Work In Progress

  • Candidate Coethnicity, Rural Dwelling, and Partisanship in Africa (Under Review) with Robin Harding Paper

Why do some citizens in new democracies attach to parties while other do not? Existing theories focus on demographic, institutional, or electoral cycle explanations. By considering voter-party linkages in sub-Saharan Africa, we investigate two major, but largely overlooked, determinants of partisanship. First, we propose that village block voting and parties' pro-rural policy bias facilitate partisanship for villagers, while transient social structures and partisan cross-pressures render partisanship less likely for urbanites. Second, we argue that coethnicity with major candidates signals substantive representation, increasing the likelihood of partisanship. By coupling an original dataset of candidate ethnicity with Afrobarometer data, we investigate these claims alongside existing explanations. We find robust support that rural villagers are more likely to be partisan. We find no support for a continent-wide relationship between partisanship and candidate coethnicity, countering prevailing wisdom. As elsewhere, partisanship is positively related to experience with multiparty democracy and the electoral cycle.

Men and women politicians face different incentives and barriers across job duty domains, which may result in performance gender gaps of different magnitudes across domains. We test this proposition by collecting information on multiple job duties and comparing reserved-seat women (``RS-women'') and men politicians' performance across 50 subnational Ugandan legislatures (where 1/3 of seats are reserved for women). We find no gender gaps in independently performed duties (e.g., facilitating school grants), while duties requiring increasing amount of collaboration with fellow politicians show an increasingly negative gap for RS-women (e.g., legislative activity). Using original network data from those 50 legislative bodies, we find RS-women are more peripheral in professional, but not personal networks. Further, RS-women have less education, lower wealth, and represent safer seats. Exclusion from professional networks, however, appears to be the primary factor accounting for the existing gender gaps, perhaps mutually-reinforcing lower performance over the term.

What determines the success of women in attaining elected political office and socioeconomic gender equality more broadly? A recent surge in the literature on this burning question has centered around whether adherence to Islam versus Christianity, or rather politico-economic factors coinciding with these religions, is responsible for cross-national variation in gender inequality. We view religion not as fixed, but fluid - its interpretation regarding gender equality is endogenous to contextual factors such as political institutions and economic development. Our new approach is thus to conduct analyses on gender attitudes within mixed religion countries in Africa, matching Muslims and Christians on individual level socioeconomic traits to better isolate the role of religion. We find cross-national variation in the within-country Muslim and Christian gap, whereby Muslims are either equally, more, or less conservative than Christians in sub-Saharan Africa. However, within-country gaps are small. On the other hand, the gaps between men's and women's attitudes towards women's leadership and equality are much larger.

  • Candidate Entry & Vote Choice in the Wake of Incumbent Performance Transparency Initiatives with Guy Grossman and Carlo Prato

While much accountability scholarship exists regarding incumbent performance transparency and vote choice, such scholarship has largely overlooked the role of strategic challenger entry. We formalize a theory of how incumbent performance, challenger entry in primaries and general elections, and citizen vote choice --- contingent on relative party advantage --- respond endogenously to civil society organization (CSO) led incumbent performance transparency initiatives occurring early in the electoral cycle. We test hypotheses derived from the theory using the 2015 primaries and 2016 general elections across 900 constituencies from 50 Ugandan subnational governments. The study leverages two sources of identification in CSO-led incumbent performance transparency: (1) matching districts where a CSO creates annual incumbent performance scorecards and disseminates to political elites to counterfactual (matched) districts; and (2) within the CSO-scorecard districts, a field experiment in which the CSO additionally disseminates the scorecard to citizens for randomly assigned incumbents for two years during the 2011-2016 electoral term. We utilize four (hitherto) unmerged datasets: general election data from the electoral commission, incumbent survey data on councilor characteristics and primary elections, CSO scorecard data, and performance data from district plenary meeting minutes. This study is a ``downstream'' study following Grossman & Michelitch's 2018 APSR publication "Information Dissemination, Competitive Pressure, and Politician Performance between Elections: A Field Experiment in Uganda" We are currently finalizing our research design for registration and preparing for analysis.


  • Same-Gender Substantive Representation: Politician Policy Priorities, Legislative Actions, and a Responsiveness Experiment with SangEun Kim

The increase in female officeholders worldwide has brought high hopes that they will substantively represent female citizens.We investigate whether female and male politicians better substantively represent same-gender citizens' policy priorities across 50 subnational Ugandan legislatures (where one-third of seats are reserved for women). We find some gender gaps in citizen policy priorities in a citizen survey. From a politician survey, we find mixed results in the overlap between politicians' and same-gender citizens' self-stated policy priorities, with male politicians having somewhat more congruence. Leveraging meeting minutes data from plenary sessions, we find that male politicians perform more actions in policy domains disproportionally prioritized by both male and female citizens. In a survey experiment, we find that politicians are equally responsive in hypothetical budget allocations to address male versus female citizens' requests for public service delivery improvements. This work challenges conventional wisdom that female office holders are willing and able to substantively represent female citizens.

We theorize that partisanship intensifies more as elections near for certain citizens due to campaign-specific factors that buoy partisan identity salience and perceived congruence with their party: (a) citizens targeted with more mobilization activities, and (b) citizens from politically-marginalized groups that share social identity with their party's nominees. Using daily cross-sectional survey data from a nationally-representative sample collected for one year prior to the US 2000, 2004, and 2008 elections, we find partisanship substantially intensifies over a campaign year (5 percentage points). The effect is larger in states receiving more mobilization activities (swing states). While black Democrats and female Republicans received increased descriptive representation from a presidential and vice-presidential nominee in 2008, respectively, only black Democrats' partisanship intensifies significantly more than comparison groups in this election. We conclude that campaigns matter because they intensify partisanship and exacerbate polarization on partisan cleavages; who becomes more polarized, however, depends on campaign-specific factors.