Published / Accepted

with Ali Cheema, Sarah Khan & Shandana Khan Mohmand

Conditionally accepted at American Political Science Review

What explains persistent gender gaps in turnout in some democracies? We advance a theory of women’s political participation in patriarchal settings which centers on the role of men as gatekeepers. We conduct a field experiment in Pakistan, and find that targeting only women with a nonpartisan GOTV campaign has no effect on their turnout in a national election. However, women’s turnout increases substantially when men are canvassed to support the participation of women in their household. Using a behavioral measure, we find that canvassing men increases their publicly expressed support for women’s role in democracy two months after the election. In households where both men and women are canvassed, we also observe an increase in political discussion, and in men’s practical support to help women vote on election day. The results suggest that constraints emanating from men are central to understanding and reducing gender gaps in political participation.

with Ali Cheema & Shandana Khan Mohmand. In Pakistan's Political Parties: Between Dictatorship & Democracy. Georgetown University Press. 2020.

Political contact and closeness takes many forms, and while the immediate reasons for the initiation of political contact may be linked to electoral campaigning or the resolution of service delivery issues, contact and closeness also serves as a primary vehicle for the transmission of citizen preferences to politicians and party workers. Using original surveys with 2,150 voters and interviews with 33 politicians, this chapter shows that those who are in contact with or close to politicians have markedly different characteristics from the average voter. To the extent that politicians derive their beliefs about citizen preferences from these forms of contact and closeness, and to the extent that they take political decisions based on their own beliefs about citizen preferences, these findings have important implications for the representation and implementation of citizen preferences. They also demonstrate the critical role that political machines play in transmitting information about citizen preferences up to higher-tier politicians.

Working Papers

Under Review

Studies on the role of information in political accountability usually ask whether voters know enough about politicians. In this paper, I reverse this standard approach by asking instead whether politicians know enough about voters to adequately represent them. I develop and test a theory of how politicians acquire and respond to information about citizen preferences. Using original surveys of 4,578 citizens and 653 local politicians in Pakistan, I show that politicians have highly inaccurate beliefs about citizen preferences and high demand for more information. In collaboration with the second-largest political party in Pakistan, I conduct a field experiment to test the conditions under which local politicians are responsive to citizen preferences. Politicians who receive information about citizen preferences make recommendations to their party leadership that are closer to what citizens prefer. Directly elected politicians are more responsive than indirectly elected ones. Politicians are more responsive to women's preferences compared to men's preferences. I construct a simple model of belief updating which suggests that the greater responsiveness to women's preferences is driven by lower confidence in prior beliefs about women. This paper shows that our understanding of low accountability in developing democracies is missing an essential ingredient: politicians' inaccurate beliefs.

with Michael Callen, Ali Cheema, Adnan Khan, Farooq Naseer & Jacob Shapiro. Under Review.

Do voters care about how connected their candidates are? We investigate this question in the 2015 local government elections in Pakistan combining: (i) data on ties between candidates, higher level politicians, and bureaucrats; (ii) a large-scale field experiment; and (iii) election outcomes. Before the election, voters considered local candidates' connections important and expected local politicians to help them access services provided by other levels of government. Providing voters information on connections increased support for more connected candidates, but information on past party performance did not. More connected candidates received more votes and were more likely to win office, but there was no electoral benefit to past service provision. The results provide novel evidence of the importance of political connections for electoral outcomes and show that forward-looking expectations based on candidate characteristics and an understanding of higher-level political process play an important role in vote choice.

Party Over Person: Preferences for Leaders in a Pakistani Megacity

with Haseeb Bajwa, Ali Cheema & Shandana Khan Mohmand.

On what basis do voters in developing megacities choose their local leaders? We examine citizens’ preferences for local leaders in the Pakistani megacity of Lahore. We conduct a conjoint experiment with 2,150 male and female voters in Lahore to examine the characteristics they prefer in local leaders, and find that these urban voters have preferences for their local leaders that considerably differ from the rural residents and slum-dwellers highlighted in the literature. Co-Partisanship trumps all other characteristics, with education being a close second. Political connections formed on a personal basis lose the importance they assume in rural settings and in informal urban settings. Voters do not exhibit any preference for a leader being born and raised in the community. Contrary to the co-ethnicity preference that is often considered the hallmark of South Asian politics, voters do not prefer leaders from the same caste group as themselves. This paper is the first attempt, to the best of our knowledge, to understand the emergence of local leaders in a formal urban setting in a developing country.

with Ali Cheema, Sarah Khan & Shandana Khan Mohmand.

Why does a gender gap in voting exist in Pakistan? Our research looks beyond the creation of democratic spaces for women’s participation, such as voter registration, to look instead at the constraints that women face in being able to use such spaces. This paper uses qualitative fieldwork undertaken in Lahore over 2017–18 to understand what enables or constrains women’s decision to turn out to vote. Standard explanations for Pakistan’s large gender gap in political participation highlight overt restrictions imposed on women voters by male family and community members. We do not believe that these explanations are complete and look instead at more subtle processes that socialize women into non-political roles, and result in a ‘gendered psyche’ that makes women feel invisible and irrelevant to the electoral process. Constraints resulting from political engagement are usually under-emphasized in academic and policy literature, and analyzing their importance is a major contribution of this paper.

Under what conditions are politicians responsive to voter needs and preferences? We use survey data from 2,150 citizens and 40 interviews with local brokers and party officials to test whether responsiveness varies by the structure of machine politics and political competition at a local level. We find that there is considerable variation in the extent to which both municipal and non-municipal service delivery improves between the last general election in 2013 and our survey in early 2017. The density and competitiveness of party structures at the local level explain this variation for locally provided municipal services, but not for centrally allocated non-municipal services, such as electricity, gas and roads. In particular, we find that both the density of party networks and the extent to which the number of party workers present in a locality are proportional across parties are both positively correlated with improvements in the delivery of municipal services. In other words, what matters is not only how many brokers are operating within a locality, but also whether or not they belong to different parties that are actively competing with each other for the vote.

In Preparation

Overseeing the Machine: Monitoring the Effort of Political Party Workers

(Preliminary draft available upon request)

Can monitoring by political parties induce their workers to expend greater effort in electoral campaigns? I answer this question through a large-scale field experiment in collaboration with a major political party in Pakistan, a context where the costs of mobilizing men are lower than the costs of mobilizing female voters. Monitoring the effort of political workers on male voters does not increase contact with male voters, but decreases contact with female voters. Monitoring the overall effort of political workers increases contact with male voters, but does not affect contact with female voters. These results shed light on principal agent relationships within political parties and norms against the involvement of women in politics.

Precision versus Proximity: Experimental Evidence on Bureaucrats' Decision Making from Pakistan

with Michael Callen, Adnan Khan & Asim Khwaja

(Preliminary draft available upon request)

Bureaucrats take decisions with enormous welfare consequences in developing countries. Researchers frequently present evidence to bureaucrats aiming to convince them to use this evidence in policy-making. Using a lab-in-the-field experiment with 746 civil servants in Pakistan, we show that key assumptions underlying this process are unwarranted. In particular, two key features of policy research are not aligned with how policy-makers respond to evidence. The first feature is that researchers produce large-N evidence. We find that policy makers update their beliefs substantially when presented with small-N evidence, but large-N evidence does not shift their beliefs more than small-N evidence does. This result casts doubt on the ability of large-N research to convince policymakers. The second feature is that policy research is conducted with high internal validity in some areas and it is expected that policy makers in ‘similar’ areas will use it to make policies in their own areas. Experimental results show that policy makers update more when given small-N evidence from their area compared to large-N evidence from other areas. This implies that to convince policy makers, researchers should supplement large-N research with small-N research in the local area.