (2) (3) (4) (5)
Municipal-level variables
Income 0.05 0.07 0.07 0.07
(.13) (.13) (.13) (.13)
Education 0.12* 0.13* 0.16** 0.16**
(.05) (.05) (.05) (.05)
Unemployment 0.22
y
0.18 0.15 0.15
(.12) (.12) (.12) (.12)
Blue collar 0.01 0.02 0 0
(.04) (.04) (.04) (.04)
Public sector 0.18*** 0.17*** 0.18*** 0.12***
(.04) (.04) (.04) (.04)
Non-western income 2.91
y
2.34 1.83 1.83
(1.53) (1.57) (1.56) (1.55)
Non-western unemployment 0.44* 0.40
y
0.15
y
0.35
y
(.21) (.21) (.12) (.21)
Non-western residents 0.06 0.12** 0.12**
(.04) (.04) (.04)
Population density 0.00 0.00 0.00
(0) (0) (0)
MP-era housing .10**
(.04)
Non-western MP .10**
(.04)
District-level variables
Non-western residents 0.01*** 0.01*** 0.01***
(0) (0) (0)
Intercept 6.40 7.61 7.56 9.30 9.30
(.15) (2.44) (2.45) (2.49) (2.49)
Proportion of variance
Municipality 6.13 5.09 5.08 4.92 4.92
(.54) (.48) (.48) (.47) (.47)
Electoral district 3.73 3.75 3.73 3.73 3.73
(.07) (.07) (.07) (.07) (.07)
y
p<.10; *p <.05; **p <.01; ***p <.001.
a
Group variable is municipal code.
172 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
residents at the municipal level significant so I include an interaction between non-western
residents and Million Programme housing in Model 5. I find that, as expected, the inter-
action effect is significant and negatively correlated with votes for the Sweden Democrats.
Together, Models 3 and 4 signify that it is not merely the proportion of immigrants in a city
which impacts voting but, rather, how they are arranged residentially.
Because I find no significance of income or blue collar across the models, which is in-
consistent with much of the literature, I suspect that municipalities were the incorrect unit
of analysis for testing economic threat and I use labor market at level two. In this model,
income becomes significant and positively correlated with anti-immigrant vote, but blue
collar employment remains insignificant (Table 2). As in municipal-level models, the
income level of non-western immigrants is the strongest predictor of vote, but changes
direction. This indicates that perhaps voters in Sweden do indeed feel wage competition
with immigrants. The significance and direction of all other variables remain the same and
are consistent with threat models. However, this model has reduced explanatory power
regarding voting behavior at the neighborhood level because only 32 percent of the variance
in the model is explained at the labor market level.
Across the models, several variables work in the direction consistent with group threat: a
city’s average level of education and protected public sector employment are inversely
correlated with anti-immigrant voting. But some key variables are not significant or
work in the opposite direction predicted by group threat. Perhaps most surprising is the
finding that income and employment are not significant. It is important to point out that this
is not evidence against group threat. A more reasonable interpretation is that the variables
in the model are predictive of anti-immigrant attitude, as much research has shown, but
not vote.
4.2 Spatial analysis
Even though multi-level models account for grouping of level one observations, they do not
account for proximity within or between the groups under study, so the models presented
in Table 1 estimate the effects of non-western immigrants in one district, but do not
account for immigrants in the surrounding neighborhoods. Since individuals are unlikely
to be confined to the neighborhood in which they reside, the proportion of immigrants in
neighboring areas should impact voting behavior if intergroup contact matters. We would
expect different levels of intergroup contact, and therefore anti-immigrant votes, when
comparing two districts with low levels of non-western residents if one borders a neigh-
borhood with similarly low numbers of non-westerners but the other abuts a neighborhood
with a high proportion of non-westerners. To test the spatial relation between non-western
residents and vote, I rely on tests of spatial autocorrelation and spatial heterogeneity using
neighborhood measures.
4.2.1 Spatial autocorrelation. To test for autocorrelation, I calculated Moran’s I—a
global correlation coefficient similar to Pearson’s r produced in standard regression
models, but constrained by space. Its values range from 1, meaning that values are
VISIBILITY AND VOTES 173
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
perfectly dispersed in space, to +1, indicating that values are perfectly correlated across the
geographic area under analysis.
I calculated Moran’s I using both inverse distance and polygon contiguity (first order)
commands in ArcGIS.
12
Inverse distance weights voting results in every district against
every other in the dataset, assuming that the impact of districts upon each other decreases
with distance.
13
Polygon contiguity calculates neighbor effects, so rather than taking into
account every district it only includes those with a shared border. Moran’s I is large and
statistically significant using both weight matrices. In general, values greater than 0.2 in-
dicate spatial autocorrelation, and both methods of calculation result in robust and sig-
nificant coefficients (Table 3).
Interestingly, Moran’s I is much larger when calculated using polygon contiguity. This
indicates that neighbor effects are strong and votes in a district are highly correlated with
those in adjacent districts. This is consistent with three clustering patterns which can be
derived from the contact hypothesis: 1) anti-immigrant voting will occur around neigh-
borhoods with large non-western populations, where ethnic majority residents are likely to
have regular, superficial contact with ethnic minority group members during daily activ-
ities; 2) anti-immigrant votes are less frequent in neighborhoods near areas with small
immigrant populations, where the ethnic majority has little routine contact with the ethnic
minority; and 3) anti-immigrant voting is less frequent in neighborhoods that have sur-
passed a certain threshold of non-western residents, where ethnic majority group members
are likely to have meaningful contact in the form of friendships with ethnic minority group
members.
Of course, it should be noted that this correlation can also be due to patterns inconsistent
with the contact hypothesis—for instance, anti-immigrant vote increasing in the absence of
non-westerners. The analysis cannot give us insight to the direction of correlation; it only
demonstrates a strong spatial correlation in voting. Therefore, I cannot determine whether
Table 2. Economic threat variables at the labor market level
(1)
Income 0.24 (.08)**
Education 0.19 (.01)***
Unemployment 0.13 (.03)***
Blue collar 0.00 (.01)
Public sector 0.04 (.01)***
Nonwestern income 1.78 (.55)**
Nonwestern unemployment 0.27 (.08)***
Intercept 7.12 (.89)
Proportion of variance
Labor market 3.75 (.70)
Electoral district 4.39 (.08)
y
p<.10; *p <.05; **p <.01; ***p <.001.
174 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
neighbor effects are due to the presence of immigrants or some other unobserved charac-
teristic of neighborhoods, so I introduce measures of non-western residents in my next
analysis. I use geographically weighted regression (GWR) because the robust and signifi-
cant spatial autocorrelation in voting violates the basic assumptions of independence in
ordinary least squares regression models.
4.2.2 Spatial heterogeneity. I estimate several GWR models with set bandwidth param-
eters ranging from five to 1,000. The first two models incorporate the nearest five and ten
electoral districts because I have a theoretical reason to suspect that patterns of daily
interaction are shaped mainly by the ethnic composition of the nearest neighborhoods. I
find that more than 90 percent of variance in anti-immigrant voting behavior can be
explained by the proportion of non-western immigrants the nearest five or ten electoral
districts (Table 4), which is consistent with my expectation that patterns of inter-group
contact impact the salience of immigration as a political issue. It should be noted that small
bandwidths reduce degrees of freedom so the adjusted R-squared values are lower, around
.80. I found no autocorrelation in the residuals, so the model accounts for all significant
spatial processes in the data.
As the number of neighbors increase, R-squared values decrease, but reveal observable
patterns of interest at the regional level. Model 5, which includes 1,000 districts as neigh-
bors, has the lowest R-squared value (though at 0.43 it is still notable in the social sciences)
and a map of regression coefficients reveals interesting regional patterns of spatial hetero-
geneity (Fig. 1). The neutral, gray shaded areas on the map indicate areas where the effect of
non-western residents in the nearest 1,000 electoral districts has almost no effect on voting.
The red areas indicate where the effect is positive, with GWR coefficient values ranging
from .02 to .10. Blue indicates a negative correlation with values ranging from .10 to zero.
A quick glance at the red areas on the map shows that residents of central Sweden are
more electorally sensitive to the presence of non-westerners than are Swedes in most other
parts of the country. There are also smaller red patches to the east of Stockholm and on the
island of Gotland, which is an interesting counter to economic threat models because the
island of Gotland and the area between the capital and the coast are generally high income
areas, but threat-based explanations posit that low income groups should feel most threa-
tened by immigration. However, because they are high income areas, they are likely similar
to central Sweden in that they are places relatively new to non-western immigration. The
larger regression coefficient in these regions could indicate that individuals in areas new to
immigration are especially sensitive to the presence of non-western newcomers.
Table 3. Spatial autocorrelation between districts
Moran’s I
Inverse distance 0.49***
Polygon contiguity (first order) 0.72***
y
p<.10; *p <.05; **p <.01; ***p <.001.
VISIBILITY AND VOTES 175
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
The analysis also reveals that voters in southern Sweden and in Stockholm are less
sensitive to the presence of non-westerners than voters in most other parts of the country.
This finding is interesting because southern Sweden is the area of the country where the
Sweden Democrats receive most votes. So, even though Ska
˚ne is the region where the
Sweden Democrats are most popular, the political expression of anti-immigrant attitudes
is relatively low, given the number of immigrants in the region. Given the long history of
immigration to this region (as well as Stockholm) and its high proportion of non-western
immigrants, this finding is consistent with the expectations of meaningful contact. When
an immigrant population surpasses a certain threshold, there is more possibility for mean-
ingful inter-group contact to occur. These findings indicate that this is likely happening in
many neighborhoods in the region.
4.3 Observational fieldwork
To examine the ways that residential segregation translates into superficial, inter-group
contact in public spaces, I conducted observational fieldwork in a mid-sized Swedish city. I
chose the city of Linko
¨ping for several reasons. Firstly, it is the seventh largest municipality
in Sweden, so it is large enough to be representative of the dynamics that likely occur in
other Swedish cities, but it is also compact enough that I could visit most areas of the city
during fieldwork. Importantly, Linko
¨ping also has a Million Programme housing project
and, according to the GWR results presented in the previous section, it is in an area of the
country where the impact of non-western residents on anti-immigrant voting is about
average. Additionally, due to its relatively high levels of employment, high-skill labor
force, and education, the city should be less susceptible to economic threat than many
cities.
Linko
¨ping is city approximately 200 kilometers southwest of the capital of Stockholm in
the county of O
¨stergo
¨tland. Historically, it served as the religious and educational center of
the region, later becoming the headquarters of the Saab aircraft industry. Today it is home
to one of Sweden’s largest universities and the country’s second largest information tech-
nology park. The population of the city is, on average, more highly educated and employed
than the national average. Of Linko
¨ping’s 144,690 residents (Statistics Sweden 2010), 13.1
percent are foreign-born (Table 5). Using the same calculations described above in the
Table 4. Geographically weighted regressions of anti-immigrant party vote
(1) (2) (3) (4) (5)
Neighbors 29 5 10 100 1,000
R
2
0.79 0.98 0.91 0.66 0.43
Adjusted R
2
0.73 0.80 0.78 0.64 0.43
AICc 22240 52679 25265 23146 25587
Residuals
Moran’s I 0.01*** 0.02*** 0.02*** 0.04*** 0.17***
y
p<.10; *p <.05; **p <.01; ***p <.001.
176 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
variables section, I find that approximately 12.9 percent of the residents of Linko
¨ping have a
non-western background. This makes the proportion of foreigners in Linko
¨ping slightly
higher than the Swedish average and, in line with national trends, they fare less well eco-
nomically than native Swedes. More than two-thirds of households receiving social assist-
ance in Linko
¨ping have a foreign-born head of household (Statistics Sweden, 2008).
Figure 1. Map of the regional effects of non-western residents on anti-immigrant party vote.
(Figure is available in colour online only).
VISIBILITY AND VOTES 177
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
Furthermore, Linko
¨ping has a segregation index of 39.3, nearly twice the Swedish average of
20.4 (Statistics Sweden 2009), though not unusual for a large or mid-sized Swedish city.
Like many old European cities, Linko
¨ping has a city center marked by cobblestone streets
lined with cafes and shops. The center also contains grocery stores, chain clothing stores, a
shopping mall, and a large, historic cathedral. Just beyond the center lie residential areas,
shops, and a university. Beyond that is a major roadway, named Industrial Road
(Industrigatan), that separates residential areas from the industrial area, a large modern
shopping mall, big-box stores surrounded by parking lots, and Ska
¨ggetorp—a Million
Programme project built for working class families in the 1960s and 1970s. This area is
referred to by city residents as the ‘immigrant neighborhood’ (personal interviews) and
contains schools, a shopping mall, restaurants, a library, and other necessary amenities.
Ska
¨ggetorp’s boundary is encircled by a small road, as is common among Million
Programme projects, and is surrounded on three sides by industrial or rural areas, so it is
geographically isolated from the rest of Linko
¨ping. For instance, to enter the neighborhood
by foot or bike, one must pass through short, tunneled underpasses.
I visited public spaces such as plazas, cafes, and shopping centers and categorized indi-
viduals, using phenotypic cues, as either ‘western’ or ‘non-western’ in an attempt to meas-
ure the visibility of non-westerners. This method is similar to that employed by Durrheim
and Dixon (2005) to measure small-scale, public segregation in desegregated South Africa.
This is an imperfect measure, so I erred on the side of undercounting ethnic minorities. For
instance, I almost certainly excluded immigrants from Russia or from the former
Yugoslavia (which is a sizable immigrant group in Sweden). However, attitudes toward
such groups may be less negative than are attitudes toward those groups that are more
ethnically distant from native Swedes (Waisman 2008). I counted a total of 4,924 individ-
uals and coded whether they were in mixed-ethnicity dyads, triads, or larger groups in
order to capture meaningful versus superficial contact. I took counts in various locations at
different times of day, week, and year during 2008 and 2009 in four areas of Linko
¨ping—1)
the historic town center; 2) the ‘immigrant neighborhood’ on the edge of town; 3) the
modern mall adjacent to the immigrant neighborhood; 4) the residential area midway
between the center and the immigrant neighborhood, which contains a commuter bike
path connecting the two (Fig. 2).
Table 5. Local and national immigrant populations
Linko
¨ping Sweden
Total population 144,690 9,340,689
Immigrants
Foreign-born (%) 13.1 14.3
Non-western foreign-born (%) 10.3 8.7
Immigrants and their children born in Sweden
Foreign background (%) 16.5 18.6
Non-western background (%) 12.9 11.3
178 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
I find that immigrants are underrepresented relative to their share of the population in
the center and midway zones of the city, an overrepresented in the immigrant neighbor-
hood. In the center they are visible at a rate roughly equivalent to three-quarters their actual
makeup of the city’s population. Their visibility is even lower in the residential zone outside
Figure 2. Road map and satellite image of Linko
¨ping, with observation areas highlighted in red.
(Figure is available in colour online only).
VISIBILITY AND VOTES 179
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
the center which is separated from Ska
¨ggetorp by Industrial Road, at just over one-quarter
their actual presence in the city. The opposite is true in the modern shopping center
adjacent to Ska
¨ggetorp, where non-westerners were visible at rates approaching twice
their proportion of the population. The ‘immigrant neighborhood’ appeared to have
roughly a 70 percent non-western population in public spaces, over five times their pro-
portion of the population in Linko
¨ping.
Furthermore, nowhere in the city did I observe much ‘meaningful’ contact. I saw groups
comprised of both ethnic minority and majority group members together at the highest
rates in the shopping center adjacent to the immigrant neighborhood, where 4.2 percent of
observed people were sitting, chatting, in mixed groups (Table 6). But I do not wish to place
too much emphasis on these findings because 61 percent of observed individuals were
alone. These observations, therefore, are most likely not reflective of the interactions that
occur in schools, or workplaces, or voluntary associations. This could be especially prob-
lematic in the adjacent zone, for example, where I noticed zero individuals engaged in
meaningful contact. However, as a residential area, most of my observations consisted of
bikers on the bike path, or people walking dogs or tending their lawns.
These findings indicate that residential segregation in Linko
¨ping is transformed into
social segregation in public places. This is most likely due to the concentration of immi-
grants in the Million Programme community. These projects were designed to be self-
contained, thereby eliminating the need for residents to travel outside of their neighbor-
hood for daily errands and, often, for work. Unsurprisingly, it appears that place of resi-
dence affects movement throughout the entire urban landscape. Therefore, if an individual
lives, works, and shops primarily inside the ‘ring road’ of the city, he is more likely to
perceive that non-westerners make up a smaller percentage of the city’s population than
they actually do. On the other hand, if one lives, works, or shops in or near the immigrant
neighborhood, he is more likely to perceive non-western immigrants as constituting a
much larger proportion of the city’s population than they actually do. I hypothesize that
individuals in this situation who already hold negative attitudes toward immigrants have an
increased probability of voting for the Sweden Democrats because the salience of immi-
gration increases under these conditions of superficial contact.
Indeed, voting patterns in the city of Linko
¨ping are consistent with this claim. In Fig. 3,
the gray lines overlaid on the map of Linko
¨ping denote electoral districts, and the colors
denote votes for the Sweden Democrats in standard deviations relative to the national
average. Blue shades indicate ‘cool spots’ with lower than average support and orange
indicates ‘hot spots’ of party voting. The only areas where voters choose the anti-immigrant
party at rates of at least 1.5 standard deviations above the national average are in and
around the immigrant neighborhood, while party support tends to be below average
inside the ring road where non-westerners are less visible. The yellow areas indicate
votes at about national average rates. It is interesting to note that the large yellow patch
on the eastern edge of the map is one large electoral district because it includes a large swath
of farmland. It may show more variation if it were divided into smaller districts because
some residents likely live very near Ska
¨ggetorp while others are quite distant. I cannot
explain the additional yellow areas to the south with the observational data I collected in the
field, but they contain two additional Million Programme projects with much smaller im-
migrant populations than Ska
¨ggetorp. Personal interviews I conducted with residents
180 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
indicate that these areas are notable for being ‘working class’ more than ‘immigrant’
neighborhoods.
5. Conclusion
Taken together, these analyses indicate that neighborhood levels of inter-group contact can
help explain anti-immigrant voting. The multi-level models show that levels of economic
threat in a city can account for some of the variation in district voting, though not as clearly
as previous literature has shown that it explains anti-immigrant attitudes. The spatial
analyses show that voting for an anti-immigrant party was highly spatially dependent in
the 2010 Swedish election and was robustly correlated with the percent of non-western
Figure 3. Satellite imagery of Linko
¨ping overlaid with votes for Sweden Democrats. (Figure is
available in colour online only).
Table 6. Visibility of non-westerners in various parts of Linko
¨ping
Non-westerners observed As a proportion of
actual non-western population
Center 9.4% 0.73
Midway 3.8% 0.29
Edge 69.5% 5.37
Adjacent 23.5% 1.82
VISIBILITY AND VOTES 181
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
residents in the five nearest electoral districts. This is consistent with my hypothesis that
superficial inter-group contact, or ‘visibility’, can increase the political salience of anti-
immigrant attitudes among the ethnic majority, but these data cannot test the mechanism
of contact. My observational data from Linko
¨ping does a better job of this. I find that votes
for the Sweden Democrats are higher than average in areas where immigrants are visible in
public spaces at rates higher than their actual proportion of the city’s population.
Furthermore, the findings presented here highlight the need for more research that
examines the micro-level effects of segregation and contact. There have been decades of
rich research on both the positive and negative effects of inter-group contact, but we know
much less about the environmental conditions that structure contact. Case studies may be
especially useful for identifying the dimensions of urban landscapes that impact inter-
group contact. This study suggests that urban planning projects can have significant, but
unintended, social consequences in terms of segregation. The Million Programme was an
ambitious project implemented by Social Democrats to create modern self-contained
communities for workers moving to cities during post World War II urbanization. And
though they were successful in creating communities for residents several decades ago, they
now serve to segregate many immigrant communities. For instance, their location on the
outskirts of cities, often bordered by a circular road, creates a feeling of geographical
isolation and contributes to turning simple residential segregation into comprehensive
social segregation.
This study also suggests that we need to conceive of more nuanced ways to measure
contact with large-N data sets. Thresholds are sometimes employed as a proxy for mean-
ingful contact because as the number of immigrants increases, so too does opportunity for
inter-group friendships to develop. This appears to be at least partially supported by my
findings that the presence of non-western immigrants in southern Sweden and Stockholm
impacts voting less than in other areas of the country. This may be due to the relatively high
concentrations of immigrants in these areas: meaningful contact is facilitated once immi-
grants make up a certain proportion of the population because there is simply more op-
portunity for friendship to develop. However, these findings could also be a function of
time. Friendships will not immediately develop simply because newcomers move into a
community, even if their numbers increase quickly beyond a meaningful threshold. In fact a
quick increase may have the opposite effect. People especially notice change when con-
fronted with large quantities of information (Kahneman and Tversky 1979) and rapid
demographic change can impact local politics (Hopkins 2010). Therefore, the arrival of
immigrants in regions, small towns, or labor sectors where they have not traditionally been
present may lead to a temporary increase in anti-immigrant attitudes that decreases as
more immigrants arrive and enough time passes that meaningful friendships become the
norm rather than the exception.
This study attempts to explicitly link attitudes with voting behavior by testing the pro-
posal that superficial contact is one mechanism through which anti-immigrant attitudes
become politically salient. We have a plethora of useful research measuring the cross-na-
tional variation in attitudes toward immigration, as well as many studies examining the
causes of these attitudes. This article builds upon those by using attitudes as a starting point
and asking when they turn into political behavior. Understanding this process is important
given the context of immigration in modern Europe. Over five million immigrants now call
182 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
Western Europe home and rates of immigration from the developing world show no signs
of stopping. Understanding the ways in which interactions between ethnic groups can
either undermine or facilitate prejudice is timely, especially since it has the potential to
aggregate into political outcomes with broad social consequences.
Funding
Lars Jonsson Family Fund Grant from the Department of Scandinavian Studies, at the
University of Washington and Mellon Foundation Pre-Dissertation Award from the
Council for European Studies (CES) at Columbia University.
Acknowledgments
I am grateful for the support of the Juan March Institute at Carlos III University. Special
thanks are also due to Steve Pfaff, Edgar Kiser, Margaret Levi, Christine Ingebritsen, and
Maureen Eger, as well as the participants of the seminar series at the Institute for Research
on Migration, Ethnicity, and Society (REMESO) at Linko
¨ping University.
Conflict of interest statement. None declared.
Notes
1. It should be noted that ethnic heterogeneity may increase overall political engagement
(Kesler and Bloemraad 2010; Putnam 2007), but for the proportion of the population
that votes and holds negative attitudes toward immigrants, intergroup contact should
increase the salience of immigration.
2. I removed 173,889 early votes (Fo
¨rtidsro
¨ster ej fo
¨rdelade efter valdistrikt) because they
were only reported at aggregate levels by municipality. Of these, 8,143 were votes for the
Sweden Democrats. This leaves me with 4,665,460 votes cast in the parliamentary
election (Statistics Sweden 2010).
3. It should be noted that this measure of ‘non-western immigrants’ includes some per-
sons from western countries such as North America and Oceania because they are not
separated in the data. However, they are a small enough proportion of the population
that they should not significantly impact my findings. In 2010, residents born in
Australia, Canada, Switzerland, New Zealand, or the United States comprised 2.0 per-
cent of the total foreign-born population of Sweden.
4. At the electoral district level, counts of children born to immigrants are not reported
according to parents’ country of origin so I estimate them by multiplying offspring (the
number of the district’s population born in Sweden to two parents not born in Sweden)
by the proportion of non-western residents (foreign-born outside of Nordic or EU-27
countries). I ran models both including and excluding these estimates and found no
difference in results. Therefore, I report results including these estimates because it is
the measure of immigration closest to my theoretical conception of the variable.
VISIBILITY AND VOTES 183
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
5. This is Statistics Sweden’s closest approximation to having earned a BA degree, though
it will include years of education in educational tracks designed for skilled labor that
would not normally lead to a university degree in the Scandinavian educational system
(i.e. technical or arts training).
6. 2009 data were unavailable for three municipalities and I substituted data collected in
2007 or 2008.
7. As noted by a reviewer, the emigration of immigrants to Sweden is underreported.
Therefore, these data may include individuals no longer residing in Sweden, which
would lead to an underreporting of income at the municipal level.
8. Some are owned by the municipality or the state, but most are owned by privately-
owned public housing companies. The figures were collected in 1990 and there have
been a few municipality splits since then. In those cases, I used the 1990 value for both
new municipalities (Nykvarn and So
¨derta
¨lje; Knivsta and Uppsala; Bora
˚s and
Bollebygd; A
¨ngelholm and Ba
˚stad).
9. Some examples are Rynkeby and Tensta in Stockholm, Rosenga
˚rd in Malmo
¨, Angered
in Gothenburg, and Ska
¨ggetorp in Linko
¨ping.
10. I ran random effects models using the xtreg command with no difference in findings.
11. To test the robustness of this finding, I removed education from the model, as income
and education are likely highly correlated, but this did not change the significance of
income.
12. In each case I row standardized so all districts will carry equal weight in the analysis.
13. For instance, when calculating the effect of neighboring districts’ votes upon voting in
electoral district A, if the center of district C is twice as far away as the center of B, the
votes in B are weighted twice as heavily as those of C.
References
Aldrich, J. H., Sullivan, J. L. and Borgida, E. (1989) ‘Foreign Affairs and Issue Voting: Do
Presidential Candidates Waltz Before a Blind Audience?’ American Political Science
Review, 83: 123–41.
Alexander, L. and Tredoux, C. (2010) ‘The Spaces between Us: A Spatial Analysis of
Informal Segregation at a South African University’, Journal of Social Issues, 66: 367–86.
Allport, G. (1954) The Nature of Prejudice. Garden City, NY: Doubleday.
Arzheimer, K. and Carter, E. (2006) ‘Political Opportunity Structures and Right-Wing
Extremist Party Success’, European Journal of Political Research, 45/3: 391–418.
Biggs, M. and Knauss, S. (2012) ‘Explaining Membership in the British National Party: A
Multilevel Analysis of Contact and Threat’, European Sociological Review, 28/5: 633–46.
Blalock, H. M. Jr (1957) ‘Percent Non-White and Discrimination in the South’, American
Sociological Review, 22: 677–82.
Brewer, M. and Miller, N. (1996) Intergroup Relations. Buckingham: Open University
Press.
Brown, R. and Hewstone, M. (2005) ‘An Integrative Theory of Intergroup Contact’,
Advances in Experimental Social Psychology, 37: 255–343.
184 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
Ceobanu, A. M. and Escandell, X. (2010) ‘Comparative Analyses of Public Attitudes
Toward Immigrants and Immigration Using Multinational Survey Data: A Review of
Theories and Research’, Annual Review of Sociology, 36: 309–28.
Chong, D. (1993) ‘How People Think, Reason, and Feel about Rights and Liberties’,
American Journal of Political Science, 37: 867–99.
Christ, O., Hewstone, M., Tausch, N. et al. (2010) ‘Direct Contact as a Moderator of
Extended Contact Effects: Cross-Sectional and Longitudinal Impact on Outgroup
Attitudes, Behavioral Intentions, and Attitude Certainty’, Personality and Social
Psychology Bulletin, 36/12: 1662–74.
Coenders, M. and Scheepers, P. (2003) ‘The Effect of Education on Nationalism
and Ethnic Exclusionism: An International Comparison’, Political Psychology, 24/2:
313–43.
Dovidio, J. F., Eller, A. and Hewstone, M. (2011) ‘Improving Intergroup Relations through
Direct, Extended and Other Forms of Indirect Contact’, Group Processes and Intergroup
Relations, 14/2: 147–60.
Duncan, O. D. and Duncan, B. (1955) ‘A Methodological Analysis of Segregation Indices’,
American Sociological Review, 20: 210–17.
Durrheim, K. and Dixon, J. (2005) Racial Encounter: The Social Psychology of Contact and
Desegregation. New York: Routledge.
Eatwell, R. (1994) ‘Why are Fascism and Racism Reviving in Western Europe?’ Politics
Quarterly, 65/3: 313–25.
Gaertner, S. L. and Dovidio, J. (2000) Reducing Intergroup Bias: The Common Identity
Model. Hillsdale, NJ: Psychology Press.
—— & Dovidio, J. F. and Bachman, B. (1996) ‘Revisiting the Contact Hypothesis: The
Induction of a Common Ingroup Identity’, International Journal of Intercultural
Relations, 20: 271–90.
——, Anastasio, P. et al. (1993) ‘The Common Ingroup Identity Model: Recategorization
and the Reduction of Intergroup Bias’, in Stroebe, W. and Hewstone, M. (eds),
European Review of Social Psychology, pp. 1–26. London: Wiley.
Givens, T. (2005) Voting Radical Right in Western Europe. New York: Cambridge
University Press.
Goldberg, D. T. (1998) ‘The New Segregation’, Race and Society, 1: 15–32.
Gorodzeisky, A. and Semyonov, M. (2009) ‘Terms of Exclusion: Public Views Towards
Admission and Allocation of Rights to Immigrants in European Countries’, Ethnic and
Racial Studies, 32/1: 401–23.
Greenland, K. and Brown, R. (1999) ‘Categorization and Intergroup Anxiety’, European
Journal of Social Psychology, 29: 503–21.
Hewstone, M. and Brown, R. (1986) ‘Contact is Not Enough: An Intergroup Perspective
on the “Contact Hypothesis”’, in Hewstone. and Brown. (eds), Contact and Conflict in
Intergroup Encounters. Oxford: Basil Blackwell.
Hopkins, D. J. (2010) ‘Politicized Places: Explaining Where and When Immigrants
Provoke Local Opposition’, American Political Science Review, 104/1: 40–60.
Huckfeldt, R., Levine, J., Morgan, W. and Sprague, J. (1999) ‘Accessibility and the Political
Utility of Partisan and Ideological Orientations’, American Journal of Political Science,
43: 888–911.
VISIBILITY AND VOTES 185
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
Islam, M. R. and Hewstone, M. (1993) ‘Dimensions of Contact as Predictors of Intergroup
Anxiety, Perceived Out-Group Variability and Out-Group Attitude: An Integrative
Model’, Personality and Social Psychology Bulletin, 19: 700–10.
Ivarsflaten, E. (2005) ‘The Vulnerable Populist Right Parties: No Economic
Realignment Fueling their Electoral Success’, European Journal of Political Research,
44/3: 465–92.
Jacobs, L. R. and Shapiro, R. Y. (1994) ‘Issues, Candidate Image, and Priming: The Use of
Private Polls in Kennedy’s 1960 Presidential Campaign’, American Political Science
Review, 88: 527–40.
Johnston, R., Blais, A., Brady, H. E. and Crete, J. (1992) Letting the People Decide: Dynamics
of a Canadian Election. Stanford, CA: Stanford University Press.
Kahneman, D. and Tversky, A. (1979) ‘Prospect Theory: An Analysis of Decision under
Risk’, Econometrica, 47/2: 263–92.
Kesler, C. and Bloemraad, I. (2010) ‘Does Immigration Erode Social Capital? The
Conditional Effects of Immigration-Generated Diversity on Trust, Membership,
and Participation across 19 Countries, 1981–2000’, Canadian Journal of Political
Science, 43/2: 319–47.
Knigge, P. (1998) ‘The Ecological Correlates of Right-Wing Extremism in Western
Europe’, European Journal of Political Research, 34: 249–79.
Krosnick, J. A. (1988) ‘The Role of Attitude Importance in Social Evaluation: A Study of
Policy Preferences, Presidential Candidate Evaluations, and Voting Behavior’, Journal of
Personal Social Psychology, 55: 196–210.
List, C. and Dietrich, F. (2011) ‘A Model of Non-Informational Preference Change’,
Journal of Theoretical Politics, 23/2: 145–64.
Lubbers, M., Gijsberts, M. and Scheepers, P. (2002) ‘Extreme Right-Wing Voting in
Western Europe’, European Journal of Political Research, 41: 345–78.
——. (2001) ‘Explaining the Trend in Extreme Right-Wing Voting: Germany 1989–1998’,
European Sociological Review, 17: 431–49.
McLaren, L. M. (2003) ‘Anti-Immigrant Prejudice in Europe: Contact, Threat Perception,
and Preferences for the Exclusion of Migrants’, Social Forces, 81/3: 909–36.
Massey, D. S. and Denton, N. A. (1988) ‘The Dimensions of Residential Segregation’,
Social Forces, 67/2: 281–315.
Miller, G. and Schofield, N. (2003) ‘Activists and Partisan Realignment in the United
States’, American Political Science Review, 97/2: 245–60.
Miller, N. and Brewer, M. B. (1984) ‘The Social Psychology of Desegregation: An
Introduction’, in Miller and Brewer (eds), Groups in Contact: The Psychology of
Desegregation, pp. 1–7. New York: Academic Press.
Norris, P. (2005) Radical Right: Voters and Parties in the Electoral Market. New York:
Cambridge University Press.
Page-Gould, E., Mendoza-Denton, R. and Tropp, L. R. (2008) ‘With a Little Help from my
Cross-Group Friend: Reducing Anxiety in Intergroup Contexts through Cross-Group
Friendship’, Journal of Personality and Social Psychology, 95/5: 1080–94.
Paolini, S., Harwood, J. and Rubin, M. (2010) ‘Negative Intergroup Contact Makes Group
Memberships Salient: Explaining Why Intergroup Conflict Endures’, Personality and
Social Psychology Bulletin, 36/10: 1723–38.
186 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
Pettigrew, T. F. (2009) ‘Contact’s Secondary Transfer Effect: Do Intergroup Contact Effects
Spread to Non-Participating Outgroups?’ Social Psychology, 40/2: 55–65.
—— and Tropp, L. R. (2008) How Does Intergroup Contact Reduce Prejudice? Meta-
Analytic Tests of Three Mediators, European Journal of Social Psychology, 38: 922–34.
—— & ——, Wagner, U. and Christ, O. (2011) ‘Recent Advances in Intergroup Contact
Theory’, International Journal of Intercultural Relations, 35/3: 271–80.
Putnam, R. D. (2007) ‘E Pluribus Unum: Diversity and Community in the Twenty-First
Century. The 2006 Johan Skytte Prize Lecture’, Scandinavian Political Studies, 30/2:
137–74.
Quillian, L. (1995) ‘Prejudice as a Response to Perceived Threat: Population
Composition and Anti-Immigrant and Racial Prejudice in Europe’, American
Sociological Review, 60/4: 586–611.
Rydgren, J. and Ruth, P. (2013) ‘Contextual Explanations of Radical Right-Wing Support
in Sweden: Socioeconomic Marginalization, Group Threat, and the Halo Effect’, Ethnic
and Racial Studies, 36/4: 711–28.
Scheepers, P., Gijsberts, M. and Coenders, M. (2002) ‘Ethnic Exclusionism in European
Countries: Public Opposition to Civil Rights for Legal Migrants as Response to
Perceived Ethnic Threat’, European Sociological Review, 18/1: 17–34.
Schmid, K., Hewstone, M., Ku
¨pper, B. et al. (2012) ‘Secondary Transfer Effects of
Intergroup Contact: A Cross-National Comparison in Europe’, Social Psychology
Quarterly, 75/1: 28–51.
Schneider, S. L. (2008) ‘Anti-Immigrant Attitudes in Europe: Outgroup Size and Perceived
Ethnic Threat’, European Sociological Review, 24/1: 53–67.
Schnell, I.and Yoav, B. (2001) ‘The Sociospatial Isolation of Agents inEveryday Life Spaces as
an Aspect of Segegation’, Annals of the Association of American Geographers, 91: 622–36.
Sears, D. O. and Lau, R. R. (1983) ‘Indicating Apparently Self-Interested Political
Preferences’, American Journal of Political Science, 27: 223–52.
Semyonov, M., Raijman, R. and Gorodzeisky, A. (2008) ‘Foreigners’ Impact on European
Societies. Public Views and Perceptions in a Cross-National Comparative Perspective’,
International Journal of Comparative Sociology, 49/1: 5–29.
Sides, J. and Citrin, J. (2007) ‘European Opinion About Immigration: The Role of
Identities, Interest, and Information’, British Journal of Political Science, 37/3: 477–504.
Snijders, T. A. B. and Bosker, R. J. (1999) Multilevel Analysis: An Introduction to Basic and
Advanced Multilevel Modeling. London: Sage Publications.
Stephan, W. G. and Stephan, C. W. (1985) ‘Intergroup Anxiety’, Journal of Social Issues, 41:
157–77.
——. (2000) ‘An Integrated Threat Theory of Prejudice’, in Oskamp, S. (ed.), Reducing
Prejudice and Discrimination, pp. 23–45. Hillsdale, NJ: Lawrence Erlbaum Associates.
Swank, D. and Betz, H. G. (2003) ‘Globalization, the Welfare State and Right-Wing
Populism in Western Europe’, Socio-Economic Review, 1/2: 215–45.
Taylor, D. M. and Moghaddam, F. M. (1994) Theories of Intergroup Relations: International
Social Psychology Perspectives. London: Praeger.
Turner, R. N., Hewstone, M., Voci, A. et al. (2007) ‘Reducing Prejudice via Direct and
Extended Cross-Group Friendship’, in Strobe, W. and Hewstone, M. (eds), European
Review of Social Psychology, pp. 212–55. Hove: Psychology Press.
VISIBILITY AND VOTES 187
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
Uslaner, E. (2012) Segregation and Mistrust: Diversity, Isolation, and Social Cohesion. New
York: Cambridge University Press.
Van der Brug, W., Fennema, M. and Tillie, J. (2005) ‘Why Some Anti-Immigrant Parties
Fail and Others Succeed’, Comparative Political Studies, 38/5: 537–73.
Van der Waal, J., de Koster, W. and Achterberg, P. (2013) ‘Ethnic Segregation and Radical
Right-Wing Voting in Dutch Cities’, Urban Affairs Review, 49: 748–77.
Waisman, G. (2008) Essays on Discrimination and Corruption. Stockholm: Stockholm
University.
Williams, R. M. (1964) Strangers Next Door. Englewood Cliffs, NJ: Prentice Hall.
Wimmer, A. (2008) ‘The Making and Unmaking of Ethnic Boundaries: A Multilevel
Process Theory’, American Journal of Sociology, 114/4: 970–1022.
Zaller, J. (1992) The Nature and Origin of Mass Opinion. New York: Cambridge University
Press.
Zick, A., Pettigrew, T. F. and Wagner, U. (2008) ‘Ethnic Prejudice and Discrimination in
Europe’, Journal of Social Issues, 64/2: 233–51.
188 S. VALDEZ
at Linkopings universitetsbibliotek on December 4, 2014http://migration.oxfordjournals.org/Downloaded from
... Bowyer, 2008;Rydgren, Ruth, 2013;Evans, Ivaldi, 2021); other studies indicate a positive relationship in line with H. Blumer's (1958) conflict theory (e.g. Lubbers, Scheepers, 2002;Coffé et al., 2007;Valdez, 2014), and still other studies indicate the lack of any relationship (e.g. Lubbers, Scheepers, 2000;de Blok, van der Meer, 2018;Van Wijk et al., 2019). ...
... In the 2019 parliamentary elections, this party obtained 6.81%, and in the 2023 parliamentary elections 7.16%. A significant part of the works dealing with the relationship between the ethnic composition of the environment and voting for RRPs are ecological analyses (e.g., Coffé et al., 2007, Bowyer, 2008Rydgren, Ruth, 2013;Valdez, 2014;Martig, Bernauer, 2018). The great advantage of these works is that they are based on actual election results and not on voting declarations, as is the case with survey research (Savelkoul et al., 2017). ...
Article
Full-text available
The main aim of this study is to analyze the impact of the presence of the Slovak national minority in Poland on voting for radical right parties (RRPs) by the native Polish population. Temporally, this work covers the years 1991–2023. Territorially, the work covers the Podhale region, where the largest number of Slovaks in Poland live. A comparative analysis of cartographic materials was the basic method used in this study. The results presented in this article indicate a positive relationship between the presence of the Slovak national minority and voting by the native Polish population for RRPs: stronger voting by the native Polish population for RRPs was observed when the share of the Slovak minority in a particular locality exceeded 20%.
... A third meta-analytic review, with 50 studies, supported moderate intervention effects on outgroup attitudes through anti-bias programs in school, but found that these were most effective when delivered one-onone and by researchers (Ülger et al., 2018). A commonality among these studies is that intervention programs typically Journal of Youth and Adolescence revealed stronger effects in outgroup attitude change for adolescents from ethnic majority groups, but also that outgroup attitudes likely improve more in schools with ethnically balanced populations (van Zalk and Kerr, 2014), indicating that such interventions may have worked differently for different groups. One of the criticisms of schoolbased intervention work targeting outgroup attitude change is that it tends to be overtly practical: psychological theories and findings are rarely incorporated, and the impact on outgroup relations or attitudes is rarely evaluated systematically (Cameron & Turner, 2010). ...
... Sweden often scores very high on the Multicultural Policy Index (2021), indicating general support for policies toward immigrant minorities, and is often recognized for its strong commitment to social welfare and equality (Berggren & Trägårdh, 2015). At the same time, research also shows that ethnic segregation is increasing in the country (Hedström, 2019), and similar to many European countries, anti-immigrant attitudes have been (Valdez, 2014), and again is on the rise (European Social Survey, 2023). Adolescents with migration backgrounds commonly describe that they experience discrimination and racism (City of Gothenburg, 2023), as well as feelings of not fitting into the dichotomous, narrow frame of being considered a "Swede" (Gyberg et al., 2018). ...
Article
Full-text available
Outgroup and diversity attitudes are important components of intercultural understanding and well-being. Despite the potential of ethnic-racial identity development as a means to foster positive outgroup and diversity attitudes, little is known about its effectiveness in rapidly diversifying contexts such as Sweden. This pre-registered study filled this gap by examining if adolescents taking part in an intervention focused on ethnic-racial identity exploration, the Identity Project, also reported change in outgroup and diversity attitudes, and whether migration background, education type, and ethnic-racial identity development predicted such change. Twenty-three tenth-grade classes in Sweden (N = 509; Mage = 16.28; SDage = 0.80; 66% female; 51% migration background) participated in the intervention and were assessed in four waves over a period of 26 weeks. Whereas ethnic-racial identity exploration and resolution increased for the intervention group, the adolescents reported no change in outgroup and diversity attitudes when compared to a control group. Increases in ethnic-racial identity exploration and resolution co-varied with increases in attitudes, but only at Time 3. The results do not provide support for the link between ethnic-racial identity development and positive outgroup and diversity attitudes, and challenge the notion of attitude change as a cascading effect of the Identity Project intervention in non-US sociocultural contexts. All aspects of the study were pre-registered on the Open Science Framework platform (https://osf.io/f5896).
... Their location between differently populated communities, according to the theoretical argument, makes these sandwiched neighborhoods particularly susceptible to conflicts over community territory (i.e., the contested question to which of the two competing ethnic enclaves the sandwiched borderland belongs). Second, the haloeffect hypothesis proposes that mainstream majority members are more likely to vote for right-wing populist parties when they live in homogenous neighborhoods that border on ethnically diverse ones or are even encircled by them (Bowyer 2008;Martig and Bernauer 2018;Valdez 2014)-the idea of the encircled neighborhood gives the halo effect its name. Here, the argument is that such halo constellations entail limited opportunities for personal inter-group contact experiences in people's direct neighborhoods, whereas the presence of immigrants on the fringes of their neighborhoods instills feelings of group threat and hostility. ...
Article
Full-text available
Does immigration erode the social integration of contemporary mass-scale societies? Continued immigration and corresponding growing ethnic, racial, and religious diversity have prompted this important and highly controversial question. This article gives a brief introduction to the scientific literature on this question and derives from it the following gap that it seeks to address: According to a small number of studies on related but different topics, immigration may be detrimental to social integration if it physically manifests as one of two specific types of ethnic residential segregation. The contested boundaries hypothesis has it that border regions sandwiched between ethnically defined neighborhoods are particularly conflict prone and characterized by increased rates of crime. The halo-effect hypothesis claims that majority members who live in homogenous mainstream neighborhoods that border on ethnically diverse ones (or are even encircled by them) are more likely to vote for right-wing populist parties. In this article, we expand both approaches to the study of social integration in theoretical and empirical terms. With respect to theory building, we discuss why social integration, as indicated by social trust and community attachment, should suffer from these two types of ethnic segregation. To test these claims empirically, we use data from the geo-coded German General Social Survey (ALLBUS/GGSS) 2016 and 2018 that we merge with 100-m × 100‑m spatial grid data from the German Census 2011. These data allow us to apply edge detection techniques to identify ethnic residential boundaries, and our recently developed donut-method to measure ethnic residential halos. To our knowledge our study is the first investigation into the arguably important question whether ethnic residential boundaries and halos erode social integration.
... Responsiveness to demographic changes is the hallmark of a group threat response (Hopkins 2010) with the assumption that members of the majority group react with hostility when minority groups challenge their dominant status. Many studies (Rink et al. 2009;Savelkoul et al. 2017;Biggs and Knauss 2013;Valdez 2014) report that the radical right is more successful in areas with larger shares of immigrants; others report the opposite association or add additional conditions for when context becomes important. Contextual effects are one way to study ethnic competition, but occupation, firms or industries are other important contexts. ...
Article
Full-text available
This article uses Swedish register data to study the labour market experiences of radical right-wing candidates standing in local elections. We look at different measures of economic insecurity (labour market participation trajectories, experience of unemployment in social networks and relative growth in the number of jobs for foreign-born workers vis-a-vis natives) and examine whether they are predictors of candidates running for the Sweden Democrats, the main radical right-wing party in Sweden, as opposed to running for mainstream political parties. We find that the labour market trajectories of such candidates are markedly different from those of mainstream party candidates. Those with turbulent or out-of-labour market trajectories are much more likely to run for the Sweden Democrats, as opposed to other parties. The same is also true for candidates embedded in social networks with higher levels of unemployment, while working in a high-skilled industry markedly lowers the probability of running for the Sweden Democrats, especially for male candidates with low educational attainment. We find mixed results for the ethnic threat hypothesis.
... The rise of anti-immigrant and xenophobic parties across much of the Global North attests to this political tension. Among its many consequences, the introduction of progressively less generous policies and ever more complex legislative provisions for those who seek protection or family reunification can be observed (e.g., Federico & Pannia, 2021;Holtug, 2021), as can the gradual normalization of discriminatory anti-immigrant sentiments (Valdez, 2014;Gorodzeisky & Semyonov, 2016;Kuntz et al., 2017). ...
Chapter
Full-text available
This chapter serves as an introduction to the volume Migration and Integration in a Post-Pandemic World: Socioeconomic Opportunities and Challenges and is a broad and selected overview of the socioeconomic field of international migration and integration as we knew it before the Covid-19 pandemic. It sets the stage for exploring how the critical event of the virus impacted and may continue to impact our understanding of diverse macro-, meso-, and micro-level challenges and opportunities in migration and integration. The chapter motivates the purpose of the volume, as well as the structure of the 15 chapters and their individual contributions ranging from migration over time, transnationalism, migration policies and implementation, the role of trade unions and civil society actors, country-of-origin sector sorting and required skills, along with immigrant discrimination and vaccine hesitancy among migrant groups.KeywordsInternational migrationIntegration: Covid-19 pandemicCritical events
Article
How is support for right-wing populist parties affected by exposure to Muslim visibility? Using an original database on French mosques, this article analyzes the relationship between the presence of mosques and support for the Front National at the polling station level in the late 2000s. It finds that the propensity to vote for the Front National increases in polling stations up to intermediate distances from mosques and then decreases, suggesting a spatial mechanism known as the halo effect. The analysis also shows that larger mosques and those with minarets are associated with an accentuated halo effect, suggesting the importance of the salience of minority groups rather than their relative size in influencing political behavior.
Article
Full-text available
The DKI Jakarta elections for 2017-2022 have captured the attention of researchers due to their intriguing nature. One particularly fascinating aspect is the unexpected victory of candidate pair (Paslon) number 3, Anies-Sandi, which was initially not favored by most opinion polls. This study explores how political preference factors, as described in the sociological model, contribute to the formation of base and non-base areas of Anies-Sandi voters in the DKI Jakarta elections for the 2017-2022 period. This study also evaluates the spatial impact of the influence of these social factors. The novelty of research that examines the application of spatial regression in the 2017 DKI Jakarta Pilkada is an interesting innovation that has never been studied before. The base and non-base areas are identified per sub-district with the location quotion (LQ) formula, whereas the identification of global spatial patterns uses the Moran index calculation formula. The free variables in the spatial regression model are designed to adopt a sociological model of political preferences that includes the free variables of sex ratio, percentage of young voters, percentage of highly educated population, percentage of formal sector workers, and percentage of non-Muslim population. The results of this study found that the Anies-Sandi base and non-voter base areas were spread to most sub-districts in DKI Jakarta, but their distribution did not form a particular global spatial pattern. However, it was found that there was an error-type spatial dependency in the first-and second-round LQ models, so it was necessary to perform spatial regression of errors. After spatial regression error, it was found that the sex ratio had a positive and significant effect on LQ in the first round, and the percentage of formal sector workers and the percentage of non-Muslim population had a negative and significant effect on LQ in the first and second rounds. Meanwhile, the lambda coefficient is positive and significant in the first and second rounds of the LQ error spatial regression models, which shows the spatial effect in the residual model. In terms of goodness of fit, the error spatial regression model is more efficient than the OLS model, which is indicated by the relatively low AIC value of the spatial error model compared with the OLS model.
Article
Research on the growing support of radical right parties cannot ignore the role of spatial, contextual effects and geographical proximity. One important spatial factor is the halo effect, which suggests that a radical right party is more likely to perform well in a homogeneous electoral district surrounded by ethnically diverse areas. Voter behaviour is influenced by certain neighbourhood effects, which can be spatially distant, originating from neighbouring areas. In this way, the presence of groups more inclined to support the radical right in an area can, through interaction and observations of signs or norms, modify an individual's party preference. In this paper, I model the 2018 election results of the Sweden Democrats at the level of the 6004 Swedish electoral districts using the Spatial Durbin Model specification. The spatial model shows a better fit and indicates a spillover effect of the level of education and median wealth in addition to the contextual role of ethnic diversity and its change. Changes in the proportion of non‐Europeans in neighbouring areas have had a particularly strong spillover effect, which is associated with the growing support for the Sweden Democrats.
Article
In this paper, we contend that party realignments occur due to the interaction of candidates and activists. We examine independent party candidates who are motivated primarily to win elections but who use activist contributions to increase vote shares. In a two-dimensional policy space, such candidates will on occasion engage in "flanking" moves so as to enlist coalitions of disaffected voters, at the risk of alienating some of their traditional activist supporters. We argue that a result of such "flanking" moves, in the early part of the century, has been a shift in emphasis from an underlying social dimension to the economic dimension. In recent decades, electoral salience has shifted back to the social dimension. The net result is that the party cleavage line is much as it was a century ago-but the parties have switched sides.
Article
Two studies examined the relationship between categorization, intergroup anxiety and intel group attitudes (intergroup bias and negative affect). Study I consisted of a survey of 236 British and Japanese nationals. Study 2 was a longitudinal study of 54 Japanese students studying in the UK. Of the three categorization variables (interpersonal, super-ordinate and intergroup), only intel group categorization was shown to have a relationship to generalized intergroup attitudes. In addition, intergroup anxiety and quality of contact were associated with ingroup bias and negative affect to the outgroup. Study 2 revealed an interaction between intergroup categorization and quality of contact in predicting negative affect. Intergroup anxiety was also associated with increased intergroup categorization. It is concluded that the effects of categorization during contact are still poorly understood and that intergroup anxiety is a far more powerful variable in contact than the current literature acknowledges. Copyright (C) 1999 John Wiley & Sons, Ltd.
Article
Radical right parties have only been successful in a few countries. Why do such a small percentage of voters choose the radical right in Germany? Why is the radical right winning more seats in Austria than in France and Germany? Terri Givens argues that radical right parties will have difficulty attracting voters and winning seats in electoral systems that encourage strategic voting and/or strategic coordination by the mainstream parties. Her analysis demonstrates that electoral systems and party strategy play a key role in the success of the radical right.
Article
The core puzzle which this book resolves is to explain why radical right parties have advanced in a diverse array of democracies--including Austria, Canada, Norway, France, Italy, New Zealand, Switzerland, Israel, Romania, Russia, and Chile--while failing to make comparable gains in similar societies elsewhere, such as Sweden, Britain, and the United States. This book expands our understanding of support for radical right parties by presenting an integrated new theory which is then tested systematically using a wealth of cross-national survey evidence covering almost forty countries.
Article
The political and legislative changes which took place in South Africa during the 1990s, with the dissolution of apartheid, created a unique set of social conditions. As official policies of segregation were abolished, people of both black and white racial groups began to experience new forms of social contact and intimacy.
Article
Support for the British National Party (BNP) has grown exponentially in the last decade. Using a leaked membership list, we locate over 12,000 members and match them with Census data on more than 200,000 neighbourhoods in Britain. Two established theories of ethnic hostility—contact and threat—provide opposing predictions about the effect of the proportion of minorities. These predictions are tested with a multilevel analysis of variation in the probability of white British adults belonging to the BNP. The probability is lower in neighbourhoods with a substantial proportion of non-whites. The probability is higher, by contrast, in cities with a larger proportion of non-whites, but only where they are also highly segregated. Within the non-white category, we find that South Asians matter rather than blacks; results for Muslims are similar. These findings show how contact and threat can be disentangled by considering different spatial scales, and also demonstrate the importance of segregation.
Article
We examine the accessibility of ideological and partisan orientations as factors affecting the political capacity of citizens. In particular, is the utility of partisan and ideological reasoning contingent on the accessibility of an individual's own self-identifications? Are people with accessible points of ideological and partisan orientation more likely to invoke these orientations in formulating political judgments and resisting efforts at political persuasion? Are they more likely to demonstrate politically compatible points of orientation? These questions are addressed in the context of a study conducted during the course of the 1996 election campaign. In order to measure the accessibility of respondents' partisan and ideological self-identifications, we record response latencies--the time required for respondents to answer particular questions. Based on our analysis, we argue that attitudes and self identifications are useful heuristic devices that allow individuals to make sense out of the complexity and chaos of politics. But some citizens are better able than others to employ these devices, and by demonstrating who these citizens are, the concept of accessibility becomes an important element in the explanation of political capacity.