ABSTRACT

These are political parties, who play the key role in candidate selection and assignment to constituencies; voters, with their potential biases towards candidates; and candidates themselves, who vary in terms of personal qualities, policy platforms and strategies. Kittilson and Tate (2005) identify three models for explaining cross-national differences in ethnic minority representation. The first focuses on social change (Katz and Mair 1994), the second on the role of partisan elites (Aldrich 1995), while their own model focuses on the political opportunity structure, including intra-party organization and the role of party gatekeepers, and changes in the legal environment and in the ideological climate regarding minorities. Changes in social attitudes, they argue, are not in themselves sufficient to promote ethnic minority representation, nor is it enough that the partisan elite come to focus on increasing minority representation. Rather, improvements result from a combination of societal change, organized pressure from minorities on the parties, and therefore change among the party elite. While we agree with the importance of this combination of factors, we argue that the key facilitating variable underlying these conditions is the actual or imputed attitudes of the voters. An important question from this perspective is whether voters themselves demonstrate prejudice against ethnic minority candidates and in favour of those with a more traditional social profile, or whether the party gatekeepers assume such prejudice and thereby justify their candidate selection criteria.5 Our main hypothesis is that the parties, whether intentionally or not, misperceive voters’ preferences vis-à-vis ethnic representatives. Where parties assume that the electorate is not prepared to choose such a candidate, they can thereby justify their status quo approach on this matter. Traditional opinion polls among French voters have produced somewhat mixed evidence on this matter. On the one hand, it is clear that the public perceives a profound lack of responsiveness among the traditional political class. Consistently since the mid-1990s, polls have shown that between 70 and 80 per cent of respondents feel that politicians care little or not at all about what ordinary people think (Bréchon 2004). This distrust of the political class is strongly related to the perception that politicians have been ineffective in addressing enduring unemployment, and to the sense of depolarization between the traditional Left and Right (Martin 2000). Rising distrust is also related to the emergence of post-materialist attitudes among citizens (Norris 1999). Finally, it may be that distrust is also linked to a broad critique of the lack of social diversity among the political elite. For example, a poll in October 2005 showed that 89 per cent of respondents thought there should be more women in Parliament, and 84 per cent thought there should be better representation of youth. While there was less consensus on the issue, a small majority (55 per cent) also felt that there should be better representation of visible minorities in Parliament.6 On the other hand, it is not clear whether French voters would respond positively to minority candidates. A 2007 poll by TNS-SOFRES showed that 30 per cent would refuse to vote for a Muslim candidate, whereas only 18 per cent would refuse to vote for a gay candidate, or for a candidate of foreign origin. Such a penalty could certainly cost the election for a major party. This is

especially so in the context of France’s electoral system (two-round majoritarian, based on single-member-districts), in which small differences in vote margins between the Left and Right can produce significant parliamentary seats differentials (Dolez and Laurent 2005). While 85 per cent of respondents to a CSA poll in 2008 said they would be prepared to vote for ‘a candidate belonging to a visible minority’, only 40 per cent believed that other voters would do the same. There is almost certainly an element of social desirability bias within the responses to the former question, in which case the latter figure may present a more accurate picture of the ideological climate in France. In short, ethnic prejudices may remain fairly widespread among the voting public, meaning that minority candidates would suffer an ethnic penalty. If this assumption is true, it validates the behaviour of mainstream political parties. After all, their (intrinsic or instrumental) goal is to maximize votes and not to propose an exact mirror of society. But is this so straightforward? All of these survey questions are susceptible to various biases that may distort, either positively or negatively, the actual position of the electorate. We have already discussed the problem of the social desirability bias. Another problem is the halo effect: asking respondents about ethnic representation, after having first questioned them about the need for more women and youth in Parliament, can inflate the number of positive responses. Social desirability bias and question order affects and can have a very large impact on the measure of an issue with strong connections to racism and prejudice. To more accurately assess the public state of mind regarding minority political representation (the main objective of this chapter), we need a better design that eliminates the possibility of such biases. We turn now to such a design.