LAUSR

Abstract Authoritarian incumbents routinely use democratic emulation as a strategy to extend their tenure in power. Yet, there is also evidence that multiparty competition makes electoral authoritarianism more vulnerable to failure. Proceeding from the assumption that the outcomes of authoritarian electoral openings are inherently uncertain, it is argued in this article that the institutionalisation of elections determines whether electoral authoritarianism promotes stability or vulnerability. By ‘institutionalisation’, it is meant the ability of authoritarian regimes to reduce uncertainty over outcomes as they regularly hold multiparty elections. Using discrete‐time event‐history models for competing risks, the effects of sequences of multiparty elections on patterns of regime survival and failure in 262 authoritarian regimes from 1946 to 2010 are assessed, conditioned on their degree of competitiveness. The findings suggest that the institutionalisation of electoral uncertainty enhances authoritarian regime survival. However, for competitive electoral authoritarian regimes this entails substantial risk. The first three elections substantially increase the probability of democratisation, with the danger subsequently diminishing. This suggests that convoking multiparty competition is a risky game with potentially high rewards for autocrats who manage to institutionalise elections. Yet, only a small number of authoritarian regimes survive as competitive beyond the first few elections, suggesting that truly competitive authoritarianism is hard to institutionalise. The study thus finds that the question of whether elections are dangerous or stabilising for authoritarianism is dependent on differences between the ability of competitive and hegemonic forms of electoral authoritarianism to reduce electoral uncertainty.