Patrimonial Economic Voting: Legislative Elections in France

Abstract

Patrimonial economic voting has been neglected in favour of classical economic voting studies. This assertion holds less, however, with French election investigations, where the neglect is relative rather than absolute. Whereas classical economic voting holds the economy to be a valence issue, patrimonial economic voting regards the economy as a positional issue. Voters who own more property, in particular high-risk assets, are held to be more right-wing in their political preferences. This patrimonial effect shows itself to be statistically and substantively strong in one of the few election data-sets with sufficient measures available – surveys on the National Assembly contests of 1978, 1988, 2002. The electoral effect exceeds that from the traditional ‘heavy variables’ of class and income. Moreover, further work might show its impact comparable to that of classic sociotropic retrospective evaluations of the national economy. Certainly a case can be made for further study of patrimonial economic voting, as compared to classical economic voting.

Notes

1. In a subsequent study, Boy and Mayer (1997: 118–23) returned to this earlier index, including home ownership as well. Presenting bivariate results, they link it to vote choice. Since then, l'effet patrimoine has not been much pursued in French election studies. Dupoirer (2004: 193), in a brief comment, says that patrimony hardly adds anything to the explanation of the vote, once occupational categories are taken into account. Cautrès (2004: 88–9) reports the relationship of patrimonial elements to the 2002 presidential vote, but does not offer comment. In his otherwise excellent analysis of the 2007 French national elections, patrimony is not considered (Cautrès and Muxel 2009). Overall, it appears that the early seeds on patrimonial voting planted in France did not produce more lasting, developed, fruit. In the first French National Election Study to attend to it, that of 1978, nine survey items were devoted to patrimony. That number went steadily down in subsequent studies, when such items were posed at all: 1988 = 8; 2002 = 6; 2007 = 4. Moreover, the idea was never empirically tested over time in a systematic, multivariate way. Lastly, the theoretical dimension which we find crucial – between high-risk and low-risk assets – was never exploited. For example, in the latest survey, that of 2007, only one high-risk item was posed. Clearly, though, the French case still offers the most opportunity for growth since, unlike other nations, a data foundation has at least been laid.

2. After 1986, when the right-wing government privatised certain public sector enterprises, and urged citizens to buy stocks and make market investments, some right-wing voters may have been led to do so. However, even if such a phenomenon occurred, we do not consider that it was widespread enough to significantly reverse the causal arrow flowing from patrimony to the vote.