In electoral autocracies, opposition coalition formation offers the best hope of getting to democracy. Yet forming electoral coalitions also entails convincing opposition voters to ignore compromises and engage in the cross-party voting necessary for opposition victory. To what extent are voters committed to defeating the autocratic incumbent even if it would result in dislikable outcomes? A survey experiment in Malaysia finds that opposition voters overwhelmingly express pretreatment support for the opposition coalition. But when exposed to a treatment vignette about which member party might lead the next government, many voters retract their support. Specifically, voters' support for the coalition declines when their least preferred member is expected to control the government and when they can vote for a closer ideological alternative outside of the coalition. Although voters are committed to opposition unity and democratic transition, that commitment is sensitive to the anticipated consequences of an opposition victory.

Governments around the world have crafted new laws to threaten, arrest, prosecute and incarcerate online political activists. While the primary effect of online repression is to silence criticism and forestall collective action, a secondary effect is to induce self-censorship among the masses. Yet scant research examines how self-censorship works, nor discusses its implications for entrenching authoritarianism and encouraging democratic backsliding. This article proposes a simple expected utility model of self-censorship, arguing that citizens will more likely self-censor when the expected costs of online political expression outweigh its benefits. Analysing the fourth wave of the Asian Barometer survey of 10,216 respondents across eight Southeast Asian countries, I find that higher income politically engaged social media users are indeed less likely to express their political opinions. Additionally, this correlation holds in states where online repression is most severe, but is non-existent in countries where online repression is moderate or low.

No abstract

No abstract