Working Papers

Politicians frequently disregard the interests of low-income Americans when making policy decisions. I argue that this tendency is partially caused by politicians’ isolation among the rich. I test this theory with experiments embedded in an original survey of 1,224 state legislative candidates from 44 states. In a first experiment, I show that politicians extrapolate from their privileged social networks to form perceptions of broader social conditions, causing them to underestimate how many of those they seek to govern are struggling financially. In a second experiment, I correct politicians’ misperceptions, and find that this causes an increase in their support for social welfare programs that provide struggling families with financial assistance and healthcare. The results are driven by Republicans, who are more likely than Democrats to have privileged social networks, underestimate financial hardship among those they seek to govern, and increase their support for welfare programs when I correct their misperceptions. [You can access this paper here.]

College is a key pathway to political participation, and lower-income individuals especially stand to benefit from it given their lower political participation. However, rising inequality makes college disproportionately more accessible to higher-income students, creating many predominantly-affluent campuses. Do low-income students gain a participatory boost from attending college? How does the prevalence of affluent students on campus affect this gain? Predominantly-affluent campuses may create participatory norms that elevate low-income students’ participation. Alternatively, they may create affluence-centered social norms that marginalize these students, depressing their participation. Using a large panel survey (201,011 students), controls on many characteristics, and quasi-experimental tests, we find that predominantly-affluent campuses increase political participation to a similar extent for all income groups, thus failing to close the gap. We test psychological, academic, social, political, financial, and institutional mechanisms for the effects. The results carry implications for the self-reinforcing link between inequality and civic institutions. [You can access this paper here.]

Democracy requires voters to hold politicians accountable for violating democratic norms. Yet we know little about what increases voters’ willingness to punish norm-violating politicians, especially when the politicians are co-partisans. We examine the case of Republican voters’ willingness to sanction Donald Trump. In a nationally representative survey of U.S. adults, we randomly assign voters to messages expressing disapproval of Trump’s norm-violating behavior that vary in the message’s source and language. Contrary to the idea that Republicans distrust party elites who criticize their party leaders, we find that Republicans become more willing to sanction Trump electorally when hearing that other Republican elites disapprove of Trump’s behavior. These effects occur when elites invoke appeals to national identity and characterize Trump’s behavior as “un-American.” Our results contribute to understanding the conditions under which voters hold politicians from their own party accountable for violating democratic norms. [You can access this paper here.] 
 

America’s two-party system is sometimes thought to work better for certain race- and class-based groups than others. When politicians perceive groups as being captured by one party or the other, it may cause them to do less to advance that group’s interests. While this perspective has the potential to help explain group-based inequalities in American politics, it has not been experimentally tested. As such we do not know whether the perception that groups are captured by one party or the other has a causal effect on how politicians treat groups. To examine this possibility, we conduct a national survey of state legislative candidates in 2018 in which we experimentally test whether party elites’ preferences over group-targeted electoral strategies are influenced by whether the group in question is stereotypically aligned with the party and whether the group’s support for the party is uncertain. We find the first causal evidence that groups benefit from being “uncaptured" in America’s two-party system. [You can access this paper here.]