Anti-Immigration and the Media
How media framing affects tolerance towards
anti-immigration parties
Niklas Jakobsson, 10832777
Master’s Thesis, Graduate School of Communication
Master’s Programme Erasmus Mundus Journalism: Media and Politics
Supervisor: Yphtach Lelkes2
Abstract
Anti-immigration parties in Europe are on the rise. With growing support for these parties in the political arena, it’s imperative to research which factors influence the public’s perception of these parties. Media effects have been extensively researched and shown to have significant effects on political processes and opinions. Therefore this study looked at how media effects, and media framing in particular, has an impact at the most basic level of acceptance between groups – political tolerance.
Through and experiment this study looked at how media framing can have an impact on political tolerance in the case of Sweden. The concrete research question was “To what
extent does media framing affect political tolerance towards anti-immigration parties in Sweden?”
The study was carried out via an online experiment distributed to 120 university students across Sweden. Results showed that introducing conflict when talking about the Swedish anti-immigration party increases political tolerance whilst introducing democratic values has no significant effect.
Introduction
Since the 1990s, anti-immigration parties have been on the rise in Europe. While some have
been accepted into their respective national political arenas, others have seen constant
political exclusion despite gaining support in elections. This has led to a widespread
condemnation by the media of anti-immigration parties and politics. But what does this mean for peoples’ perception of these parties?
Political science sees conflict as something which breads tolerance (Duch and Gibson, 1992).
There exists much research in the field of media theory which highlights how conflict mainly
influences tolerance negatively (Nelson et al., 1999; Ramirez and Verkuyten, 2011; Sullivan
and Transue, 1999). This study sets out to explore which of the two fields takes precedence by
using the impact of media framing on tolerance towards the Swedish anti-immigration party,
the Sweden Democrats.
This study will examine how the effects of different frames of the same event (the 2014
Swedish budget crisis) in English- language news has an impact on how tolerant respondents
3
political tolerance, this study used an online experiment to study correlations between media
framing and political tolerance. The experiment was carried out amongst people currently
enrolled at a Swedish Higher Education Institute (Högskola or Universitet).
University students or equivalent were selected due to their general predisposition to be
overall more politically tolerant (Golebiowska, 1995). This study shows that media framing
has an impact on political tolerance towards anti-immigration parties. Introducing conflict
into media content means that university students become more politically tolerant.
Furthermore, democratic framing does not have the same statistically significant effect.
Theoretical Framework
There is a clear trend of excluding anti-immigration parties from policy making and political
debate on a national level (Van Spanje and Van der Brug, 2009). Downs (2002) sees this as
one of several ways to deal with anti-immigration parties. The other possibilities include
ignore, isolate, co-opting their programmes and collaboration. As some of these are mutually
exclusive, Van Spanje and Van der Brug (2009) divide them into political collaboration or
exclusion. But despite these tactics, the success of anti-immigration parties across Europe
does not seem to be slowing down.
Support for anti-immigration parties, and their subsequent success, has been approached from
several different angles (Van der Brug, 2005). The most prominent model, socio-structural,
(Stern and Betz, 1994, Betz, 1998 and Lubbers and Scheepers, 2000) argues that support for
these parties comes from citizens who feel threatened by the rapid changes in society.
Furthermore, there is the reoccurring argument of the protest vote (Stern and Betz, 1994).
This means that voters support these parties due to the deficiencies of mainstream parties
4
While the impact of this has been researched from a political science perspective, it is
appropriate to expand on this research by looking at how media has an impact on how people
perceive these parties through the use of media framing. Framing has been shown to have
significant effects on several aspects of political interaction, such as voting behavior (Hobolt,
2006) and overall support for political issues (de Vreese, van der Brug and Hobolt, 2011).
Therefore it is imperative to look at how media framing has an impact on the more superficial
level of political tolerance, which is seen as the most basic level of political relations between
groups (Vogt, 1997).
Frames are “central organizing ideas that provide coherence to a designated set of idea elements” (Ferree, Gamson, Gerhards & Rucht, 2002, p.105). In essence, a frame sets the framework for what an issue is and is not. It creates a border which distinguishes the content
within the frame from content which is different (Gamson, 2004). Entman (1993) argues that
to frame is to selectively emphasize certain aspects of a perceived reality. This is done in
order to make something more salient in a communicating text, “in such a way as to promote
a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described” (p.52).
Patterson and Iyengar (1991) argue that people attempt to understand political issues by
reducing them to simple questions of responsibility. In terms of media, media framing has been defined by Gamson and Modigliani (1987) as “a central organizing idea or story line that provides meaning to an unfolding strip of events… The frame suggests what the controversy is about, the essence of the issue” (p.147).
Many view framing effects as evidence of citizen incompetence – an ability to use an elitist
position to manipulate the common citizen (Druckman, 2001). However, Druckman’s (2001)
research shows that citizen incompetence is an isolated issue and citizens generally use frames in a “competent and well-reasoned manner” (p.225). The ability to manipulate or alter
5
opinions through framing can therefore have substantial effects on important political issues
such as political support and political tolerance.
Questions of political tolerance are integral for increasingly diverse societies and functioning
democracies (Ramirez and Verkuyten, 2011). Willhoite (1977) sees political intolerance as a natural response to one’s political enemies. However, intolerance is less likely to present itself when other values are highlighted, what Stouffer (1955) calls the sober second thought.
Furthermore, Sullivan, Piereson and Marcus (1978) define tolerance as the willingness to
extend liberties and protections to disliked or even hated groups. Duch and Gibson (1992)
argue that people regularly faced with ideological diversity are likely to have a higher level of
political tolerance.
One of their main indicators for this ideological diversity is the extent of which there is
support for radical right or left-wing parties (Duch and Gibson, 1992). A country which has
gone through several elections with parties on the extreme should therefore see a steady
increase in political tolerance. However, it is important to note that tolerance is not something you show when you are in support of a group or political party. “You do not demonstrate "tolerance" toward groups whose ideas you support or about which you don't care. For
example, if you are sympathetic to the views of a pro-life group, or are neutral toward their
stance, then you should not describe yourself as "tolerant" toward the group” (Avery, 2002,
pp.270).
Tolerance increases greatly towards groups or organizations that support democracy and a
willingness to extend basic rights to other groups, essentially groups “playing by the rules”.
Changing the tolerance towards these groups is therefore difficult. However, people are less
likely to be tolerant towards groups which are seen as anti-democratic and not extending basic
rights. Furthermore, tolerance towards these groups is also more easily influenced (Petersen,
6
In creating a clear definition for “anti-immigration party”, Fennema (1997) argues that there is only a single variable that defines this stance; a political party’s opposition to immigration. Fennema (1997) argues that parties do not need to share fundamental ideologies to be defined
as anti-immigration. Furthermore, it is imperative to note that there is a clear distinction
between being anti-immigrant and anti-immigration in this definition.
Anti-immigration parties from across the political spectrum have established themselves in
several Western European countries over the last few decades. Sweden is one of the latest
countries to see a surge in support for its anti-immigration party (Rydgren and Ruth, 2011).
Fennema (1997) argues that anti-immigration parties can exist across the political spectrum.
That is, as mentioned previously, due to the clear definition that an anti-immigration party
needs to only hold the single outspoken position against immigration to be labelled as
anti-immigration.
Recent research has shown that the anti-immigration Sweden Democrats have a mixed level
of influence on local policies concerning refugees and immigration (Bolin, Lidén and Nyhlén,
2014). However, the current political situation in Sweden began when SD had its first major
political breakthrough in the 2010 election (Hellström, Nilsson and Stoltz, 2012). The Sweden
Democrats managed to win their first seats in the parliament, acquiring enough votes to pass
the 4% bar. This not only gave them a small amount of political influence, but substantially
increased the amount of media coverage surrounding the party and its members. Boomgarden
and Vliegenhart (2007) found that in the Netherlands, an increased amount of media coverage
relating to immigration issues led to a higher share of vote intentions for anti-immigration
parties.
The media’s coverage of anti-immigration parties already has certain impacts on readers and voters. Hellström, Nilsson and Stoltz (2012) argue that the position of the Swedish
7
party to stress their views on “Swedishness and social cohesion” (p.186). Cultural racism has
been central to the emergence and popularity of the Sweden Democrats (Deland and Westin,
2007) which arguably makes their position and coverage in the mainstream media provocative
and opinion-splitting. However, the ideological positioning of the Sweden Democrats has evolved due to the interaction between the party’s self-image and the counter-arguments adopted by its opponents (Hellström and Nilsson, 2010). Hellström and Nilsson (2010) argue
that mainstream parties approach of distancing themselves from the party would tame in the
run-up to the 2010 election, but instead it intensified.
Since SD became a member of the Swedish parliament, only one other political party has
voiced support for a stricter immigration policy1. SD is now the third largest party in the
country and inferring from the research on local influence (Bolin, Lidén and Nyhlén, 2014)
this should give them an increased amount of influence on the refugee and immigration
policies on a national level. However, the December Deal2 made by the leading coalition and
opposing coalition following the 2014 parliamentary budget vote shows that a majority of the
other political parties refuse to let SD influence the country’s policies, arguably indicating a
lack of political tolerance at the elitist level.
Just as the media has an impact on perception of anti-immigration parties, the behavior of
other political parties plays an integral role in their success or failure. Van Spanje and Van der
Brug (2009) have shown that ostracizing anti-immigration parties does not have any negative
effect on their support. Since SD grew by 7% in a four-year span it also shows that being
ostracized by mainstream Swedish political parties did not have any negative impact on their
support, on the contrary, support saw a steady increase. There are other factors, such as the
continuing poor economic situation and the increase in immigration which could also have
1
KD announces policy for stricter immigration rules, last accessed 6 June 2015 http://www.expressen.se/nyheter/kd-vill-sanka-bidragen-for-invandrare/
2
December Deal, last accessed 6 June 2015 http://www.reuters.com/article/2014/12/27/us -sweden-politics-poll-idUSKBN0K505120141227
8
contributed to this increase. Hellström and Nilsson (2010) mistook the possible change of
position by political parties in relation to the Sweden Democrats. However, they did
hypothesize that the policies and approach by SD would be more accepted which arguably is
the case.
Furthermore, Sweden is seen as the most tolerant country towards immigrants in the European Union. “Sweden stands out. Time after time, it shows itself to be the most tolerant country. The differences between Sweden and the other countries are not always large. But they are consistent” (Sniderman et. al, 2014, p.23). So the Swedish public has a tendency of tolerance towards immigrants – but it is questionable to what extent this translates into tolerance
towards those intolerant towards immigration.
The accepting of SD’s political stance infers that there may be an increasing level of political
tolerance towards the party among voters belonging to other political parties. Avery (2002)
defines political tolerance as the willingness to extend basic rights and civil liberties to those
with whom you disagree. However, the possible rise of political tolerance cannot be directly
attributed to the more accepting approach suggested by Hellström and Nilsson (2010). Adding
media influence could potentially explain this change.
Nelson et al., (1997) argue that controversies which bring important values into conflict lead
to unstable opinions. Furthermore, these opinions are affected by how the controversy is
portrayed. Sullivan and Transue (1999) showed that those who endorse democratic values are more likely to adopt a tolerant position. Therefore, people’s value priorities in combination with news frames can be expected to influence political tolerance for extreme political groups.
Keeping in line with the approach that diverse political landscapes breed tolerance (Duch and
Gibson, 1992) should mean that homogenous frames – which do not highlight conflict -
9
should increase the political tolerance. However, research on media framing and its impact on
voting shows that positive framing, which agrees with the reader’s value priorities, should
have a positive impact and negative framing, which disagrees with the reader’s value
priorities, should have a negative impact (Nelson et al., 1997; Sullivan and Transue, 1999) on
tolerance.
The two opposing theoretical arguments need therefore be tested to see whether the impact of
the overarching value system or the media effects takes precedence in this setting.
In their research into tolerance towards extremist organizations in the United States of
America, Nelson, Clawson and Oxley (1997) found that media framing highlighting values
deeply engrained in the population increases tolerance towards extremist organizations. As
the Swedish anti-immigration party came out of the fascist movement3 it can be argued that
they are still given their “extremist” label for historic reasons despite their cohesion with the
political system.
The fostering of democratic values has a long-standing history in the Swedish school system
(Ohrn, 2001). The Swedish democratic values in this regard are the “inviolability of human
life, individual freedom and integrity, the equal value of all people, equality between men and
women and solidarity with the weak and vulnerable” (Ohrn, 2001, p.321). Furthermore, these
democratic values include the dimension of democratic relations such as generosity, tolerance,
understanding respect and consideration (Ohrn, 2001). While the Swedish society does not
have the same deeply rooted civil liberties values as Nelson, Clawson and Oxley (1997) find
have an impact in the US, the democratic values argued by Ohrn (2001) are arguably equally
engrained in the Swedish society as civil liberties are in the US.
3
From Boots to Suits: Sweden Democrats extreme roots, last accessed 6 June 2015 https://euobserver.com/eu-elections/123316
10
Building on the research by Nelson, Clawson and Oxley (1997) it can be argued that highlighting the Swedish equivalent of “deeply rooted civil liberties” (democratic values) would have a similar impact towards political tolerance.
Research Question and Hypotheses
Based on the theoretical framework laid out in the previous section the following research
questions and hypotheses have been created in order to test the extent of which media framing
in the context of Swedish politics has an impact on political tolerance towards the Sweden
Democrats.
RQ: To what extent does media framing affect political tolerance towards anti-immigration parties in Sweden?
Sub-RQ1 – Which of the two frames will have the greatest effect on political tolerance compared to the control group – regardless of whether the effect is positive or negative?
Hypotheses:
H1a – Exposure to the democracy frame will lead to overall greater tolerance towards anti-immigration parties.
H1b - Exposure to the democracy frame will lead to overall less tolerance towards anti-immigration parties.
Political theory leads us to H1b whilst communication theory leads us to H1a. For the purpose
of this study, H1a will be tested.
H2 – Exposure to the conflict frame will lead to an overall lower level of tolerance towards anti-immigration parties.
11
The ostracizing of anti-immigration parties in Europe has been substantially proven (Van
Spanje and Van der Brug, 2009), and therefore it is likely that the political tolerance towards
the Sweden Democrats will be lower in the conflict frame compared to the democracy frame
and control group
H3 – Respondents with high level of political tolerance will still be spread out across the left-to-right political spectrum.
Duch and Gibson (1992) argue that a broad political spectrum and regular political conflict
leads to higher political tolerance. So when measuring general political tolerance, it is
plausible to argue that people with an overall high political tolerance will also be spread out
across the political spectrum and not belong to a certain party or ideology.
Methodology
The study was conducted by distributing an online self-completion experiment to university
students in Sweden. Access to the respondents was provided by a Danish public opinion
company which allowed access to their nationally representative sample. Respondents were
not offered any credit for partaking in the experiment, however they were included in a draw
to win an electronic device. In order to determine whether the respondents fit within the
sample group, a question was asked whether they are currently enrolled in a Swedish Higher Education Institute. Respondents which answered “No” were sent back to the public opinion company and their answers were not recorded.
The participants were not given any incentives to participate in the study. In all, there was a
total of 120 respondents divided across three conditions with a gender division of 53 males
and 67 females. A total of 153 started the survey, but eight dropped out and 25 were screened
12
Political tolerance varies greatly across socio-economic groups. However, there is a clear
correlation between increased political tolerance and a higher level of education
(Golebiowska, 1995). Higher education (university/college equivalent) “leads to individual
value priorities that are conducive to greater openness to political diversity” (p.23). However,
studies in the United States show that students are tolerant to certain groups and not others.
Students with different political leanings show similar traits in terms of political intolerance
towards certain other groups (Brandt et al., 2014). It is unclear to what extent this research
translates onto a Swedish student population. Although there exist contradictory arguments
regarding the overall tolerance level of students, this group was still selected due to their
slight overall tendency to be more politically tolerant than the average population
(Golebiowska, 1995).
The questions and conditions for this experiment were all in English. As the Swedish media is
seen to be biased against the Sweden Democrats, introducing the democracy frame via
Swedish-language media content could create reliability issues. Therefore the conditions were
made up of English- language articles. Reuters was chosen due to its well-known name and
reputation. All articles were dated on the same day with the same byline. Sweden is ranked as
the third most proficient country of non-native English speakers4, making the language barrier
for the questions and the conditions minimal. Students are also seen as the most proficient
English- language speakers in Sweden – above the national average.
Respondents were first asked to complete a series of demographic questions for the control
variables. For the purpose of this experiment, control variables included age (mean 33.38,
standard deviation 10.176), gender (53 male, 67 female), religion (34 protestant, 2 Roman Catholic, 1 Orthodox, 2 Muslim, 45 Atheist, 23 Agnostic, 10 Don’t Know/Rather not Say), self-assessment of social status (18 Lower Class, 90 Middle Class, 4 Upper Class),
4
13
employment status (57 Unemployed, 63 Employed) and marital/relationship status (55 Single,
36 in Domestic Partnership, 25 Married, 2 Separated, 2 Divorced). Furthermore, respondents
were asked whether they voted in the 2014 general election (111 Yes, 5 No, 4 Rather Not
Say). To avoid issues of social desirability, the questions was worded in such a way that
makes it socially acceptable to say that a respondent did not vote in the election (Appendix B,
Q15).
After the control questions, respondents were randomly assigned to one of three conditions.
Condition one consisted of the Democratic Frame, condition two of the Conflict Frame and
condition three was a control group with the original article. Respondents were asked to
carefully read the text provided. The two different frames and the control condition were fully
randomized amongst respondents, with no previous answer triggering a certain version of the
article. Each article consisted of 650-700 words and was based on a Reuters article published
in December 20145 (Appendix B, “Base”, “Dem”, “Con”). The two conditions were altered in
terms of framing, putting emphasis on either conflict or democratic values. In order for the
respondents in the control group to get a treatment, the original article was included.
The democracy frame was modeled after framing experiments carried out surrounding the US
civil liberties debate (Nelson, Clawson and Oxley, 1997) where it was shown that highlighting
values deeply engrained in the culture and education of the population has an impact in terms
of increased tolerance. The conflict frame was modeled around highlighting
“incompatibilities, disagreements or oppositional tensions between individuals, groups and institutions” (Putnam and Shoemaker, 2007, p.167).
Frames were operationalized by introducing words, phrases and information that highlighted
the different conflict or democratic aspects. In terms of highlighting the different aspects, the following two versions were given; 1. Conflict frame: “The anti-immigration Sweden
5
14
Democrat party, which has set out to collapse the ruling coalition;” 2. Democracy Frame: “The Sweden Democrats, which hold the balance of power in parliament.” In the following paragraph the emphasis on conflict continued with the following alterations; 1. Conflict Frame: “Prime Minister Stefan Lofven said last-ditch talks with Alliance leaders to resolve the crisis sparked by the unaligned Sweden Democrats, who want to cut the number of asylum
seekers by 90 percent, had proved fruitless;” 2. Democracy Frame: “Prime Minister Stefan
Lofven said last-ditch talks with Alliance leaders to resolve the crisis sparked by the
democratically different Sweden Democrats, had proved fruitless. ” Alterations made up
approximately 15 per cent of the text with a variety of words and phrases similar to the
examples given. Furthermore, sub-headings were altered into Conflict Frame: “Anti-Immigration” and Democracy Frame: “Reform Key”. This was done in order for the
treatments to still have an impact if readers skimmed through the text and took cues from the
sub-headings.
Following the stimulus material, a series of questions were asked in order to measure their
political tolerance and attitudes towards the Sweden Democrats. Respondents were asked to
evaluate to what extent they see SD as a threat to their own political ideology, to what extent
they believe the party exercised their democratic rights within the context of the article and
who they assign blame to in the same context. The third series of questions set out to measure
political attitudes and tolerance towards all parties in the Swedish parliament at the time of the experiment. Questions included which political party’s political beliefs do they agree or disagree with the most. Respondents were also allowed to give a “Do not know/Rather not say” response to these questions. After the final questions, respondents were led to a de-briefing page where the experiment was explained and respondents were given contact details
15
The key variables are comprised of (Duch and Gibson, 1992) micro-level origins of political
intolerance. Although measuring intolerance, these indicators also, directly, measure political
tolerance. Out of the 10 indicators, 5 were related to demographic attributes such as gender,
age, religion, social status, ideology. The other five indicators were translated into the
questions relating to perception of threat, attitude towards radical social change, attitude
towards protection of basic rights the extent to which they believe that a democracy is the ideal governance system in today’s society and the extent of which they think the Sweden Democrats are politically alienated/has political efficacy.
These questions made up the total Tolerance variable. Each question was measured on an
11-point scale (0-10) with three of the questions having a positive wording and two of the
questions having a negative wording. Negatively worded questions had their answers recoded
to be on the same scale as the positive questions. There were no “Don’t Know/Rather Not Say” options given to these questions. When analyzed, three of the questions together had a Cronbach’s Alpha of .650 while two were seen as measuring something different. Those two combined had a Cronbach’s Alpha of .640. The first set of three questions therefore made up
the political tolerance variable (ToleranceSmall).
Results
In all regression analyses, respondents which indicated their support for the anti-immigration
party were excluded making n=114. Respondents assigned to the different conditions did not
differ significantly with regards to age, gender, region, religion and political leanings (Table
1).
Table 1 Participant Characteristics
CF DF Control
16 Age, Mean 31.93 34.54 33.38 Gender male 21 18 18 female 19 21 23 Religion (a) protestant 12 10 12 atheist 16 18 10 agnostic 9 7 7 PolLeaning, mean 5.31 5.22 5.18
a. Only the major religious groups of respondents accounted for
Framing’s impact on political tolerance
The democracy frame increased the overall tolerance mean of respondents by .264 compared
to the control group (mean 8.096 versus mean 8.360, t(113)=2.281). However, this was not
statistically significant. In H1 it was posited that exposure to the conflict frame would increase the reader’s tolerance level towards the Swedish anti-immigration party.
Conflict framing has a statistically significant positive impact on respondents’ tolerance level
towards the Swedish anti-immigration party at the p<.05 level when comparing it to the
control group. The conflict frame increased the tolerance mean by .818 (M=8.914, SD=.391,
p=.039).
When running the regression analysis, demographic variables were introduced in order to
further assign the effect to the conflict frame. The variables of age, gender, religion,
employment status and marital status had no significant effect on the tolerance level (Table 3).
H2 theorized that exposure to the conflict frame would decrease the level of political
tolerance in the experiment. However, the results show a causation between exposure to a
conflict frame and a statistically significant increase in the level of political tolerance.
In relation to H3 an analysis was run to see whether any specific political ideology had an
impact on higher levels of tolerance. Neither frame showed any interaction effects with
17
Furthermore, political ideology, without the interaction, had no significant effect on the
tolerance variable across all conditions, rendering H3 plausible.
Table 2 Interaction Effects
n=114 B Std.error b Std.error b Std.error
Constant DemocracyIdeology ConflictIdeology Conflict 8.096 .021 - .818** .306 .118 - .399 8.096 - .114 .244 .288 - .115 .678 8.096 - - .818** .278 - - .391 Democracy .216 .706 .264 .409 .264 .396
a. Dependent Variable: ToleranceSmall
Table 3 Regression Model
b Std.error b Std.error b Std.error
n=114 Constant Conflict 8.096 .818** .278 .391 7.046 .935** 1.037 .403 7.450 .925** 1.104 .403 Democracy .264 .396 .346 .412 .368 .412 Gender - - -.302 .334 -.299 .334 Age - - .017 .019 .013 .019 Employment - - .221 .256 .223 .256 Religion - - .090 .063 .097 .064 MaritalStatus - - -.074 .197 -.109 .199 DisagreeSwedenDem - - - - -.392 .368
a. Dependent Variable: ToleranceSmall
However, the abovementioned regression analysis does not take into account respondents’
ideological attitude towards the Sweden Democrats. So the same regression analysis was
computed but adding the variable of which political party the respondent disagrees the most
with. All respondents for Sweden Democrats were coded as 1 and all other parties were coded
as 0. In Table 3 we see that when taking into account respondents’ actual thoughts on the political parties’ ideology, the conflict frame is still the main variable increasing political tolerance.
18
Democracy vs Conflict
Sub-RQ1 was set out in order to see which of the two frames has the largest effect on political
tolerance – regardless of whether the effect is positive (increased tolerance) or negative
(decreased tolerance). In the first column of Table 3 we looked at the change of tolerance
between the three groups regardless of any control variables. Results show that although both
frames have an impact on tolerance, Conflict is still the only statistically significant.
Furthermore, it increases the level of tolerance 3.099 times more than the Democracy frame
(Figure 1).
Discussion
This study set out to find out to what extent media framing has an impact on political
tolerance towards anti-immigration parties. By using the case of Sweden, an experiment was
conducted with two distinct frames.
In terms of the overall research question this study shows that media framing, to a certain
extent, has an impact on political tolerance towards anti-immigration parties. Introducing
7,6 7,8 8 8,2 8,4 8,6 8,8 9
Control Democracy Frame Conflict Frame
Figure 1 - Condition Means
Tolerance Increase Control Tolerance
19
conflict or democratic values into a news article has a positive impact on political tolerance to
a certain degree.
Although results showed that there was an increase in political tolerance for respondents
exposed to the Democracy frame, these results were not statistically significant. H1a was
therefore rendered not plausible in this study but the theoretical foundation and the small
sample could have had an impact on the results. Therefore I suggest that further research
should be carried out, introducing democracy frames to a larger sample.
Another possibility is to what extent the democratic values are still deeply engrained within the respondents. Ohrn’s (2001) research was at the time of the experiment 15 years old and Sweden has since then gone through several recessions and changes in the education system.
Whilst this might not have a huge direct impact, it could explain why highlighting the
democratic values has a lesser effect than when highlighting conflict.
Results show that introducing additional conflict into media content leads to an increase in
political tolerance levels. H2 was built on a variety of media studies research relating to
conflict and its effects on readers. However, results show that conflict had the opposite effect
of what was theorized in H2, following along the lines of political science research (Duch and
Gibson, 1992). Although this was taken into consideration throughout the process of the
experiment, it was not believed that it would be this prominent in the results.
Whilst Duch and Gibson (1992) argue that societies exposed to political conflict over a long
period are more politically tolerant, this study argued that since the Sweden Democrats has
only spent five years in parliament at the time of the experiment, this should not be defined as
a long period of time. However, Sweden saw far-right movements and unsuccessful bids from
20
argued that these early dealings with extremist parties might have had a larger impact on peoples’ perception on conflict.
What the results do show in relation to H2, and the debate of dominating theories from
different fields, is that political science theory on conflict and political tolerance seems to play
a dominant role in shaping political tolerance. Although H2 is rendered not plausible, the
results are positively surprising from a research point-of-view. It is clear that the media is
distancing itself from the Sweden Democrats (Hellström, Nilsson, Stoltz, 2012) and that
anti-immigration parties, in general across Europe, are ostracized. However, these results indicate
that the approach taken by other parties and mainstream media might have a reverse effect.
In terms of H3 it was found that high political tolerance was spread out across the political
spectrum. Respondents not only identified their political leanings, but also identified which
party they supported at the time of the experiment. This falls in line with the expected results
based on previous research and theoretical framework (Duch and Gibson, 1992).
In terms of Sub-RQ1, it was created to look at which of the two frames would have the most
significant effect. As previously discussed, both frames had effects but only conflict was
statistically significant. In terms of the relationship between the two frames, the Conflict
frame increased tolerance 3.099 times more than the Democracy frame (Figure 1). Although it
was originally theorized that the two frames would have different effects (positive and
negative), these results still shed light on the difference in impact the two frames have.
It is also interesting when interpreting these results in relation to Swede’s general tendency to be tolerant. The public’s tolerance towards immigration has been well documented
(Sniderman et. al, 2014) and it seems as though this general level of tolerance translates onto
those intolerant. This also gives insight into how the Sweden Democrats are perceived in terms of “playing by the rules” (Petersen, Slothuus, Stubager & Togeby, 2011). It could be
21
possible that highlighting conflict in media content shows the democratic process in action.
However, results could possibly be different if the same study was carried out with an
organisation which is not political.
Media framing has an impact on political tolerance towards the Swedish anti-immigration
party. However, there is a need to further build on this research from a cross-national
perspective and see to what extent this research could be relevant for other countries.
Countries which are culturally and geographically similar, and have gone through a similar
political transition, could possibly draw conclusions from this research. Denmark, Norway
and Finland all fit within the Nordic frame, with a similar cultural heritage, historic
background and current political situation. A replication of this experiment in those countries
would most likely lead to similar results – also taken into account the fact that these countries
are generally seen as very tolerant as well.
It is also important to note the limitations of this study. The Swedish higher education system
has, to a larger degree than most other countries, a sizeable amount of people aged 25 and
above in the system. This also has an impact on the sample and the demographics of the
respondents. In a country with an overall younger respondent pool, the results could differ
substantially. Furthermore, although the language issues are minimal, respondents could miss
out on the nuances of certain aspects of the frames. The limited amount of respondents means
that there could be aspects and variables which did not have enough respondents to have a
statistically significant impact. Furthermore, the effects of both frames could potentially be
more prominent if the sample was increased. The combination of questions which made up
the ToleranceSmall variable were only three out of the five which were theorized as indicators
of political tolerance (Duch and Gibson, 1992). There would also be a need in a replication of this experiment to revisit the two variables which were excluded due to a poor Cronbach’s
22
Alpha and see whether better question wording would make them statistically relevant in a
future analysis.
Conclusion
This paper gives an indication of how the current media landscape’s dealing with anti-immigration parties can have a significant impact on how these parties are perceived.
Although there are other factors which, in combination with media effects, influences the publics’ perception of these parties, framing effects should not be disregarded. This study has further given insight into the dynamic and difficult relationship between the media, public and
extremist political parties.
References
A. Gamson, W. (2004). Bystanders, Public Opinion and the Media. In: D. Snow, S. Soule and H. Kriesi, ed., The Blackwell Companion to Social Movements, 1st ed. John Wiley & Sons, pp.242-261.
Avery, P. (2002). Avery, P. G. (2002). Teaching Tolerance: What Research Tells Us.(Research and Practice). ,. Social Education, 66(5), p.270.
Betz, H. (1998). Introduction. In: H. Betz and S. Immerfall, ed., The new politics of the right:
Neo-populist parties and movements in established democracies, 1st ed. Basingstoke, UK:
MacMillan, pp.1-10.
Bolin, N., Lidén, G. and Nyhlén, J. (2014). Do Anti-immigration Parties Matter? The Case of the Sweden Democrats and Local Refugee Policy. Scandinavian Political Studies, 37(3), pp.323-343.
Boomgaarden, H. and Vliegenthart, R. (2007). Explaining the rise of anti-immigrant parties: The role of news media content. Electoral Studies, 26(2), pp.404-417.
Brandt, M., Reyna, C., Chambers, J., Crawford, J. and Wetherell, G. (2014). The Ideological-Conflict Hypothesis: Intolerance Among Both Liberals and Conservatives. Current Directions
in Psychological Science, 23(1), pp.27-34.
de Vreese, C., van der Brug, W. and Hobolt, S. (2011). Turkey in the EU?: How cultural and economic frames affect support for Turkish accession. Comp Eur Polit, 10(2), pp.218-235.
Downs, W. (2002). How effective is the cordon sanitaire? Lessons from efforts to contain the far right in Belgium, France, Denmark and Norway. Journal für Konflikt und
23
Duch, R. and Gibson, J. (1992). "Putting Up With" Fascists in Western Europe: A
Comparative, Cross-Level Analysis of Political Tolerance. The Western Political Quarterly, 45(1), p.237.
Entman, R. (1993). Framing: Toward Clarification of a Fractured Paradigm. J
Communication, 43(4), pp.51-58.
Fennema, M. (1997). Some Conceptual Issues and Problems in the Comparison of Anti-Immigrant Parties in Western Europe. Party Politics, 3(4), pp.473-492.
Gamson, W. and Modigliani, A. (1987). The changing culture of affirmative action. In: R. Braungart and M. Braungart, ed., Research in political sociology, 1st ed. Greenwich, CT: JAI Press, pp.Vol. 3 137-177.
Gamson, W., Gerhards, J. and Rucht, D. (2002). Shaping Abortion Discourse. Cambridge University Press.
Gibson, J. (1992). Alternative Measures of Political Tolerance: Must Tolerance be "Least-Liked"?. American Journal of Political Science, 36(2), p.560.
Golebiowska, E. (1995). Individual value priorities, education, and political tolerance. Polit
Behav, 17(1), pp.23-48.
Hellstrom, A. and Nilsson, T. (2010). 'We Are the Good Guys': Ideological positioning of the nationalist party Sverigedemokraterna in contemporary Swedish politics. Ethnicities, 10(1), pp.55-76.
Hellström, A., Nilsson, T. and Stoltz, P. (2012). Nationalism vs. Nationalism: The Challenge of the Sweden Democrats in the Swedish Public Debate. Government and Opposition, 47(02), pp.186-205.
Hobolt, S. (2006). How Parties Affect Vote Choice in European Integration Referendums.
Party Politics, 12(5), pp.623-647.
Lubbers, M. and Scheepers, P. (2000). Individual and contextual characteristics of the German extreme Right-Wing vote in the 1990s. A test of complementary theories. Eur J
Political Res, 38(1), pp.63-94.
Nelson, T., Clawson, R. and Oxley, Z. (1997). Media Framing of a Civil Liberties Conflict and Its Effect on Tolerance. The American Political Science Review, 91(3), p.567.
Ohrn, E. (2001). Marginalization of democratic values: a gendered practice of schooling?.
International Journal of Inclusive Education, 5(2-3), pp.319-328.
Patterson, T. and Iyengar, S. (1992). Is Anyone Responsible? How Television Frames Political Issues. The American Political Science Review, 86(4), p.1060.
Petersen, M., Slothuus, R., Stubager, R. and Togeby, L. (2010). Freedom for All? The Strength and Limits of Political Tolerance. Brit. J. Polit. Sci., 41(03), pp.581-597.
24
Putnam, L. and Shoemaker, M. (2007). Changes in conflict framing in the news coverage of an environmental conflict. J.Disp Resol, p.167.
RAMÍREZ, C. and VERKUYTEN, M. (2011). Values, Media Framing, and Political
Tolerance for Extremist Groups. Journal of Applied Social Psychology, 41(7), pp.1583-1602.
Rydgren, J. and Ruth, P. (2011). Voting for the Radical Right in Swedish Municipalities: Social Marginality and Ethnic Competition?. Scandinavian Political Studies, 34(3), pp.202-225.
Scheufele, D. (1999). Framing as a theory of media effects. Journal of Communication, 49(1), pp.103-122.
Sniderman, P., Petersen, M., Slothuus, R. and Stubager, R. (2014). Paradoxes of Liberal
Democracy. Princeton: Princeton University Press.
Stern, F. and Betz, H. (1994). Radical Right-Wing Populism in Western Europe. Foreign
Affairs, 73(6), p.172.
Stouffer, S. (1955). Communism, conformity, and civil liberties. Garden City, N.Y.: Doubleday.
Sullivan, J. and Transue, J. (1999). THE PSYCHOLOGICAL UNDERPINNINGS OF DEMOCRACY: A Selective Review of Research on Political Tolerance, Interpersonal Trust, and Social Capital. Annu. Rev. Psychol., 50(1), pp.625-650.
Sullivan, J., Piereson, J. and Marcus, G. (1978). Ideological Constraint in the Mass Public: A Methodological Critique and Some New Findings. American Journal of Political Science, 22(2), p.233.
Van Der Brug, W. (2005). Why Some Anti-Immigrant Parties Fail and Others Succeed: A Two-Step Model of Aggregate Electoral Support. Comparative Political Studies, 38(5), pp.537-573.
Van Spanje, J. and Van der Brug, W. (2009). Being intolerant of the intolerant. The exclusio n of Western European anti-immigration parties and its consequences for party choice. Acta
Politica, 44(4), pp.353-384.
Vogt, W. (1997). Tolerance & education: learning to live with diversity and difference.
Choice Reviews Online, 35(04), pp.35-2257-35-2257.
Willhoite, F. (1977). Evolution and Collective Intolerance. The Journal of Politics, 39(03), p.667.
25
Appendix A: Output
Linear Regression ModelsModel Summary
Model R R Square Adjusted R Square
Std. Error of the Estimate 1 ,302a ,091 ,029 1,72600 a. Predictors: (Constant), MaritalStatus, Employment, Democracy, Gender, Religion, Age, Conflict
26 ANOVAa
Model Sum of Squares df Mean Square F Sig. 1
Regression 30,708 7 4,387 1,473 ,185b Residual 306,844 103 2,979
Total 337,552 110 a. Dependent Variable: ToleranceSmall
b. Predictors: (Constant), MaritalStatus, Employment, Democracy, Gender, Religion, Age, Conflict
Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) 7,046 1,038 6,791 ,000 Conflict ,935 ,403 ,256 2,318 ,022 Democracy ,346 ,412 ,093 ,840 ,403 Gender -,302 ,334 -,087 -,906 ,367 Age ,017 ,019 ,097 ,908 ,366 Employment ,221 ,256 ,087 ,865 ,389 Religion ,090 ,063 ,148 1,423 ,158 MaritalStatus -,074 ,197 -,039 -,379 ,706 a. Dependent Variable: ToleranceSmall
Model Summary
Model R R Square Adjusted R Square
Std. Error of the Estimate 1 ,318a ,101 ,030 1,72488 a. Predictors: (Constant), MaritalStatus, Employment, Democracy, Gender, DisagreeSwedenDem, Religion, Age, Conflict
ANOVAa
27
1
Regression 34,082 8 4,260 1,432 ,192b Residual 303,470 102 2,975
Total 337,552 110 a. Dependent Variable: ToleranceSmall
b. Predictors: (Constant), MaritalStatus, Employment, Democracy, Gender, DisagreeSwedenDem, Religion, Age, Conflict
Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) 7,450 1,104 6,748 ,000 Conflict ,925 ,403 ,253 2,293 ,024 Democracy ,368 ,412 ,099 ,894 ,374 Gender -,299 ,334 -,086 -,896 ,372 Age ,013 ,019 ,074 ,678 ,499 Employment ,223 ,256 ,087 ,874 ,384 Religion ,097 ,064 ,159 1,524 ,131 DisagreeSwedenDem -,392 ,368 -,107 -1,065 ,289 MaritalStatus -,109 ,199 -,056 -,546 ,586 a. Dependent Variable: ToleranceSmall
Model Summary
Model R R Square Adjusted R Square
Std. Error of the Estimate 1 ,199a ,039 ,022 1,71667 a. Predictors: (Constant), Democracy, Conflict
ANOVAa
Model Sum of Squares df Mean Square F Sig. 1 Regression 13,445 2 6,723 2,281 ,107b
28
Residual 327,112 111 2,947 Total 340,558 113
a. Dependent Variable: ToleranceSmall b. Predictors: (Constant), Democracy, Conflict
Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) 8,096 ,278 29,074 ,000 Conflict ,818 ,391 ,225 2,091 ,039 Democracy ,264 ,396 ,071 ,666 ,507 a. Dependent Variable: ToleranceSmall
Interaction Models
ANOVAa
Model Sum of Squares df Mean Square F Sig. 1
Regression 15,286 3 5,095 1,674 ,177b Residual 319,686 105 3,045
Total 334,972 108 a. Dependent Variable: ToleranceSmall
b. Predictors: (Constant), ConflictIdeology, Democracy, Conflict
Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) 8,096 ,283 28,604 ,000 Conflict ,244 ,842 ,065 ,290 ,772 Democracy ,264 ,403 ,071 ,655 ,514 ConflictIdeology ,114 ,139 ,177 ,818 ,415 a. Dependent Variable: ToleranceSmall
Bootstrap for Coefficients
Model B Bootstrapa
29 Lower Upper 1 (Constant) 8,096 -,003 ,288 ,001 7,528 8,647 Conflict ,244 -,009 ,678 ,717 -1,148 1,574 Democracy ,264 ,009 ,409 ,522 -,555 1,104 ConflictIdeology ,114 ,001 ,115 ,288 -,147 ,335 a. Unless otherwise noted, bootstrap results are based on 1000 stratified bootstrap samples
ANOVAa
Model Sum of Squares df Mean Square F Sig. 1
Regression 13,154 3 4,385 1,565 ,202b Residual 302,588 108 2,802
Total 315,742 111 a. Dependent Variable: ToleranceSmall
b. Predictors: (Constant), DemocracyIdeology, Conflict, Democracy
Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig. B Std. Error Beta 1 (Constant) 8,096 ,272 29,818 ,000 Conflict ,818 ,382 ,232 2,144 ,034 Democracy ,216 ,773 ,060 ,279 ,781 DemocracyIdeology ,021 ,129 ,033 ,160 ,873 a. Dependent Variable: ToleranceSmall
Bootstrap for Coefficients
Model B Bootstrapa
Bias Std. Error Sig. (2-tailed) 95% Confidence Interval Lower Upper 1 (Constant) 8,096 -,008 ,306 ,001 7,441 8,684 Conflict ,818 ,008 ,399 ,045 ,027 1,596 Democracy ,216 ,009 ,706 ,760 -1,146 1,595 DemocracyIdeology ,021 -,001 ,118 ,847 -,219 ,246
30
a. Unless otherwise noted, bootstrap results are based on 1000 stratified bootstrap samples
Demographics
Religion
Frequency Percent Valid Percent Cumulative Percent Valid Protestant 34 28,1 29,1 29,1 Roman Catholic 2 1,7 1,7 30,8 Orthodox 1 ,8 ,9 31,6 Muslim 2 1,7 1,7 33,3 Atheist 45 37,2 38,5 71,8 Agnostic 23 19,0 19,7 91,5 10 10 8,3 8,5 100,0 Total 117 96,7 100,0 Missing
Rather not say 3 2,5 System 1 ,8 Total 4 3,3 Total 121 100,0
Employment
Frequency Percent Valid Percent Cumulative Percent Valid Full Time 14 11,6 11,7 11,7 Part Time 49 40,5 40,8 52,5 Not employed 57 47,1 47,5 100,0 Total 120 99,2 100,0 Missing System 1 ,8 Total 121 100,0 MaritalStatus
Frequency Percent Valid Percent Cumulative Percent Valid Single 55 45,5 45,8 45,8 In a Domestic Partnership 36 29,8 30,0 75,8 Married 25 20,7 20,8 96,7 Separated 2 1,7 1,7 98,3
31 Divorced 2 1,7 1,7 100,0 Total 120 99,2 100,0 Missing System 1 ,8 Total 121 100,0 Voting
Frequency Percent Valid Percent Cumulative Percent Valid No 5 4,1 4,3 4,3 Yes 111 91,7 95,7 100,0 Total 116 95,9 100,0 Missing 99 4 3,3 System 1 ,8 Total 5 4,1 Total 121 100,0 Q25
Frequency Percent Valid Percent Cumulative Percent Valid Vänster 17 14,0 14,8 14,8 Socialdem 2 1,7 1,7 16,5 Miljö 11 9,1 9,6 26,1 Centerp 2 1,7 1,7 27,8 Kristd 1 ,8 ,9 28,7 Moderaterna 7 5,8 6,1 34,8 Sverigedemokraterna 75 62,0 65,2 100,0 Total 115 95,0 100,0 Missing 99 5 4,1 System 1 ,8 Total 6 5,0 Total 121 100,0 Q26
Frequency Percent Valid Percent Cumulative Percent Valid
Vänster 21 17,4 17,5 17,5 Socialdem 13 10,7 10,8 28,3 Miljö 29 24,0 24,2 52,5
32 Center 12 9,9 10,0 62,5 Kristd 6 5,0 5,0 67,5 Folkp 5 4,1 4,2 71,7 Moderaterna 11 9,1 9,2 80,8 Sverigedemokraterna 6 5,0 5,0 85,8 99 17 14,0 14,2 100,0 Total 120 99,2 100,0 Missing System 1 ,8 Total 121 100,0
Appendix B: Survey and Treatments
The effect of English language news on Swedish University students
Q30 Are you currently a student at a Swedish higher education institution (Högskola eller Universitet) Yes
No
Q27 The effect of English language news on Swedish readers. Thank you for participating in our survey. This study is conducted by students from the University of Amsterdam in The Netherlands. In
33 our survey we want to ask you some questions about media and politics. All the information we collect in this study is confidential, and there will be no way to trace back responses to individual respondents. We only look at the general patterns that emerge among larger groups of people. If you agree to participate in this survey please click on the “Next” button below. Thank you for participating in our survey!
Q3 What is your gender? Male
Female
Q1 What is your age?
Q4 What is your religious preference? Protestant
Roman Catholic Jewish
an Orthodox church such as the Greek or Russian Orthodox Church Muslim
Atheist Agnostic Rather Not Say
Something else (please specify) ____________________ Q5 Are you now employed full-time, part-time or not employed Full time
Part time Not employed
Q6 What is your marital status? Single/never been married In a domestic partnership Married
Separated Divorced Widowed
Q24 When thinking about your current situation, which one of the following social classes fits your life best at this point in time?
34 Lower Class
Middle Class Upper Class Rather not say
Q15 With a busy modern life, it is common that people don't have the time to vote in general elections due to other important commitments. Did you vote in the 2014 Swedish general election? Rather not say
No Yes
Not applicable
Q7 How would you describe your political leanings on a scale of left to right? Extreme Left 2 3 4 5 Centre 7 8 9 10 Extreme Right
35 Q20 Next will be presented with a news report from Reuters, published in December 2014. Please read the text carefully.
Base (Reuters) - Sweden's minority, center-left government teetered on the brink of collapse on Tuesday after just two months in office when a far-right party announced it would vote against the 2015 budget, effectively dooming it to defeat. The anti-immigration Sweden Democrat party, which holds the balance of power in parliament, said it would support an alternative budget proposed by the center-right Alliance opposition bloc, leaving the government isolated. Prime Minister Stefan Lofven said last-ditch talks with Alliance leaders to resolve the crisis sparked by the unaligned Sweden Democrats, who want to cut the number of asylum seekers by 90 percent, had proved fruitless. "There is no one on the other side of the table, it is meaningless to hold talks," Lofven told reporters after the meeting at the government headquarters, saying he would decide how to proceed after Wednesday's debate in parliament. "We may call snap elections later, when the constitution allows. We could also resign and there are other alternatives." The leaders of the Moderate, Centre, Christian Democrat and Liberal parties, which make up the Alliance, stood their ground despite the olive branch offered by Lofven, saying they would not budge from their pledge to vote for their own bill. Parliament is due to vote on the budget on Wednesday. Lofven, head of a coalition of Social Democrats and Greens, faces the risk of becoming Sweden's shortest-serving prime minister since the 1930s, having ruled out staying in government and implementing an opposition budget. With options rapidly running out, he could still send the budget back to committee for amendments to try to win backing from the center-right, though prospects of success are remote. He could also resign and try to put together a new
government. A last resort would be to call a snap election - something that has not happened since 1958 - risking a period of political and market uncertainty. A vote could be called in late December at the earliest, with the elections taking place within the following three
months. "Sweden hasn't been this close to a snap election for many decades," said Andreas Johansson Heino, political scientist at liberal think tank Timbro. The Swedish crown weakened after the Sweden Democrats' decision, losing around 3 ore to stand at around 9.32 to the euro at 5.50 p.m. ET. "In the long term, the most important thing is that we have stable government finances," Annika Winsth, chief economist at banking group Nordea, said. "Investors are going to continue to have confidence that that is the case unless this drags on for a long time." IMMIGRATION KEY The budget was meant to draw a line under eight years of tax cuts under previous center-right administrations that made many Swedes richer, but also raised worries over declining standards in healthcare and education and over increasing social division. The Social Democrats and Greens planned extra spending on schools, welfare and job creation financed by tax hikes, including for high-income earners. The Alliance is proposing more cautious spending as well as vehicle tax hikes and higher duties on tobacco and alcohol. Acting Sweden Democrat leader Mattias Karlsson said his party was flexing its muscles to force a reversal of Sweden's generous stance on immigration. "If the Alliance doesn't change its policies (on immigration), we would try to bring down a government of those parties too," he said. Costs for asylum seekers including housing, language lessons and welfare allowances totaled 1.5 percent of the country's 2013 budget, with Sweden the biggest per-capita recipient of asylum seekers and refugees last year, according to the Organisation for Economic Co-operation and Development. That humanitarian generosity has fed the rise of the far-right. The Sweden
Democrats doubled their support in the September election, taking 13 percent of the vote and becoming the third largest party in parliament. Mainstream parties have shunned them, and the budget gives them a rare opportunity to show their political strength. "They have a mandate
36 from their constituents to stir up trouble," said Henrik Ekengren Oscarsson, professor in political science at Gothenburg University.
Con (Reuters) - Sweden's minority, center-left government teetered on the brink of collapse on Tuesday after just two months in office when a far-right party announced it would vote against the 2015 budget, effectively dooming it to defeat. The anti-immigration Sweden Democrat party, which has set out to collapse the ruling coalition, said it would support an alternative budget proposed by the center-right Alliance opposition bloc, leaving the government isolated. Prime Minister Stefan Lofven said last-ditch talks with Alliance leaders to resolve the crisis sparked by the unaligned Sweden Democrats, who want to cut the number of asylum seekers by 90 percent, had proved fruitless. "There is no one on the other side of the table, it is meaningless to hold talks," Lofven told reporters after the meeting at the government headquarters, saying he would decide how to proceed after Wednesday's debate in parliament. "We may call snap elections later, when the constitution allows. We could also resign and there are other alternatives." The leaders of the Moderate, Centre, Christian Democrat and Liberal parties, which make up the Alliance, stood their ground despite the olive branch offered by Lofven, saying they would not budge from their pledge to vote for their own bill. Parliament is due to vote on the budget on Wednesday. Lofven, head of a coalition of Social Democrats and Greens, faces the risk of becoming Sweden's shortest-serving prime minister since the 1930s, having ruled out staying in government and implementing an opposition budget, an expected outcome due to the Sweden Democrats political sabotage. With options rapidly running out, he could still send the budget back to committee for amendments to try to win backing from the center-right, though prospects of success are remote. He could also resign and try to put together a new government. A last resort would be to call a snap election - something that has not happened since 1958 - risking a period of political and market uncertainty. A vote could be called in late December at the earliest, with the elections taking place within the following three months. "Sweden hasn't been this close to a snap election for many decades," said Andreas Johansson Heino, political scientist at liberal think tank Timbro. The Swedish crown weakened after the Sweden Democrats' shocking decision, losing around 3 ore to stand at around 9.32 to the euro at 5.50 p.m.
ET. "In the long term, the most important thing is that we have stable government
finances," Annika Winsth, chief economist at banking group Nordea, said. "Investors are going to continue to have confidence that that is the case unless this drags on for a long
time." ANTI-IMMIGRATION POLICY The budget was meant to draw a line under eight years of tax cuts under previous center-right administrations that made many Swedes richer, but also raised worries over declining standards in healthcare and education and over increasing social division. The Social Democrats and Greens planned extra spending on schools, welfare and job creation financed by tax hikes, including for high-income earners. The proposed plan will now see itself undermined by the carelessness of the Sweden Democrats. The Alliance is proposing more cautious spending as well as vehicle tax hikes and higher duties on tobacco and alcohol. Acting Sweden Democrat leader Mattias Karlsson said his party was flexing its muscles to force a reversal of Sweden's generous stance on immigration. "If the Alliance doesn't change its policies (on immigration), we would try to bring down a government of those parties too," he said. Costs for asylum seekers including housing, language lessons and welfare allowances totaled 1.5 percent of the country's 2013 budget, with Sweden the biggest per-capita recipient of asylum seekers and refugees last year, according to the Organisation for Economic Co-operation and Development. That humanitarian generosity has fed the rise of the far-right. The Sweden