See all ›
96 CitationsSee all ›
64 ReferencesSee all ›
1 FigureEven in Sweden: The Effect of Immigration on Support for Welfare State Spending
Maureen A. Eger
Abstract
While the politics of globalization and welfare state retrenchment have garnered much attention in recent years, scholarly research on public support for welfare state expenditure is comparatively sparse. Furthermore, new pressures, specifically international immigration and resulting ethnic heterogeneity, add a new challenge to the welfare state. In this article, I analyse support for social welfare expenditure in Sweden—the country that spends the greatest percentage of its GDP on social expenditure and, until recently, remained relatively ethnically homogeneous. Results from multilevel models reveal that multiple measures of immigration at the county-level have significant negative effects on support for the welfare state. Moreover, recent immigration has a negative effect on attitudes towards universal spending. Thus, this analysis provides clear evidence that ethnic heterogeneity negatively affects support for social welfare expenditure—even in Sweden.
Even in Sweden: The Effect of
Immigration on Support for
Welfare State Spending
Maureen A. Eger
While the politics of globalization and welfare state retrenchment have garnered much
attention in recent years, scholarly research on public support for welfare state
expenditure is comparatively sparse. Furthermore, new pressures, specifically international
immigration and resulting ethnic heterogeneity, add a new challenge to the welfare state.
In this article, I analyse support for social welfare expenditure in Sweden—the country that
spends the greatest percentage of its GDP on social expenditure and, until recently,
remained relatively ethnically homogeneous. Results from multilevel models reveal that
multiple measures of immigration at the county-level have significant negative effects on
support for the welfare state. Moreover, recent immigration has a negative effect on
attitudes towards universal spending. Thus, this analysis provides clear evidence that ethnic
heterogeneity negatively affects support for social welfare expenditure—even in Sweden.
Introduction
Scholars have devoted considerable energy to under-
standing how global changes affect the sustainability of
welfare states—especially those states that spend
substantial chunks of their GDP on welfare. Many
have feared that globalization, defined typically as the
extension of worldwide capitalism, would lead to a
veritable ‘race to the bottom’ and to either the death of
the European welfare states or its convergence on a less
generous model (Scharpf, 1991; Kitschelt, 1994;
Steinmo, 1994; Rhodes, 1998; Stephens, Huber and
Ray, 1999). The logic of this argument stems from
Okun’s (1975, p. 51) famous claim that economic
‘efficiency is bought at the cost of inequalities in
income and wealth’. In reality, there is very little
evidence connecting globalization to welfare state
retrenchment (Huber and Stephens, 2001, 2005).
Nevertheless, over the past 20 years, most countries
have made changes to social programs that ultimately
reduce the provision of social welfare benefits; how-
ever, these changes have been small, incremental, and
have not necessarily come at the expense of equality
(Pontusson, 2005).
Considering the durability of institutions in general
(North, 1990), this is not too surprising; yet, new
pressures, in the form of immigration-generated diver-
sity, add a new challenge to the welfare state. One only
needs to pick up a newspaper or turn on the television to
see that immigration and increasing ethno-cultural
heterogeneity have become some of the greatest sources
of political turmoil in recent years. In some journalistic
accounts of these demographic changes, the politics of
immigration are explicitly linked to attitudes about
social welfare expenditure. For example, in a piece about
the Nordic welfare states, a Swedish blue-collar union
negotiator explains, ‘Sweden is a small country ....Up
to 10 years ago it was very homogeneous as a country.
Everything was very alike. Up until then all Swedes
looked the same; almost thought the same. Because
we are all so equal, we can share the pain of the
problems .... As Sweden gets more divided, it’s more
difficult to keep this idea of sharing the pain’.
1
The politics of institutional change and welfare state
retrenchment have garnered much attention in recent
European Sociological Review VOLUME 26 NUMBER 2 2010 203–217 203
DOI:10.1093/esr/jcp017, available online at www.esr.oxfordjournals.org
Online publication 31 March 2009
ßThe Author 2009. Published by Oxford University Press. All rights reserved.
For permissions, please e-mail: journals.permissions@oxfordjournals.org
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
years; however, scholarly research on public support
for welfare state expenditure is comparatively sparse.
Furthermore, while other research has considered the
sustainability of welfare state spending due to demo-
graphic change, none has looked systematically at the
effect of increasing ethnic heterogeneity on individuals’
attitudes towards social expenditure—in particular,
spending that benefits the collective. Therefore, this
article seeks to explore the effects of demographic
change on citizens’ support for the welfare state in
Sweden—the country that spends the greatest percent-
age of its GDP on social expenditure and, until
recently, remained relatively ethnically homogeneous.
Case Study: Sweden
Sweden stands out as arguably the most egalitarian,
humanitarian, and democratic country in the world.
Indeed, all the key indicators point to just that. With
consistently high economic growth and a strong
welfare state that ensures low inequality,
2
Swedes
enjoy a high standard of living, universal health care,
education, and generous unemployment benefits.
Further, the Economist Intelligence Unit’s (2007)
index of democracy ranks Sweden as the most
democratic country, with a near perfect score.
3
And,
in a country where on average 80 per cent of
the population votes in national elections, Swedish
economic and social policies reflect a desire to
include marginalized and potentially marginalized
populations.
Sweden remained ethnically, linguistically, and
religiously homogeneous well into the 20th century.
In 1900, less than 1 per cent of the population was
foreign-born; by 1950 this figure had only increased to
2.8 per cent. Because of this traditional homogeneity,
economic and social inequality once constituted the
only salient divisions in Sweden (Hammar, 1985,
p. 22). The belief that Swedes were ‘all in the same
boat’
4
contributed to the development of the most
generous welfare state in the world and legislators
crafted policies in order to eradicate inequality based
on class and gender. However, in the recent years,
successive and relatively large waves of immigration
have prompted the Swedish government to pass new
legislation to protect other minority groups from
discrimination as well as create economic incentives
for both employers and workers in order to integrate
new populations of immigrants into the labor market.
Currently, Sweden devotes more than 30 per cent of its
GDP on social expenditure—more than any other
country in the world.
Although Sweden is not generally thought of as a
country of immigration, in reality, it has become just
that (see Table 1). In a short time, Sweden has
transitioned from a fairly homogenous society to one
characterized by ethnic, linguistic, and cultural diver-
sity. In 1960, foreign-born persons made up a mere 4
per cent of Sweden’s population, with well over half
emigrating from Finland, Denmark, and Norway.
Furthermore, a relatively small percentage of these
immigrants were refugees. By the late-1970s and
Table 1 Population by country of birth
Country of origin 1970 1980 1990 2000 2006
Sweden 7,539,318 7,690,282 7,800,185 7,878,994 7,938,057
Other Nordic countries 320,913 341,252 319,082 279,631 273,991
EU27 without Nordic countries 139,194 151,349 175,679 187,883 216,488
Rest of Europe 41,161 54,402 71,167 174,482 203,431
Africa 4,149 10,025 27,343 55,138 75,405
Asia 5,949 30,351 124,447 220,677 309,606
North America 15,626 14,484 19,087 24,312 27,168
South America 2,300 17,206 44,230 50,853 58,221
Oceania 558 962 1,866 2,981 3,642
Soviet Union
a
7,244 6,824 7,471 7,584 6,667
Foreign-born total 537,585 626,953 790,445 1,003,798 1,175,200
Total residents 8,076,903 8,317,235 8,590,630 8,882,792 9,113,257
Foreign born as percentage of all residents 6.7 7.5 9.2 11.3 12.9
Source: Befolkningsstatistik, ‘‘Tabeller o
¨ver Sveriges befolkning fo
¨r 2006,’’ Statistiska Centralbyra
˚n, 2007.
(http://www.scb.se/statistik/_publikationer/BE0101_2006A01_BR_BE0107TAB.pdf).
a
Includes data for Russia and other former Soviet states through 1990. Some recent immigrants from the former Soviet Union have not been
naturalized into any of the 15 states, which is why there are data for 2000 and 2006.
204 EGER
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
early-1980s, immigration patterns had changed sub-
stantially and the number of asylum-seekers sky-
rocketed. Over the past few decades, refugees have
come from as far away as the former Yugoslavia, Chile,
North Africa, and the Middle East. By the mid-1990s,
over 10 per cent of all Sweden’s inhabitants were
foreign-born, and 13 per cent were either foreign-born
or Swedish-born with two foreign-born parents.
Currently, 13 per cent of the population is foreign-
born—which is the same percentage foreign-born as in
the United States—and 16.7 per cent is either foreign-
born or Swedish-born with two foreign-born parents.
Twenty per cent has at least one foreign-born parent.
5
Although the issue was curiously omitted from the
political debates leading up to the 2006 national
election, it is obvious that immigration has become a
national phenomenon. As a percentage of the total
population, the largest numbers of foreign-born people
reside in the county of Stockholm, which includes the
capital city of Stockholm, its districts, and 26 other
municipalities. However, the municipality with the
highest concentration of foreign-born residents
includes the southern city of Malmo
¨, located only a
bridge’s distance from Copenhagen, Denmark.
Segregation in and near these cities is rampant, with
a large proportion of foreign nationals and foreign-
born living in suburban ghettos, such as Stockholm’s
Rinkeby or Malmo
¨’s Rosega
˚rd.
Today, almost all immigrants come to Sweden
seeking asylum or family reunification (Swedish
Migration Board, 2007a,b).
6
According to the United
Nations High Commission for Refugees, in 2006, Iraqis
became the largest nationality to seek asylum in
Europe. In that year, over 40 per cent of all Iraqi
applications for asylum in Europe went to Sweden, and
according to Statistics Sweden, 91 per cent, or 9,500
Iraqis, received residence permits. This increased the
number of Iraqis residing in Sweden to 83,000. In
2007, the total number of asylum seekers doubled from
the previous year to over 36,000. For a country of only
9.1 million people, this is no small thing. While most
countries require that refugees prove that their lives are
in extreme peril, from 2003 to 2007, Sweden made an
exception for Iraqis, who needed only to demonstrate
that he or she was fleeing central or southern Iraq to
receive asylum. Moreover, Sweden’s updated Aliens
Act (Sweden’s Ministry for Foreign Affairs, 2006)
expands the definition of refugee to include people
who are persecuted because of their gender or sexual
orientation.
Unlike other European countries that have been
reluctant to recognize that immigrants do not necessarily
desire to return to their countries of origin, Sweden’s
forward-thinking policies have made it relatively easy for
refugees to obtain citizenship. In fact, it is easier for
foreigners who come to Sweden with asylum permits
to apply for citizenship than it is for those who come
with work or study visas. Granting citizenship, of course,
stimulates more immigration, and according to the
Swedish Integration Board (2006), family reunification is
the most prevalent reason for immigrating to Sweden.
Thus, in a relatively short period of time, Sweden
has evolved into a multiethnic society. Do Swedes still
possess the same sense of solidarity and believe they
are all in the same boat, or have emergent in-group/
out-group boundaries changed how individuals feel
about universal social welfare benefits? I argue that
increasing ethnic heterogeneity makes real and per-
ceived differences more salient. This in turn affects
people’s willingness to contribute to the collective
welfare when the collective now includes people not in
one’s in-group.
In the following sections, I first offer a critique of
previous explanations of support for the welfare state
and show that the majority of these approaches assume
that popular attitudes are either functions of economic
position or regime type. Moreover, these approaches
assume attitudes are relatively static and immune to
broader societal change. Then, I review the micro-level
theory of in-group bias, from which I derive my
hypothesis about the effects of immigration-generated
diversity on attitudes towards the welfare state.
Explanations of Support for Welfare State
Expenditure
According to the power-resource model, which is
considered the dominant approach to the study of the
welfare state, individual class position predicts one’s
level of support for welfare state expenditure. More
specifically, low-income individuals favour redistribu-
tion and therefore support generous welfare state
spending. Those with high incomes do not. Thus,
this theory sees the welfare state as the outcome of
distributive conflicts among class-based actors. It is the
relative power of these actors that is significant for
policy outcomes (Korpi, 1980; Esping-Andersen, 1985,
1990; Huber and Stephens, 2001).
Iversen (2005, p. 5) maintains that the power-
resources model fails to see the relationship between
production and social protection. He argues that
capitalists also have an interest in a strong welfare
state as well because of the ‘Janus-face’ of this
particular institution. The welfare state is not just
a mechanism for redistribution; it is also social
EFFECT OF IMMIGRATION ON SUPPORT FOR WELFARE STATE SPENDING 205
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
insurance. ‘The welfare state is simultaneously an arena
for distributive struggles and a source of comparative
advantage’ (Iversen, 2005, p. 13; emphasis in original).
According to Iversen, protection against risk is
essential for a successful economy, especially as
countries enter the world market. Without this
insurance, employers or employees would have little
incentive to invest in new technologies or human
capital. Capitalists and other wealthy individuals,
therefore, who may not enjoy the redistributive
properties of the welfare state, nonetheless support its
existence in order to insure against risk.
Both these theories contend that one’s individual
economic position determines support for social
welfare expenditure; in fact, the later posits that it is
in the economic interest of all—both low income and
high income—individuals to support the welfare state.
However, variation in welfare state regimes (Esping-
Andersen, 1990; Hall and Soskice, 2001; Huber and
Stephens, 2001; Pontusson, 2005) suggests the role of
institutional variables in explaining these attitudes. The
comparative welfare state literature maintains that
different institutional arrangements affect the structure
of support for social welfare expenditure. Specifically,
regimes with ‘universal’ versus ‘selective’ or means-
tested social programs are more likely to garner higher
levels of public support (Korpi, 1980; Esping-
Andersen, 1990, 1999; Forma, 1997; Svallfors, 1997;
Gilens, 1999). In other words, social programs that
potentially benefit all are more likely to enjoy general
popularity, whereas programs that benefit the poor
or another subset of the population are less likely
to receive widespread support. According to this
approach, individual income—as well as age and
gender—help explain attitudes towards the means-
tested programs.
Other scholars argue that preferences, norms, values
embedded in different institutional arrangements or
welfare state regimes explain variation in attitudes
towards social welfare expenditure. Brooks and Manza
(2007) find between-country variation in the relative
contribution of economic considerations, social cleav-
ages, and institutions on welfare state policy preferences.
However, their ‘embedded preferences’ approach to
explaining both attitudes and welfare state output
suggest that policies and the support for them are
merely artifacts of a particular country or institutional
type. Similarly, Rothstein (1998) argues that a country’s
political institutions give rise to social norms; policies
congruent with the social norms act to reinforce those
norms, thus indirectly legitimating the political institu-
tion. According to Rothstein, ‘a just system must
generate its own support’ (Rothstein, 1998, p. 143).
However, Rothstein’s approach to explaining support
for the Swedish welfare state implies that citizens’
preferences are uniform and do not change.
Many have identified ‘American exceptionalism’, or
Americans’ distinctive cultural values, as the key
obstacle to widespread support for social spending in
the United States (e.g., Lipset, 1963). Tocqueville
([1848]1969), who admired Americans’ experiment
with democratic rule, was the first to document
Americans’ reverence for the individual and preoccu-
pation with their own self-interest. Feldman and Zaller
(1992) find that economic individualism, or ‘the
commitment to merit as the basis for the distribution
of rewards in society and the belief that people ought
to work hard’, contributes to negative attitudes
towards welfare. According to these authors,
Americans do not like the idea of others getting
something for doing nothing, whether or not that is
the reality of the situation, and this generates
opposition to increases in social expenditure. Values
and norms may help us understand mean differences
in aggregated country-level attitudes, but by relying on
these approaches we gain little in explaining variation
in support for social spending.
A fourth approach points to institutional processes
to explain attitudes towards social welfare expenditure.
Pierson (1994, 2000) finds that individual interests,
even the interests of the ruling party, matter less than
institutional effects. Specifically, path dependency, or
increasing returns, makes it more difficult to undo
policies than implementing them in the first place. He
argues that welfare states, by creating constituencies,
secure their own support. Even if citizens hate paying
taxes, they nonetheless benefit and therefore remain
attached to public provision. Pierson shows that even
when a conservative, market-liberal party is in power
(as in the cases of Reagan and Thatcher) ‘... the
welfare state stands out as an island of relative stability’
compared to other government policies (Pierson, 1994,
p. 5).
Each of these approaches argues that individual
attitudes are functions of either one’s country of
residence or one’s location in the class structure.
Attitudes may vary from class position to class position
or from welfare regime to welfare regime, but that is it.
While I do not discount the importance of economic
interests and class position on attitudes nor do I
pretend that regime types do not affect support for
social expenditure, I argue that these explanations are
limited in that they presume attitudes are relatively
constant. In other words, these explanations imply
that attitudes change only if one moves up (or down)
the economic ladder or to another country.
206 EGER
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
More importantly, none of these explanations consider
other sources of variation within countries.
Theoretical Framework
A more promising approach to explaining attitudes
towards the welfare state takes into account the politics
of difference, specifically the effects of racial and ethnic
diversity on support for social welfare. In an analysis of
the United States, Luttmer (2001) finds a negative
relationship between nonblacks’ support for welfare
7
and the percentage of black welfare recipients in a
respondent’s community. He also finds a negative
relationship between blacks’ support for welfare and
the percentage of nonblack welfare recipients in one’s
community. Interestingly, he finds no relationship
between the percentage of nonblack recipients and
nonblacks’ support for welfare, or does his analysis
demonstrate a relationship between the percentage of
black recipients and blacks’ support for welfare
spending. This suggests it is not the presence of
‘poor people’ in general or even the presence of black
recipients that affects welfare state attitudes but instead
the presence of welfare recipients who are members of
ethnic out-groups. Indeed, other research confirms
that the greatest difference in support for redistribu-
tion is not between the rich and poor but instead
between blacks and whites (Kinder and Sanders, 1996;
Alesina and La Ferrara, 2000).
Gilens (1999) finds that racial stereotypes about
blacks’ work ethic negatively affect white Americans’
attitudes towards social welfare spending. Furthermore,
he finds the magnitude of this effect greater than the
effects of ‘American individualism’ and economic self-
interest on welfare state attitudes. Fox (2004) assesses
the relationship between the presence of Latinos and
whites’ attitudes towards welfare spending. Fox finds
that the percent Latino does matter in shaping whites’
beliefs about Latinos’ work ethic; however, she also
finds that the type of stereotype that whites hold has
no effect on whites’ welfare spending preferences. In
states with a small population of Latinos, whites are
more likely to believe Latinos are lazy and want to
spend less on welfare. In disproportionately Latino
states, whites are more likely to find Latinos hard-
working. Yet, holding constant their views of blacks,
whites in these states also want to spend less on
welfare. Put more simply, whites’ stereotypes about
Latinos, both positive and negative, decrease support
for policies intended to assist poor Americans. This
finding suggests that the presence of an ethnic
minority itself contributes to attitudes about social
welfare expenditure.
In a relatively short period of time, increased
migration has altered the demographic makeup of
countries worldwide—including Sweden. Do these
changes negatively affect citizens’ attitudes about the
generosity or universality of social welfare benefits?
Recent sociological research demonstrates that the
presence of ethnic minority groups has a negative
effect on Americans’ support for social welfare. I
argue that this relationship is not limited to the
United States and hypothesize that ethnic hetero-
geneity, in this case produced by international
migration, has a negative effect on Swedes’ attitudes
as well.
In an effort to understand better the mechanisms
involved in this relationship as well as provide
empirical support for my argument, I turn to cognitive
social psychological theory and robust experimental
evidence that demonstrates the effects of group
boundaries and in-group bias on resource allocation.
8
Whatever informal or formal group an individual
associates with or belongs to at a certain point in time
constitutes an in-group. These groups can range from
one’s family to one’s sports team or from one’s ethnic
group to one’s nation of origin. According to Festinger
(1954), in-group formation results from the cognitive
process social categorization, and in-groups comprise
individuals who are similar on some salient dimension.
One’s out-group includes individuals who do not
share that salient physical or social trait. In-group
formation is not necessarily a conscious process but
instead an automatic response to one’s environment.
Furthermore, these groups need not be solidaristic or
instrumentally created organizations. More impor-
tantly, an in-group’s boundaries are neither definite
nor immutable. Depending on the circumstances, these
group boundaries may fall in and out of salience as
well as shift to expand or contract over time.
9
In-group bias, or the tendency to favour one’s own
group, is a simple concept but one that has significant
implications for society—especially in the treatment of
in-group and out-group members. Experiments show
that people strongly prefer members of their own
group, even when one’s ‘group’ is the result of random
assignment and has no consequence for the subject
(Tajfel et al., 1971; Brewer, 1979). Using normless
‘minimal’ group situations, social psychologists
demonstrate how easily people will group themselves
using categorical distinctions and then engage in
differential treatment of in and out-group members
(Tajfel, 1970, 1978, 1981, 1982; Tajfel and Billig, 1974).
It is important to note, however, that in-group bias
does not always lead people to discriminate against an
out-group in an overt fashion. Allport (1954, p. 42)
EFFECT OF IMMIGRATION ON SUPPORT FOR WELFARE STATE SPENDING 207
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
maintains that in-group attachment, preference, and
favouritism are not necessarily reciprocally related to
out-group contempt, hostility, or overt discrimination,
and Brewer’s (1999) research confirms that positive in-
group identification does not systematically correlate
with negative attitudes towards or treatment of out-
groups. In other words, although immigration makes
national and ethnic in-group/out-group boundaries
salient, in-group bias does not necessarily lead to anti-
immigrant sentiment or hostility towards foreigners.
An illustrative example comes from Norris’s (2005)
analysis of cross-national variation in the rise of radical
right parties. She does not find a significant relation-
ship between the percentage of a country’s residents
that is foreign-born or the number of asylum-seekers
with residence permits and the electoral success of
political parties with xenophobic platforms.
Although in-groups do not always confer costs to
out-groups, in-group members are nevertheless willing
to differentially distribute benefits to their own group
versus members of other groups (Tajfel et al., 1971;
Turner, 1975, 1981, 1984; Brewer, 1979; Tajfel and
Turner, 1986; Mullen, Brown, and Smith, 1992). This
robust finding is attributed to members’ motivation to
maintain a group’s positive identity, distinctiveness,
and boundaries. These behaviours emerge not only in
‘minimal’ group settings but also amongst natural
groups (Brewer and Campbell, 1976).
I argue that as countries and sub-regions experience
increases in ethnic heterogeneity due to international
migration, citizens within that geographic unit are less
likely to support generous welfare state expenditure
that benefits the collective—even though this means
that they too would receive fewer benefits (see
Figure 1). Indeed, experimental research shows that
in-group members will allocate resources so that there
is the greatest difference between groups, even if this
means that in-group members receive a lesser amount
(Tajfel et al., 1971). Thus, I posit that immigration will
have a negative effect on attitudes towards not only
means-tested spending but also universal spending.
I use data from Sweden because of its strong social
democratic tradition and welfare state institutions as
well as its recent history of increasing immigration.
Therefore, an analysis of Swedes’ attitudes will be an
especially good test of the theory.
Data and Methods
I rely on attitudinal data, A
˚sikter om den Offentliga
Sektorn och Skatterna, or ‘Opinions about the Public
Sector and Taxes’, collected by Stefan Svallsfors and his
colleagues at Umea
˚University in Sweden (Svallfors,
1986; Svallfors et al., 1992; Svallfors and Edlund, 1997,
2002). The sample is a pooled cross-section with data
from 1986, 1992, 1997, and 2002. I sort subjects by la
¨n
code, Sweden’s first level administrative and political
subdivision,
10
and attach regional-level census data
from the Swedish census, Statistiska Centralbyra
˚n. The
primary responsibility of Sweden’s 21 la
¨n is the
provision of welfare benefits such as healthcare and
education.
For my dependent variable, I use an index of
attitudes towards social expenditure. For each of the
seven spending items, I coded responses so that
respondents who want to see an increase in
spending receive a 2; those who want spending to
stay the same receive a 1; and those who want to see
spending decreased receive a 0. Missing values for
any of the seven items were assigned the mean value
for that year. In order to create an index that ranges
from 0 to 100, respondents’ scores were then
aggregated to give a range of 0–14, divided by the
maximum value of 14, and multiplied by 100.
11
Figure 2 shows the mean scores for the spending
index (left y-axis) and means scores by spending item
(right y-axis) over time.
I also created a spending index that includes only
items that are universal in nature. This second dependent
variable measures attitudes towards health care, elder
care, and primary/secondary school expenditure.
Ethnic heterogeneity - Public support for social welfare
Social identity Individual support for social welfare
Social categorization
In-group bias
Figure 1 Theoretical model
208 EGER
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
The first set of independent variables includes
individuals’ sociodemographic characteristics.
12
I
include a continuous measure of age and a dichot-
omous measure of gender. Education is an ordinal
variable, where respondents choose their highest level
of education (0 ¼primary; 1 ¼secondary; 2 ¼some
university/college; 3 ¼university or college degree).
Because one of the items included in the dependent
variable measures attitudes towards social spending on
families with children, I also include a question about
whether or not there is a child in the respondent’s
home. I use Kingston’s (2000, p. 85) conception of
‘income classes’ to capture my subjects’ economic
position. I created four income categories based on the
median disposable income for each year: below the
median; 100–200 per cent above the median; 200–300
per cent above the median; and above 300 per cent of
the median.
While all Swedes enjoy universal access to health
care, education, childcare, elder care, and economic
support for children under the age of 16, the next set
of independent variables captures the respondents’
relationship to means-tested spending. I include two
variables that ask whether or not respondents have
been unemployed for at least a week in the past 3 years
and whether or not respondents have received social
assistance in the last 3 years. The next two questions
measure political affiliation and union membership.
Sweden is a multi-party, parliamentary democracy;
however, the Social Democrats, who are the architects
and long-time protectors of Sweden’s universal and
generous welfare state, have governed Sweden 65 out
of the last 75 years. The Left Party and the Green Party
are also mainstream, relevant parties that are consis-
tently pro-welfare state.
13
Therefore, I use a dichot-
omous measure of party affiliation where respondents
are coded as either supporting the Social Democratic,
Left, or Green Party or supporting any other party.
Union membership is also a dichotomous variable.
In order to assess the effects of demographic,
political, economic variables on Swedes’ attitudes
towards social expenditure, I sort subjects by county
and year and then attach regional-level census data
that corresponds with the survey year. I use three
different measures of immigration in my analyses: the
percentage of the county population who immigrated
to Sweden in that calendar year, the percentage of
the county population that is a foreign citizen, and
Figure 2 Mean scores for attitudes towards social expenditure, by type and year
EFFECT OF IMMIGRATION ON SUPPORT FOR WELFARE STATE SPENDING 209
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
the percentage of the county population that is
foreign-born. I use three other county-level variables
as controls: the percentage of the county’s working-age
population that is employed, the percentage of the
population that receives social assistance, and the
percentage of the votes cast for the Social Democratic,
Left, and Green parties in the last election. Finally,
I include a year variable in order to control for
unobserved effects of time.
Results
I fit a two-level, random intercept model using the
program gllamm for Stata (Rabe-Hesketh and
Skrondal, 2005). In Table 2, I report results from my
analysis of Swedish attitudes towards social expendi-
ture, 1986–2002.
Model 1 is the empty model, where we can observe
the amount of the total variance in the dependent
variable. In Model 2, I assess the effects of the
individual-level variables. As expected, individual-level
variables affect support for social expenditure.
14
One’s
economic position, educational attainment, and history
of receiving social assistance affect Swedes’ support for
the welfare state in the manner predicted by the
power-resource model. Gender, parenthood, union
membership, and left party affiliation positively affect
support.
When the effects of county-level variables are
assessed in models 3–5, we see that different measures
of immigration consistently affect attitudes towards the
welfare state. Each unit increase in the percentage of
new immigrants in that year decreases support for the
welfare state by 8.84. This measure varies between
0.18 and 1.00 per cent (m¼0.46), demonstrating that
in counties with a high proportion of recent
immigrants, respondents are significantly less suppor-
tive of the welfare state. To ensure that this variable is
Table 2 Multilevel models of support for social welfare expenditure
(1) (2) (3) (4) (5)
Intercept 66.15 (0.34) 59.09
(1.31) 55.24
(11.13) 60.77
(9.71) 53.02
(11.38)
County-level variables
Percent immigrant (year) 8.84 (1.59)
Percent foreign citizen 0.55
(0.17)
Percent foreign born 0.47 (0.14)
Percent employed 0.01 (0.11) 0.08 (0.09) 0.02 (0.11)
Percent social assistance 0.30 (0.23) 0.54y(0.28) 0.63(0.25)
Percent left vote 0.04 (0.05) 0.02 (0.07) 0.03 (0.06)
Year 3.01 (0.43) 2.33
(0.42) 3.07 (0.53)
Individual-level variables
Gender (ref. ¼male) 4.01
(0.53) 3.69 (0.52) 3.69
(0.51) 3.70 (0.51)
Age 0.00 (0.02) 0.00 (0.02) 0.01 (0.02) 0.01 (0.02)
Education 1.49
(0.22) 1.59 (0.23) 1.61
(0.23) 1.61 (0.23)
Income (ref. ¼100–200 per cent)
Below median 1.14(0.52) 1.99 (0.47) 1.96
(0.48) 1.95 (0.47)
200–300 per cent median 6.55
(0.73) 6.71 (0.75) 6.69
(0.73) 6.64 (0.72)
Above 300 per cent median 9.09
(1.02) 8.76 (1.03) 8.80
(1.07) 8.69 (1.07)
Child (ref. ¼none) 2.13
(0.53) 1.95 (0.57) 2.01
(0.56) 2.00 (0.57)
Union (ref. ¼no) 2.86
(0.65) 2.94 (0.69) 2.97
(0.70) 2.93 (0.70)
Left party (ref. ¼no) 7.82
(0.67) 8.01 (0.67) 7.97
(0.67) 7.99 (0.67)
Unemployed (ref. ¼no) 1.68(0.74) 1.08 (0.73) 1.09 (0.71) 1.09 (0.72)
Social assistance (ref. ¼no) 10.17
(1.07) 9.81 (1.06) 9.98
(1.03) 9.87 (1.04)
Variances and covariances of random effects
Level 1 288.74 (10.41) 244.36 (8.39) 237.28 (8.38) 237.71 (8.43) 237.42 (8.53)
Level 2 4.49 (1.19) 1.60 (1.23) 0.81 (0.54) 1.31 (1.23) 0.88 (.60)
Log-likelihood 19,689.696 1,4291.243 14,238.435 14,243.140 14,239.754
N4,628 3,427 3,427 3,427 3,427
yP50.10; P50.05; P50.01; P50.001.
Group variable is la
¨n code. Robust standard errors reported in parentheses.
210 EGER
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
not acting as a proxy for some other feature of Swedish
counties, in Model 4, I replace percent immigrant with
the percentage of residents that are foreign citizens and
in Model 5 with the percentage of residents that are
foreign-born. Each unit increase in percent foreign
citizen reduces support by 0.55. This measure varies
between 0.86 and 10.05 per cent (m¼4), thus this
measure of immigration also has a large effect on
attitudes. Nevertheless, Model 5, which assesses the
effects of percent foreign-born, has a better fit than
Model 4. This is not surprising, as it may be difficult
or impossible to distinguish between foreign citizens
and naturalized Swedish citizens, although signs of
cultural integration (i.e., style of clothing) could
indicate the latter. A unit increase in the proportion
of residents who are foreign-born decreases support for
the welfare state by 0.47. This measure has a
minimum value of 2.84 per cent and maximum
value of 18.12 per cent (m¼7.74), producing an
effect that rivals many of the individual-level variables.
Employment rates do not have an effect on
individual-level attitudes towards social spending.
This finding may be partly due to the nature of the
Swedish welfare state itself. This institution depends on
high employment in order to sustain itself; therefore,
unemployment is treated like the common cold rather
than a chronic disease. The state provides education
and assistance to the unemployed in order to ensure
that its tax base remains high. Consequently, inequality
is low and the standard of living in Sweden is relatively
high, which affects the salience of economic differ-
ence.
15
To be sure, economic hardship is not as visible
as it is in other countries, which suggests that class
is not as salient a social division as ethnic difference.
As a result, the latter is more likely to trigger social
categorization and in-group bias. Nevertheless, the
proportion of county residents who receive social
assistance appears to have a positive effect on attitudes.
I also regress a dependent measure of universal
spending items (education, health care, and elder care)
on the same county-level and individual-level variables.
In Table 3, I report the results from the only full
model where a measure of immigration is significant.
Conclusion
Previous research on the political sociology of the
welfare state has not considered seriously the role of
immigration on support for social expenditure. While
institutional theories help explain differences in
countries’ aggregate levels of support and welfare
state output, they are less helpful in accounting for
attitudinal heterogeneity within a population. Results
from the current analysis certainly support power
resource theory and the importance of economic
interests for welfare state attitudes. However, this
research also provides evidence that the proximity and
salience of an out-group negatively affects attitudes
about the allocation of resources.
Both recent immigration and proportion foreign-
born reduce support for the Swedish welfare state.
Indeed, immigration is the only county-level variable
that negatively affects Swedes’ attitudes, and it is the
only regional variable that is significant across models.
When universal spending is assessed by itself, recent
immigration still has a negative effect on support,
which demonstrates that these analyses are not merely
picking up the effect of immigration on support for
redistribution. Regardless, the redistributive nature of
the Swedish welfare state is an integral part of the
institution and is inextricably linked to its ideological
roots—the idea that all Swedes are ‘all in the same
boat’. In an analysis of attitudes in eight countries in
1992, Svallfors (1997) finds Swedes’ support for the
redistributive nature of the welfare state to be lower
than other social democratic and even conservative
Table 3 Multilevel model of support for Universal
Social Welfare Expenditure
bSE
Intercept 58.30 13.82
County-level variables
Percent immigrant (Year) 6.47 2.38
Percent employed 0.05 0.13
Percent social assistance 1.11 0.22
Percent left vote 0.07 0.07
Year 7.80 0.36
Individual-level variables
Gender (ref. ¼male) 4.37 0.63
Age 0.01 0.02
Education 0.89 0.27
Income (ref. ¼100–200 per cent)
Below median 1.25y0.68
200–300 per cent median 4.25 1.05
Above 300 per cent median 5.42 1.22
Child (ref. ¼none) 2.13 0.55
Union (ref. ¼no) 3.29 0.54
Left party (ref. ¼no) 3.14 0.57
Unemployed (ref. ¼no) 0.46 0.87
Social assistance (ref. ¼no) 2.14 1.23
N3,427
yP50.10; P50.05; P50.01; P50.001.
Group variable is la
¨n code. Robust standard errors reported.
EFFECT OF IMMIGRATION ON SUPPORT FOR WELFARE STATE SPENDING 211
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
regimes. Results from the current analysis suggest that
increasing immigration may be part of this story.
These analyses only include data from 1986 to 2002;
thus, it is difficult to know if overall support or
support for specific programs was higher in previous
decades or has declined in the recent years. Rates of
immigration, however, have increased, and results
from this analysis suggest that recent immigration
would negatively affect current levels of support.
Would, then, a shift in public opinion pose a real
threat to the Swedish welfare state? Institutional change
itself tends to be incremental (North, 1990); however,
if the democratic theory of politics (Downs, 1957;
Dahl, 1961) is correct, changes in public opinion will
eventually lead to changes in leadership and policy.
Recent research does demonstrate that public opinion
plays a significant role in policy outcomes. In fact,
when public opinion is considered, other political
factors such as party strength or political organization
recede in importance (Burstein, 1999; Burstein and
Linton, 2002). Brooks and Manza (2007) find that
mass public opinion is central in explaining welfare
state expenditure and the variation between welfare-
state regimes in their welfare state effort. The
researchers also find that aggregate public opinion
tends to change slowly and over time due to the effects
of partisan hegemony and institutions.
Sweden, however, has already seen a change in
leadership, which proves a rare event in a country
where one political party has had incredible, if not
hegemonic, authority and influence. While no right-
wing parties with a xenophobic platform garnered
enough votes to earn a seat in Parliament, the 2006
election did result in a historic shift towards more
liberal economic policies and the ousting of Social
Democrats from the prime ministry and the govern-
ment’s ruling coalition. Prime Minister Fredrik
Reinfeldt, leader of the Moderate Party, and the
ruling coalition have already brought about changes
to the welfare state—reducing spending on both
universal and means-tested programs.
16
Despite the fact that Sweden’s economy grew 5.6 per
cent in the last quarter of 2006, employment and social
welfare expenditure were the big issues of the election.
The centre-right alliance campaigned for a reduction
in taxes and the parties’ leaders said that too many
people relied on welfare for too long. Immigration
as an issue did not come up, as both coalitions claimed
to desire liberal immigration policies. Nevertheless,
downsizing the welfare state is one way to limit
immigrants’ desire to come to Sweden. A reduction
in taxes and expenditure also could mean that fewer
benefits go to immigrant populations.
In Sweden, it is taboo to discuss immigration as a
political issue, even in an election year, which is one
way that its political climate differs greatly from other
countries—the United States being the most obvious
example. Public discussions about ethnic, religious, or
cultural diversity are viewed as politically incorrect,
and political rhetoric describes multiculturalism as a
public good versus a challenge that needs attention.
Nevertheless, economic policies and demographic
make-up are inextricably linked even if politicians do
not make that connection salient. The only party that
weighs in on the growing number of immigrants is the
right wing party, the Swedish Democrats, who won
local seats in southern Sweden in the 2006 election.
Unlike New Democracy, which experienced brief
electoral success in 1992, this party’s platform is both
anti-immigrant and pro-welfare.
Recent research points to the effects of ethnic and
racial heterogeneity on actual social spending. Alesina
and Glaeser (2004) find that racial and ethnic
fractionalization explains a significant portion of the
variation between social expenditure in the United
States and in European welfare states. Such findings
suggest that ethnic heterogeneity and in-group bias,
disguised as individualism, may have played a role in
the development of the American liberal welfare state.
Lipset and Marks (2000, p. 267) maintain that ‘ethnic,
racial, and religious cleavages were more powerful
sources of political identity for most American workers
than was their commonality as workers’ and that
working-class cleavages contributed to the failure of
socialism in the United States. While the exceptional
diversity that the United States has experienced since
its birth as a nation distinguishes it from other
advanced post-industrial societies, globalization and
increased immigration may serve to decrease
‘American exceptionalism’ as other Western democra-
cies also become increasingly multiethnic. Future
research is needed to investigate other sociological
implications of immigration-generated diversity.
Data limitations do not permit me to answer a
number of important questions about attitudinal
change. This research relies on pooled, cross-sectional
data and therefore cannot speak to whether or not
individual attitudes have changed over time. As
previously mentioned, the data only cover a 16-year
period. Thus, we do not know if aggregate support for
social expenditure was higher in previous decades or if
overall support for the welfare state declined since
2002. Future research is necessary. Comparative
analyses are also important to gauge the effect of
immigration on welfare state attitudes in Sweden
relative to its effect in other countries. Is this effect
212 EGER
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
stronger in countries with weaker welfare states or in
countries with a different history of immigration? Do
cross-national differences in immigration and integra-
tion policies temper or amplify the effects of in-group
bias?
Dahl (1996, p. 642) noted that altruistic tendencies
diminish as the scope of the group increases. In a
homogeneous group, ‘egoism merges indistinguishable
with altruism;’ yet, ‘as the group expands in numbers,
as homogeneity declines, and as conflicting interests
increase’ it becomes more difficult to sacrifice one’s
own interests in order to advance another’s. Due to
recent large-scale immigration, ethnicity has become a
salient social division in Sweden. Results from the
current analysis demonstrate that the presence of
foreign-born residents negatively affects support for
social expenditure in Sweden—the country that boasts
the most generous and inclusive welfare state in the
world. Immigration is not the entire story, of course,
but this analysis reveals something significant that
previous research has not.
Notes
1. Cowell, A. Letter from Sweden: An Economy
with Safety Features, Sort of Like a Volvo. The
New York Times, 10 May 2006. The quote
belongs to Ingemar Goransson.
2. In 2000, Sweden’s Gini coefficient was 25 versus
the United State’s 40.81.
3. Incidentally, the United States was ranked 17 of
28 ‘full democracies’.
4. Vi sitter alla i samma ba
˚t.
5. Befolkningsstatistik (2007), Statistiska
Centralbyra
˚n, http://www.scb.se.
6. Tobias Billstro
¨m, Minister of Migration and
Asylum Policy, argues that Sweden must shift
from an immigration policy based solely on
asylum-seeking to one that also includes high-
skilled labor migration. He said that diversifying
the ways in which one migrates to Sweden
should have an effect on attitudes towards
immigration in general (Personal interview with
Tobias Billstro
¨m, 13 June 2007). In May 2008,
the government submitted a bill to parliament
proposing new rules that would make it easier
for Swedish firms to recruit skilled workers from
outside Sweden and the EU.
7. It is important to note that in the United States,
the general public perceives ‘welfare spending’ to
include only unemployment insurance and other
forms of means-tested spending. Thus, it is safe
to assume that this question measures attitudes
towards this type of spending and not attitudes
towards the prototypical welfare state, which
includes other types of expenditure. Luttmer’s
analysis and conclusions also imply this.
8. While testing this micro-level hypothesis is
beyond the scope of the current project, I rely
on robust experimental evidence in order to
strengthen my argument.
9. For example, one could argue that for a period
after September 11, many ethnic, religious, and
political group boundaries within the United
States declined in salience as the terrorist attack
and aftermath heightened other group bound-
aries based on nationality. Or, to use an example
from Sweden, in the postwar period, Finns
constituted the largest and most prominent
out-group in Sweden (Hammar, 1985).
Although Finland is Sweden’s Nordic neighbor,
Finns are not linguistically (and some have
argued not culturally) Scandinavian. Today,
however, the Finnish experience in Sweden
is much closer to the experience of the native-
born. Finnish immigrants and those with Finnish
background are seen as more culturally similar
than immigrants who come from the Middle East
or northern Africa.
10. Ideally, I would also sort respondents by a lower
geographic unit, the municipality or kommun,
and run additional analyses to measure the
effects of local variables on the dependent
variable. However, because there are 290 munici-
palities in Sweden, the number of cases in the
survey is not large enough to allow for structur-
ing the data set this way. Regardless, because
people’s knowledge of current affairs is not
limited to one’s immediate surroundings (i.e.,
the neighborhood, town, or city), I believe the
la
¨n, or county, is an appropriate choice for a
study that assesses the effects of immigration on
attitudes. People do not need to interact with
immigrants to have knowledge that they live
and/or work nearby. Moreover, considering
over 80 per cent of la
¨n expenditure goes to
the provision of healthcare and education, this
geographic unit has practical relevance for its
EFFECT OF IMMIGRATION ON SUPPORT FOR WELFARE STATE SPENDING 213
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
residents. Furthermore, although the proportion
of municipalities’ population that is foreign-born
varies within la
¨n, it varies less than one might
expect. For example, in 2002, the Stockholm
county municipality with the largest percent
foreign-born was Botkyrka (32.85 per cent).
The municipality with the smallest proportion
of foreign-born residents was Ekero
¨(8.30 per
cent). The city of Stockholm and its districts had
19.22 per cent. The mean was 15.67 per cent with
a standard deviation of 5.68 per cent (N¼26).
Finally, immigrating to Sweden does not mean
that one remains where he or she initially lives.
Although asylum reception centres are located in
over 30 municipalities throughout Sweden,
currently there is no law that mandates that
refugees or their families live in a particular
kommun.
11. The measure I created follows the guide outlined
by Svallfors (2006) and is similar to his six-item
index.
12. An alternative hypothesis is that foreign-born
bring anti-welfare state attitudes with them when
they immigrate to Sweden; therefore in areas
where there is a high concentration of foreigners,
we would find less support for the welfare state.
Thus, it would be ideal if I could also control for
whether or not the respondent is foreign-born or
has foreign background. These measures were not
included in all four data sets. According to the
PI, the sample is nationally representative: all
(legal) residents were included in the sampling
frame, regardless of ethnicity or citizenship, and
foreign-born populations were not over-sampled
in areas with relatively large foreign-born popu-
lations. Furthermore, the questions were asked in
Swedish, which also suggests to me that it is
unlikely that the most recent waves of immi-
grants were included at any of the time points.
Nevertheless, because a measure of foreign back-
ground is available in the 1992 survey, I ran
separate analyses for that year. Respondents who
are either foreign-born or native-born with two
foreign-born parents make up 12.25 per cent of
the sample in 1992, which is consistent with
national statistics. The respondents’ ethnic back-
ground and the dependent variable are weakly
correlated (r¼0.05), and it is important to
emphasize that the relationship is positive.
Furthermore, I regress all individual-level vari-
ables, including ethnic background, on the
dependent variable. Having foreign-background
versus Swedish background does not have a
significant effect on one’s attitudes towards the
welfare state. Even if the findings for 1992 are
not generalizable and recent immigrants are less
supportive of the welfare state, a negative effect is
not inconsistent with the theory. I argue that
ethnic heterogeneity negatively affects support for
the welfare state and not that the presence of
foreigners only affects native-born Swedes’ atti-
tudes. Based on the theory of in-group bias, there
is little reason to believe that immigrants would
be any more likely to support the allocation of
resources to ethnic out-groups than native-born
Swedes.
13. The Left Party has won seats in parliament in
every election between 1985 and 2002. The Green
Party won seats in four of the six elections
during that time.
14. Using attitudes to help explain another attitude is
analytically problematic, especially if the attitudes
are theoretically related. Yet, to be sure that
respondents’ attitudes towards those who receive
means-tested support is not driving their support
for the apparatus of redistribution, I control for
those attitudes in an unreported analysis. Not
surprisingly, attitudes towards people who are
unemployed and people who receive social
assistance have a negative effect on support for
social expenditure; however, including this vari-
able does not alter the results presented here.
15. To verify the robustness of these findings, I also
ran analyses with other level-2 economic con-
trols: wealth tax contributions per capita and real
estate tax contributions per capita. Including
these variables instead of percent employed does
not change the results. Furthermore, neither of
these measures is significant when included
instead of the measures of immigration.
16. At the start of 2007, the government reduced
sickness and unemployment benefits by increas-
ing the fees for benefits and simultaneously
reducing pay from those benefits. The cost of
public transportation increased and museums
began to charge admission instead of remaining
214 EGER
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
free to the public. Union fees are no longer
tax-deductible. In 2008, property taxes and the
wealth tax were abolished.
Acknowledgements
The author thanks Paul Burstein, Trey Causey, Michael
Hechter, Christine Ingebritsen, Edgar Kiser, Sabino
Kornrich, Becky Pettit, Steve Pfaff, Erin Powers, Jens
Rydgren, Sarah Valdez, Charles Westin, and the
participants in the Onsdagsseminar in the Department
of Sociology at Stockholm University and the Seminar
in Institutional Analysis in the Department of
Sociology at University of Washington. Gratitude is
also owed to two anonymous reviewers. This research
was funded in part by a Fulbright Fellowship and a
FLAS Fellowship.
References
Alesina, A. and Glaeser, E. L. (2004). Fighting Poverty
in the U.S. and Europe. Oxford: Oxford University
Press.
Alesina, A. and La Ferrara, E. (2000). Participation
in Heterogeneous Communities. The Quarterly
Journal of Economics,115, 897–904.
Allport, G. W. (1954). The Nature of Prejudice.
Cambridge, MA: Addison-Wesley.
Brewer, M. B. (1979). In-Group Bias in the Minimal
Intergroup Situation: A Cognitive Motivational
Analysis. Psychological Bulletin,86, 307–324.
Brewer, M. B. (1999). The Psychology of Prejudice:
Ingroup Love or Outgroup Hate. Journal of Social
Issues,55, 429–444.
Brewer, M. B. and Campbell, D. T. (1976).
Ethnocentrism and Intergroup Attitudes: East
African Evidence. New York: Halsted Press.
Burstein, P. (1999). Social Movements and Public
Policy. In Giugni, M., McAdam, D. and Tilly, C.
(Eds.), How Social Movements Matter. Minneapolis:
University of Minnesota Press, pp. 3–21.
Burstein, P. and Linton, A. (2002). The Impact of
Political Parties, Interest Groups, and Social
Movement Organizations on Public Policy: Some
Recent Evidence and Theoretical Concerns. Social
Forces,81, 380–408.
Brooks, C. and Manza, J. (2007). Why Welfare States
Persist: The Importance of Public Opinion in
Democracies. Chicago: The University of Chicago
Press.
Dahl, R. (1961). Equality and Power in American
Society. In William, V. and Ehrlich, H. J. (Eds.),
Power and Democracy in America. South
Bend, Indiana: University of Notre Dame Press,
pp. 73–91.
Dahl, R. A. (1996). Equality Versus Inequality. PS:
Political Science and Politics,29, 639–648.
Downs, A. (1957). An Economic Theory of Democracy.
New York: Harpers.
Esping-Andersen, G. (1985). Power and Distributional
Regimes. Politics and Society,14, 223–256.
Esping-Andersen, G. (1990). The Three Worlds of
Welfare Capitalism. Princeton, NJ: Princeton
University Press.
Esping-Andersen, G. (1999). Social Foundations of
Postindustrial Economies. Oxford: Oxford
University Press.
Feldman, S. and Zaller, J. (1992). The Political Culture
of Ambivalence: Ideological Responses to the
Welfare State. American Journal of Political
Science,36, 268–307.
Festinger, L. (1954). A Theory of Social Comparison
Processes. Human Relations,7, 117–140.
Forma, P. (1997). The Rational Legitimacy of the
Welfare State: Popular Support for Ten Income
Transfer Schemes in Finland. Policy and Politics,25,
235–249.
Fox, C. (2004). The Changing Color of Welfare? How
Whites’ Attitudes toward Latinos Influence for
Welfare. American Journal of Sociology,10,
580–625.
Gilens, M. (1999). Why Americans Hate Welfare.
Chicago: University of Chicago Press.
Hall, P. A. and Soskice, D. (Eds.) (2001). Varieties of
Capitalism. New York: Oxford University Press.
Hammar, T. (Ed.) (1985). European Immigration
Policy. Cambridge: Cambridge University Press.
Huber, E. and Stephens, J. D. (2001). Development
and Crisis of the Welfare State: Parties and Policies
in Global Markets. Chicago: The University of
Chicago Press.
Huber, E. and Stephens, J. D. (2005). State Economic and
Social Policy in Global Capitalism. In Janoski, T.,
Alford, R. R., Hicks, A. M. and Schwartz, M. A.
(Eds.), Handbook of Political Sociology: States, Civil
Societies, and Globalization, pp. 607–629.
Iversen, T. (2005). Capitalism, Democracy, and Welfare.
New York: Cambridge.
Kinder, D. R. and Sanders, L. M. (1996). Divided by
Color: Racial Politics and Democratic Ideals.
Chicago: The University of Chicago Press.
Kingston, P. W. (2000). The Classless Society. Stanford,
CA: Stanford University Press.
EFFECT OF IMMIGRATION ON SUPPORT FOR WELFARE STATE SPENDING 215
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
Kitschelt, H. P. (1994). The Transformation of
European Social Democracy. New York: Cambridge
University Press.
Korpi, W. (1980). Social Policy and Distributional
Conflict in the Capitalist Democracies: A
Preliminary Comparative Framework. Western
European Politics,3, 291–316.
Lipset, S. M. (1963). The Value Patterns of Democracy:
A Case Study in Comparative Analysis. American
Sociological Review,28, 515–531.
Lipset, S. M. and Marks, G. W. (2000). It Didn’t
Happen Here: Why Socialism Failed in the United
States. New York: W. W. Norton & Co.
Luttmer, E. P. (2001). Group Loyalty and the Taste for
Redistribution. Journal of Political Economy,109,
500–528.
Mullen, B., Brown, R. and Smith, C. (1992). Ingroup
Bias as a Function of Salience, Relevance, and
Status: An Integration. European Journal of Social
Psychology,22, 103–122.
Norris, P. (2005). Radical Right: Voters and Parties in
the Electoral Market. New York: Cambridge
University Press.
North, D. C. (1990). Institutions, Institutional Change
and Economic Performance. Cambridge: Cambridge
University Press.
Okun, A. (1975). Equality and Efficiency: The Big
Tradeoff. Washington DC: Brookings Institution
Press.
Pierson, P. (1994). Dismantling the Welfare State.
Cambridge: Cambridge University Press.
Pierson, P. (2000). Increasing Returns, Path
Dependence, and the Study of Politics. American
Political Science Review,94, 80–104.
Pontusson, J. (2005). Inequality and Prosperity: Social
Europe vs. Liberal America. Ithaca: Cornell
University Press.
Rabe-Hesketh, S. and Skrondal, A. (2005). Multilevel
and Longitudinal Modeling Using Stata. College
Station, TX: StataCorp Press.
Rhodes, M. (1998). Globalisation, Labour Markets
and Welfare States: A Future of ‘‘Competitive
Corporatism’’? In Rhodes, M. and Me
´ny, Y. (Eds.),
The Future of European Welfare. A New Social
Contract? Basingstoke: Macmillan, pp. 178–203.
Rothstein, B. (1998). Just Institutions Matter.
Cambridge, MA: Cambridge University Press.
Scharpf, F. (1991). Crisis and Choice in European Social
Democracy. Ithaca: Cornell University Press.
Statistiska Centralbyra
˚n/Statistics Sweden. (2007).
Statistical Databases: Befolkningsstatistik,
Arbetsmarknad,Socialtja
¨nst, and
Medborgarinflytande.http://www.scb.se.
Steinmo, S. (1994). The End of Redistribution:
International Pressures and Domestic Policy
Choices. Challenge,37, 9–17.
Stephens, J., Huber, E. and Ray, L. (1999). The Welfare
State in Hard Times. In Kitschelt, H. et al. (Eds.)
Continuity and Change in Contemporary Capitalism.
Cambridge: Cambridge University Press,
pp. 164–93.
Svallfors, S. (1986). A
˚sikter om den Offentliga Sektorn
och Skatterna, 1986. Department of Sociology,
Umea
˚University, Umea
˚, Sweden. http://www.ssd.
gu.se/.
Svallfors, S. (1997). Worlds of Welfare and Attitudes
to Redistribution: A Comparison of Eight
Western Nations. European Sociological Review,3,
283–304.
Svallfors, S. (2006). The Moral Economy of Class. Class
and Attitudes in Comparative Perspective. Stanford,
CA: Stanford University Press.
Svallfors, S., A
˚berg, R., Edlund, J. and Hallero
¨d, B.
(1992). A
˚sikter om den Offentliga Sektorn och
Skatterna, 1992. Department of Sociology, Umea
˚
University, Umea
˚, Sweden. http://www.ssd.gu.se/.
Svallfors, S. and Edlund, J. (1997). A
˚sikter om den
Offentliga Sektorn och Skatterna, 1997. Department
of Sociology, Umea
˚University, Umea
˚, Sweden.
http://www.ssd.gu.se/.
Svallfors, S. and Edlund, J. (2002). A
˚sikter om den
Offentliga Sektorn och Skatterna, 2002. Department
of Sociology, Umea
˚University, Umea
˚, Sweden.
http://www.ssd.gu.se/.
Sweden’s Ministry for Foreign Affairs. (2006). Swedish
Aliens Act. Regeringskansliet, Stockholm, Sweden.
http://www.sweden.gov.se/content/1/c6/06/61/22/
fd7b123d.pdf.
Swedish Integration Board. (2006). Pocket Facts—
Statistics on Integration. Norrko
¨ping, Sweden:
Integrationsverket.
Swedish Migration Board. (2007a). Asylum Seekers to
Sweden During 1984–2006. Norrko
¨ping, Sweden.
http://www.migrationsverket.se/pdffiler/statistik/
tabe2.pdf.
Swedish Migration Board. (2007b). Facts and Figures.
Norrko
¨ping, Sweden. http://www.migrationsverket.
se/infomaterial/om_verket/oh/ohstatistik_en.pdf.
Tajfel, H. (1970). Experiments in Intergroup
Discrimination. Scientific American,November,
96–102.
Tajfel, H. (1978). Interindividual Behavior and
Intergroup Behavior. In Tajfel, H. (Ed.),
Differentiation Between Social Groups: Studies in
the Social Psychology of Intergroup Relations.
London: Academic Press, pp. 27–60.
216 EGER
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
Tajfel, H. (1981). Human Groups and Social Categories.
Cambridge: Cambridge University Press.
Tajfel, H. (1982). Social Psychology of Intergroup
Relations. Annual Review of Psychology,33, 1–39.
Tajfel, H., Billig, M. G., Bundy, R. P. and Flament, C.
(1971). Social Categorization and Intergroup
Behavior. European Journal of Social Psychology,1,
149–178.
Tajfel, H. and Billig, M. G. (1974). Familiarity and
Categorization of Intergroup Behavior. Journal of
Experimental Psychology,10, 157–170.
Tajfel, H. and Turner, J. C. (1986). The Social Identity
Theory of Intergroup Behavior. In Worchel, S. and
Austin, W. G. (Eds.), Psychology of Intergroup
Relations. Chicago: Nelson-Hall Publishers,
pp. 7–24.
Tocqueville, A. ([1848]1969). Democracy in America.
New York: Perennial Classics.
Turner, J. C. (1975). Social Comparison and Social
Identity: Some Prospects for Intergroup Behavior.
European Journal of Social Psychology,5, 5–34.
Turner, J. C. (1981). The Experimental Social
Psychology of Intergroup Behavior. In Turner, J.
C. and Giles, H. (Eds.), Intergroup Behavior.
Oxford: Blackwell, pp. 66–101.
Turner, J. C. (1984). Social Identification and
Psychological Group Formation. In Tajfel, H.
(Ed.), The Social Dimensions: European
Developments in Social Psychology. London:
Cambridge University Press.
Author’s Address
Department of Sociology, University of Washington,
Box 353340, Seattle, WA 98195-3340 USA.
Email: meger@u.washington.edu
Manuscript received: May 2008
EFFECT OF IMMIGRATION ON SUPPORT FOR WELFARE STATE SPENDING 217
at Stockholms Universitet on February 19, 2015http://esr.oxfordjournals.org/Downloaded from
- ... As the worst economic crisis since the Great Depression of the 1930s, the Great Recession of 2007-2010 disrupted economic security, social status, and ways of life of millions of citizens [3]. Coupled with high levels of immigration, this made the actual or anticipated life chances of native workers more precarious, causing them to project their fears and anxieties onto immigrants and triggering support for anti-immigrant policies [4] and disapproval of economic redistribution and expansion of the welfare state [5][6][7][8][9]. 1 A vast body of cross-national research has examined the connection between economic decline and anti-immigrant sentiment [1,10]. A major perspective motivating this research is competitive threat theory, which argues that deteriorating economic conditions intensify economic competition with immigrants for scarce resources such as jobs, wages, and welfare benefits, thus contributing to anti-immigrant attitudes [11]. ...... 1 While some research shows that levels of immigration are associated with favorable attitudes toward the welfare state [67], these studies are mostly conducted at the national level. Studies in a variety of national settings at the subnational level that capture nuanced differences among states, counties, or cities generally show a negative effect of immigrant levels on support for the welfare state [5][6][7][8][9]. ...
- ... Until recently Sweden appeared to have the most open labor migration system of all OECD countries, given the absence of skill requirements, salary thresholds and visa restrictions as well as the renewability of permits (OECD, 2011;Emilsson, 2014), despite the fact that in June 2015 Sweden tightened its policy on admission of asylum seekers. Compared with the majority of other countries Sweden is very attractive to both highly skilled and less skilled immigrants, mainly because of its open migration policy, relatively good economic development, low unemployment rate, generous welfare benefits, high living standard, political freedom, friendly environment and the high degree of legal equality it offers to immigrants (Eger, 2010;Koopmans, 2010;Giulietti & Wahba, 2013;OECD, 2014;. ...... This transformation has increased the need for highly skilled workers (Edin and Topel 1997;Cerna 2012), providing job opportunities for skilled immigrants. Second, Sweden stands out as one of the most humanitarian, egalitarian and democratic countries in the world (Eger 2010) and, furthermore, offers generous welfare benefits, devoting around 30 % of its GDP to social expenditures (Eger 2010;Giulietti and Wahba 2013;). The Swedish social welfare system may be an important positive factor in the choice of destination country for immigrants, including the highly skilled ones, who want to minimize the risks of migration. ...... This transformation has increased the need for highly skilled workers (Edin and Topel 1997;Cerna 2012), providing job opportunities for skilled immigrants. Second, Sweden stands out as one of the most humanitarian, egalitarian and democratic countries in the world (Eger 2010) and, furthermore, offers generous welfare benefits, devoting around 30 % of its GDP to social expenditures (Eger 2010;Giulietti and Wahba 2013;). The Swedish social welfare system may be an important positive factor in the choice of destination country for immigrants, including the highly skilled ones, who want to minimize the risks of migration. ...
- ... While Luttmer (2001) uses data on individual support for redistribution in the United States, this section examines the group loyalty effect based on a broad panel of countries. In line with recent research on the topic ( Habyarimana et al., 2007;Fong and Luttmer, 2009;Eger, 2010), our hypothesis is that a higher degree of diversity is negatively related to redistribution. Due to past comparability issues with the redistribution variable, cross-country evidence on the effect of diversity on fiscal policy and the welfare state is surprisingly scarce. ...
- ... It is rather difficult to overlook that a very prominent role in the process of policy-based politicization of immigration in Sweden has historically been played by the Swedish Social Democrats (SAP). SAP immigration policies con- cerned both immigrant entry (Hinnfors, Spehar, & Knapp, 2012), asylum (Abiri, 2000;Eastmond & Ascher, 2011), or immigrants' social and welfare rights in Sweden (Eger, 2010;Sainsbury, 2006;Schierup & A lund, 2011;Schuster, 2000). Despite being a social-democratic party with an "ideo- logical foundation on public sector support, solidarity, inclusiveness and socialism with international roots' (Hinnfors, Spehar, & Knapp 2012, p. 586), the SAP often legitimized its immigration policy with its primary role of a defender of interests of the Swedish national community-as part of its key idea of Swedish Folkhemmet (Andersson, 2009;Pred, 2000). ...
- ... Similarly, studies of radical left voting have suggested different mediation effects-left-wing attitudes reduce the negative effects of socioeconomic position on support for radical left parties ( Ramiro, 2016;Visser et al., 2014)-indicating that those who experience economic difficulties are likely to be in favor of welfare redistribution and are therefore prone to vote for the radical left. And it has long been shown that those with lower incomes and the unemployed are significantly more likely to be in favor of welfare redistribution (see Eger, 2010;Finseraas, 2009;Mau & Burkhardt, 2009). It could also be the case that these voters support radical parties because they are discontented with politics (see Bergh, 2004). ...
- ... While there seems to be a general consensus that immigration and support for redistribution tend to be negatively correlated (Alesina, Glaeser and Sacerdote 2001;Luttmer 2001;Lind 2007;Dahlberg, Edmark and Lundqvist 2012), the specific causal mechanism behind this relationship is far from being settled. Moreover, an emerging body of literature points to a number of demographic, economic, and political factors moderating the relationship between immigration and support for social policy ( Soroka et al. 2004;Crepaz 2006;van Oorschot 2006;Crepaz 2008;Finseraas 2008;Larsen 2008;Crepaz and Damron 2009;Mau and Burkhardt 2009;Eger 2010). This literature, however, has obtained conflicting results claiming both high and low effects of institutions. ...... An emerging strand of literature tries to reconcile these inconsistencies by pointing out to nuanced heterogeneous effects, or to powerful omitted alternative explanations. There is evidence that the link between anti-immigrant sentiment and support for redistribution varies across levels of foreign-born population, multicultural policies, economic attitudes and conditions, political strategy, public policy preferences, and demographic changes among other factors (Fellowes and Rowe 2004;Soroka et al. 2004;Crepaz 2006;Finseraas 2008;Senik, Stichnoth and Van der Straeten 2008;Hopkins 2009;Mau and Burkhardt 2009;Eger 2010;Rugh and Trounstine 2011;Lieberman and McClendon 2013). ...
- ... The increasing labor mobility caused by globalization has made many more societies racially, ethnically, and religiously heterogeneous, including Scandinavian social democracies. Because people are more likely to pay taxes (or make other sacrifices) for other in-group members than for members of other groups (Lieberman 2009), increasing heterogeneity leads to increasing opposition to high taxes to fund large welfare states (Eger 2010, Kymlicka & Banting 2006). The difficulty of maintaining ...
- ... Along similar lines, altruism is salient for particularly the welfare state to function as altruistic behavior is characterized as being helpful without expecting anything in return. Previous studies show that if immigration is perceived as exploiting the welfare state, the support for the welfare state goes down (e.g.Eger, 2009). While the NYSA is not a direct exploitation of the welfare state, the assaults may be perceived as an exploitation of the openess of the receiving society and culture, which includes gender equality. ...... This assumption is backed empirically. In a European context,Eger (2009)shows that immigration has reduced support for the Swedish welfare state, while more recent studies in Germany point in the same direction, showing that native-born Germans become more reluctant to support welfare state redistribution when the proportion of immigrants at the regional level increases. This effect is particularly strong in the initial phase of mass immigration waves, while the effect is being moderated by the economic context, including the unemployment rate (Schmidt-Catran et al., 2016). ...
- ... Intolerance is costly as it contributes to health problems, discrimination, and violence (Bask, 2005;Bunar, 2007;Socialstyrelsen, Folkhälsorapport, 2009;Wamala, Boström & Nyqvist, 2007). Growing immigration to Sweden activates perceptions of threats and opposition to the universal spending (Eger, 2010;Eger & Bohman, 2016;Hjerm & Nagayoshi, 2011). Thus, ensuring tolerance of contemporary and future citizens is of utmost interest to the Swedish Government (Regeringskansliet, 2012). ...
- ... Table 1 provides descriptive statistics on the individual-, regional-, and countrylevel data employed in the analyses. While single-country studies may use variables that capture attitudes toward levels of social expenditure (Eger, 2010 ), cross-national analyses must rely on welfare state attitudes that are comparable across countries (Svallfors, 1997). Statements about what the government ought to do, therefore, are ideal for this type of research (Brady and Finnigan, 2014 ). ...... By the end of 2008, of the roughly 70 million immigrants residing on the European continent, approximately 75 percent immigrated to Western Europe (United Nations, 2010) – the region that boasts the most robust and redistributive welfare states in the world. As foreigners make up an increasing proportion of populations in these democracies, many have asked whether immigration undermines the social solidarity required to maintain popular support for the welfare state (e.g.Banting and Kymlicka, 2006;Eger, 2010). One of the most significant achievements of the world's advanced countries, the welfare state has many functions involving the reduction of economic risk and labor market uncertainty (Barr, 2000;Iversen, 2005). ...... Furthermore, cross-national studies find negative relationships between ethnic diversity and social expenditure (Alesina et al., 2001;Steele, 2016) and immigration and welfare state growth (Soroka et al., 2006). Therefore, it is no surprise that some cite immigration as a potential challenge for Western European welfare states (Alesina and Glaeser, 2004;Eger, 2010;Larsen, 2011). However, a growing body of research on immigration and European welfare attitudes has produced mixed findings, leading some to conclude that this relationship is weak or non-existent (e.g.Brady and Finnigan, 2014;Crepaz, 2008). ...
Project
- Private Profile
- Mikael Hjerm
- Andrea Bohman
Funded by Marianne and Marcus Wallenberg Foundation (MMW) and Sveriges Riksbankens jubileumsfond (RJ). Research team: Mikael Hjerm, Maureen Eger, and Andrea Bohman, Umeå University.
The aim of thi…" [more]
Project
- Private Profile
- Mikael Hjerm
- Andrea Bohman
- [...]
- Jeffrey Mitchell
Funded by the Swedish Research Council for Health, Working Life and Welfare (FORTE).
Social scientists have contributed significantly to our understanding of the origins and nature of prejudice, …" [more]
Article
Full-text available
A growing body of research connects diversity to anti-welfare attitudes and lower levels of social welfare expenditure, yet most evidence comes from analyses of US states or comparisons of the United States to Europe. Comparative analyses of European nation-states, however, yield little evidence that immigration – measured at the country-level – reduces support for national welfare state... [Show full abstract]
Article
Full-text available
Previous cross-national research on the link between immigration and the welfare state has focused exclusively on the relationship between the size of a country’s foreign-born population and support for redistribution, neglecting that people vary in their responses to immigration. In this article, the authors revisit the progressive’s dilemma by testing its theoretical proposition—that... [Show full abstract]