My job market paper draws on, and extends in new directions, work from my book project. My book project argues that party systems are at the root of broader patterns of social reform, redistribution, and levels of inequality in democracies across space and time. Democracies with more stable multi-party systems are more likely to redistribute income and possess lower levels of inequality than those with inchoate, two-party, or hegemonic counterparts. However, a cautionary corollary is that the development of strong parties – and especially those with interests that run counter to elites’ – can be a double-edged sword that undermine the very foundation of democracy itself.
Redistributivist theories claim that highly unequal democracies are more likely to breakdown rather than more egalitarian counterparts because they create incentives for both the masses to stage revolutions or soak the rich, as well as elites to use repression or stage a coup to protect their wealth (e.g. Boix 2003; Acemoglu and Robinson 2006). Yet, revolution from below has been exceptionally rare, and elites have long ‘locked in’ institutions that preemptively curb the emergence of redistribution in the future (Albertus and Menaldo 2017) – often producing conditions favorable to the long-term durability of highly unequal democracies. I argue that elites have little to fear from democracies, even highly unequal ones. Rather, what they fear most are credible threats of redistribution. Credible threats of redistribution are most likely to come in the form of a political party, considering their unique capacity to build coalitions comprised of disparate groups around a central ideology and capture formal political institutions. In turn, in highly unequal societies when emergent parties advancing redistributive goals overcome institutional barriers to threaten elites, democracy is unlikely to survive.
To develop this alternative causal pathway of democratic breakdown, I qualitatively draw on the deviant case of Brazil. I also employ a survey experiment on contemporary Brazilian economic elites that shows that elites randomly primed to think about the electoral success of a redistributive threat (in this case, the Partido dos Trabalhadores) are more likely to express more negative views towards democracy. Finally, I test the broader theory on a time-series cross-sectional dataset of 20 Latin American and Caribbean countries over the period of 1900-2020, finding strong empirical support that democratic breakdowns are significantly more likely to occur in conjunction with a redistributive threat.
The latest version of the manuscript can be found here.