A Dynamic Model of Political Party Equilibrium: The Evolution of ENP in Canada, 1870-2015

The effective number of political parties (ENP) in a single member plurality rule electoral system is analyzed as a dynamic process whereby the tournament nature of the election contest induces excessive entry and sunk entry costs promote persistence even as Duverger-Demsetz type political competition works to winnow unsuccessful minor candidates and parties. The result is a fringe of parties that continue to circulate in long run equilibrium. The factors hypothesized to affect the entry and exit of candidates and parties are analyzed for Canada from 1870 through 2015 first using an auto-regressive distributed lag (ARDL) model and then allowing for asymmetric adjustment by adapting NARDL panel estimation techniques. After finding evidence of asymmetry at the party level, the NARDL results uncover two new stylized facts for ENP at the national party level: (i) a continual rise in the short run fringe after 1945; and (ii) a concomitant long run decline in ENP to about 1.5 (below Duverger’s prediction of 2). The long run time path in ENP at the party level is inversely correlated with uncertainty-based measures of electoral competition, suggesting that movements in long run ENP closer to 2 are an indicator of more rather than less electoral competition.