Anti-Immigrant Parties, Local Presence and Electoral Success

Abstract

Does the local organisational presence of anti-immigrant parties affect their chances for electoral success? In order to answer this question, the article explores the potential of a supply-oriented explanation to anti-immigrant party success by examining the electoral advancements the Sweden Democrats (SD) made in the 2006 and 2010 elections. Our results indicate that traditional demand-side explanations to anti-immigrant party success can be successfully complemented by an ‘internal supply-side argument’ to make the electoral fates of these parties more intelligible. Whether the SD had a local organisational presence had a substantial effect on its results in the national election and on the probability of gaining representation in local councils. Thus, the party’s fate in the national as well as local elections was largely determined by whether or not it had a local organisational presence in Swedish municipalities.

Key Words:

• Anti-immigrant parties
• elections
• Sweden
• ‘internal supply-side’ approach

Notes

 1. Many researchers have dubbed these parties as ‘radical right wing’. However, we agree with Fennema (1997) and Van der Brug and Fennema (2007) that such parties are more correctly categorised as ‘anti-immigrant’. They cannot be easily fitted into the left–right divide. Their only common denominator is campaigning on anti-immigrant issues.

 2. New Democracy was undoubtedly a Swedish populist, anti-immigrant and right-wing party. However, they never succeeded in establishing themselves on the political scene, and organisationally, it would be hard to argue that they constituted a viable political party organisation for more than one electoral period (1991–1994).

 3. Even before their modest success in 2002, the SD was called the ‘most important political party outside the national parliament’ (Larsson and Ekman 2001). It was formed in 1988 and had intimate affiliations to outright racist and Nazi movements in Sweden. In fact, Larssons and Ekmans’ (2001) mapping of the party’s origins leaves no doubt that those who founded the party in the late 1980s were deeply tapped into neo-Nazi networks (for example the notorious Bevara Sverige Svenskt [Keep Sweden Swedish] movement, Sverigepartiet [The Sweden Party] and Framstegspartiet [The Swedish Progress Party]).

 4. Rydgren and Ruth (2011) conclude that the varying fates for Sweden Democrats in local elections are explained by the ‘the social marginality hypothesis’: support is negatively correlated with the average level of education and with gross domestic product (GDP) per capita, and positively correlated with the unemployment rate. They also find support for the ‘ethnic competition hypothesis’ and find a positive correlation between the proportion of immigrants and support for SD. However, they do not mention, hence do not test, any version of the internal supply-side approach.

 5. Another important institutional feature highlighted by Arzheimer and Carter (2006, p. 423) is the territorial organisation of states – that is, the degree of centralisation – which they argue can explain the long-term fate of anti-immigrant parties. More centralised political systems seem to decrease the likelihood of their success. This variable, however, is not of interest in this study since it is the fortune of one anti-immigrant party within the same political system which has not changed its local–central relations, that is under scrutiny.

 6. One of the reasons why we refrain from gauging the effects of, for example, party competition, is that there is lack of variation concerning such political opportunity structures between municipalities (Bäck 2008).

 7. Inglehart has, to our knowledge, never explicitly analysed party emergence or new party success, although a hypothesis along this line of reasoning is suggested in Inglehart and Flanagan (1987, p. 1300f).

 8. Already in 2002 Rydgren implied precisely this, and wrote: ‘an upcoming protest party […], with sufficient appeal, might very well change this situation’, that is Sweden’s status as a country without influential anti-immigrant parties (Rydgren 2002, p. 46). Obviously, in retrospect, Rydgren seems to have been proven right. The mechanisms bringing this change about, nevertheless, are invisible through the lens of a demand-oriented perspective, and need to be complemented by a supply-oriented one.

 9. The reason for excluding Nordic immigrants is that these are, as Lindberg (2011) shows, viewed as equals to Swedish citizens by the SD.

10. Descriptive statistics for the variables are found in Table A1 in the appendix.

11. Since all municipalities are included, one could argue that the significance levels are not relevant, but since we are discussing a general mechanism, that is, the importance of local presence, we consider the significance level to matter, also as an indication of robustness. The associated odds ratios for the variable ‘the SD running with a formal ballot’ in the full models are 18.9 for both years.

12. To check if there is a problem with multicollinearity, we have estimated Variance Inflation Factors (VIF). The VIF ranges between 1.1 and 2.2 for the different independent variables in the two models. There is no consensus for when the VIF is too high. However, a common rule of thumb is that if VIF is below five (some have proposed 10), no serious problem with multicollinearity exists. Our estimated VIFs are well below that threshold.