Bochsler, D. (2017) “The strategic effect of the plurality vote at the district level”, Electoral Studies 47, 94-112 DOI: 10.1016/j.electstud.2016.11.019

It is common knowledge that plurality voting systems (first-past-the-post (FPTP) systems) usually produce two-party systems and narrow the chances of success at the polls for small parties. Therefore, at the individual level, FPTP systems are conducive to strategic voting for the supporters of small parties, i.e. that these voters support a major party instead of the actually preferred party to avoid “wasting” their vote. Analyzing district-level data from six countries using a FPTP system, in his article, the author argues and empirically confirms, that this strategic effect in FPTP systems is uneven across districts. More precisely, the strategic effect is strongest in districts where intra-district competitiveness is high and the electoral race between the two strongest candidates is close. From a theoretical perspective, the effect is amongst others contained by two prerequisites for strategic voting: First, small party supporters should especially vote strategically in districts where a significant gap between their preferred candidate and the two candidates staying ahead exists. Secondly, small party supporters will only defect from their preferred candidate if there is no anticipated absolute majority for the strongest candidate.

Methodologically, the analysis is based on probabilistic models. In this respect, two functions are estimated in a regression with cubic splines due to the non-linearity of these functions: first, an empirical function depicting the relationship between the vote share of each of the small parties and the vote share of the largest party for each constituency. This function should capture the potential strategic effect. The second “natural” function, in contrast, does not include this effect as the first function is re-estimated without the range of values on the independent variable where the strategic effect should manifest itself the strongest. The empirical function reveals that especially when the race between the two leading candidates is close, candidates of the 3rd-largest party receive a lower share of votes as they usually would according to the natural function. However, the analysis finds no similar effect, rather small gains, for the candidates of the 4th- and 5th-largest party. Finally, the study also shows that party nationalization of small parties is not significantly diminished by the strategic effect.

Considering these findings, the article adds a new perspective to the literature on strategic behavior of political actors by highlighting the dependence of tactical voting on district competitiveness. Furthermore, the paper presents an innovative methodological approach that is highly robust to many alternative model specifications. Interestingly, the effect is stable for the whole sample, even though it comprises established as well as non-established democracies. However, the sample did not include the USA which would – as an archetypal two-party system – surely be a relevant case to analyze. At any rate, the approach presented in this article depicts a promising starting point for further research on strategic effects in FPTP systems.