Downsizing Democracy: Why Ordinary People Acquiesce to Authoritarianism (book manuscript in progress)

Support for Democracy: Attitudinal versus Revealed-Preference Measures, co-authored with Austin Jang (Current draft: January 2024)

We present a large-scale, multi-country assessment of the most frequently employed survey measures of support for democracy. We contrast such conventional, attitudinal measures to revealed-preference measures. The latter are based on voting for candidates in experimentally manipulated scenarios that mimic real-world electoral trade-offs between democratic principles and competing political considerations, such as partisan loyalty or policy preferences. Only a subset of attitudinal measures are both monotonic and discriminating in experimentally revealed commitment to democracy, two criteria that we develop to assess their performance. Two popular measures of support for democracy exhibit either no or reverse relationship to experimentally revealed commitment to democracy, resulting in potentially misleading findings. These findings are robust to a range of statistical techniques, including conventional and machine learning approaches for detecting treatment effect heterogeneity, and hold across a number of countries with diverse levels and histories of democracy. We propose a number of recommendations for practice, including improved formulations of established measures of support for democracy as well as set of alternative measures.

Partisan Mistrust and the Erosion of Democratic Norms, co-authored with Matt Graham

We investigate the nature and consequences of systemic partisan mistrust: suspicions across party lines about the other side’s commitment to fundamental democratic principles. We develop a model of self-enforcing democracy that identifies partisan mistrust as a distinct source of democratic fragility and guides our empirical analysis. In a series of incentivized experiments conducted ahead of the U.S. 2020 general and 2022 midterm elections, a representative sample of Americans allocated money for donation to real-world partisan candidates and advocacy groups, some of which promoted policies that undermine democracy. A subsequent sample was used to estimate higher-order beliefs about both in and out-partisans’ actions. This design allows us to infer i) citizens’ willingness to prioritize democratic principles over partisan interests, ii) beliefs within and across party lines about others’ willingness to prioritize democratic principles, and iii) the role of partisan mistrust as a preemptive rationale for the erosion of democracy. In the latter case, partisan actors disregard democratic norms not because they do not value them, but because they do not trust their opponents to do so.

The Singapore Model of Authoritarianism, co-authored with Lee Morgenbesser

We propose a novel explanation for the emergence, persistence, and decline of dominant parties. We develop a model in which a political party’s temporary electoral advantage results in a self-sustaining cycle of quality-based incumbency advantage. In equilibrium, i) the dominant party recruits candidates of higher quality than its competitors, ii) highest quality candidates aspire to join the dominant party because it is more likely to win elections, iii) voters correctly anticipate the dominant party’s quality advantage and re-elect it. We evaluate this framework by combining a range of qualitative and quantitative evidence from Singapore. Using original data on all candidates for political office since independence, we show that the governing People’s Action Party (PAP) indeed fields more competent candidates than its competitors. Using survey experimental data, we show that Singaporeans value competence and associate it with the PAP, and that when Singaporeans vote for the PAP, they are in effect trading-off democracy for competent governance. Our analysis clarifies why dominant parties are most vulnerable to defectors from their own ranks, why the opposition in dominant party regimes tends to be more ideological and fragmented, and why the demise of dominant parties typically follows a prolonged period of economic decline.

Testing Formal Models of Political Competition: Evidence from Conjoint Experiments, co-authored with Daniel Goldstein

We assess the assumptions and predictions of formal models of political competition employing data from candidate-choice conjoint survey experiments. An advantage of the latter is that, by design, candidates adopt the full range of policy positions, a variation that is rarely observed in real-world elections. Using over 20,000 candidate-choices, we define and estimate candidates’ “empirical best response functions” and estimate the empirical equilibria of several prominent models of electoral competition, including the Downs-Hotelling model, models with valence advantage, multidimensional models, and models of turnout. We find that voters often, though not always, have single-peaked preferences, tend vote according to those preferences, and that optimal candidate platforms weakly converge to the median voter’s ideal point.

Who (Really) Supports Democracy? Experimental Evidence from Revealed Preference Measures, co-authored with Austin Jang  and Filip Milačić (Current draft: Aigust 2023)

We present a large-scale, multi-country analysis of the predictors of citizen-level commitment to democracy that is based on revealed-preference rather than attitudinal measures. We find that only a small set of personal characteristics consistently accounts for nearly all variation in commitment to democracy across individuals and countries. These findings are robust to a range of statistical techniques, including conventional and machine learning approaches for detecting treatment effect heterogeneity, hold across a number of countries with diverse levels and histories of democracy, and predict voting for real-world parties and candidates with authoritarian tendencies.

Don’t Back No Losers! The Supply of Political Labor and the Political Organization of Clientelism (Current draft: April 2014)

The tacit promise of employment or some other office-based favor in return for electioneering activities is a key element of the relationship between candidates and their political operatives in many developing democracies. Yet the implications of the politically conditional nature of this exchange — the fact that the operatives involved in these activities obtain the promised rewards only if their candidate wins an election — have not yet been explored. We show that the politically conditional nature of this exchange has far-reaching implications for the political organization of clientelism, patronage politics, and election campaigns. We present a unified framework that explains i) which parties adopt clientelistic as opposed to programmatic campaign platforms; ii) how patronage politics amplifies the advantages typically attributed to incumbency; iii) why there is usually only one clientelistic party per party system;  and iv) why a system-level transition from clientelism to programmatic politics is more likely to arise out of technological change than candidate initiative. We support these claims by developing a new measure of clientelistic effort, which we apply to a newly collected dataset of precinct-level election results from the city of Chicago throughout the period 1907-1983.

Electoral Competition with Campaigns (Current draft: April 2015)

We develop a model of electoral competition in which a candidate’s ability to reach voters depends on the number of campaign activists willing to work on his behalf. There are two types of activists: opportunists, who only want to work for a winner, and idealists, who only want to work for a candidate whose platform is close to their ideal point. In equilibrium, campaigns amplify pre-existing differences in electability between candidates, generate an incumbency advantage, and result in platform divergence. In the last case, less electable candidates adopt more extreme platforms because they are deserted by opportunistic activists and must court idealistic ones in order to run a campaign. We thus identify a new rationale for platform divergence in electoral competition: the constraints on candidates’ platforms implied by the need to attract campaign labor.

Democracy and Dictatorship in the Age of Elections: Inferring Regime Type and Change from the Intensity of Electoral Competition (Current draft: April 2014)

The defining feature of democracy, according to most political scientists, are free and competitive elections. Yet the most frequently employed measures of regime type incorporate only a very limited amount of information about electoral competitiveness: these measures take the form of either binary indicators or index-based scales, neither of which explicitly measures how competitive elections are. This approach fails to reliably differentiate democracies from dictatorships when almost every country in the world holds multiparty elections. In this paper, I develop a new approach to the measurement of regime type by using data on the margins of victory in executive and legislative elections throughout the world. This approach has a number of advantages over existing metrics: i) it explicitly incorporates information about the degree of political competitiveness implied by electoral margins of victory; ii) the number of regimes and points of transitions between them are both inferred from the data rather than assumed at the outset; and iii) it comes with an explicit statement about the degree of uncertainty in each regime type estimate. I illustrate the value of this approach by estimating the number and changes in regime type around the world throughout the period 1815-2012.

Do Minority Rights Make a Difference? Democracy, Citizenship, and Conflict in Multi-ethnic Societies (co-authored with Bonnie Weir, current draft: April 2015)

We conduct the first empirical analysis of the consequences of linguistic minority rights for those affected by them. We take advantage of unusual provisions for linguistic minority rights in Slovakia and Romania, where a number of such rights are provided only when the minority meets a municipal-level minority population threshold. Threshold-based minority right provisions are exceptionally well-suited to the investigation of their causal effects and allow us to administer an original, regression discontinuity-based survey. Our findings help us assess the validity of a range of widely-presumed consequences of linguistic minority rights, including attitudes toward national citizenship and local community, ethnic identification, inter-ethnic levels of trust, civic life, and political engagement. We highlight the implications of our findings for foundational questions about the political role of nationality, ethnicity, and language in democratic participation, modern citizenship, and civil conflict.

The Paradox of Absolute Power: The Reputational Foundations of Personal Leadership and Political Order (Current draft: July 2012)

Even the most despotic leaders must rely on a large number of individuals in order to govern. How do leaders overcome such dependence and exercise seemingly unlimited authority? We develop a model in which a leader successfully deters challenges to his position by cultivating a reputation for invincibility. Such reputation prevents an otherwise indispensable ally from establishing a political base of his own. The process of maintaining such a reputation has several appealing features: a reputation deteriorates if not repeatedly cultivated, it is optimally exploited when needed, and it is established by seemingly arbitrary costly actions. These features match puzzling historical accounts according to which many despots periodically but unexpectedly humiliated, dismissed, but also resurrected their closest allies. These actions have been typically attributed to cultural factors or individual pathologies. Our analysis clarifies their political micro-foundations and yields insights beyond the study of authoritarian politics by highlighting the reputational foundations of personal authority.