Research

Who’s Watching? Observer effects in public opinion research with Irfan Nooruddin.

Research on social desirability effects in public opinion surveys consider question and interviewer effects. Here, we take advantage of a survey item recording the presence of other people at the time of the interview, to document a new kind of observer effect. We analyze twelve survey questions pertaining to political interest, participation, and opinion in the Indian National Election survey in 2009 (n=34,410). We find that women are less like to report an interest in or an opinion on politics if their husband or other adult family members are around at the time of the interview. Men are more likely to report participating in political activities if a small crowd is present. These findings have important implications for our understanding of political participation and the use and analysis of public opinion surveys in large parts of the developing world. Working paper version

Powering through – political uncertainty in hybrid service delivery regimes

The ability to maintain control over public goods is at the heart of the debate on distributive politics. This study investigates outsourcing of state-provided public goods to private stakeholders – what happens to the relationship between citizens and political leaders, when the latter no longer influence service delivery? Taking advantage of the varying quality of privatized electricity in the megacity of Karachi, Pakistan, and the hybrid provision of water in the city, I suggest that the privatization of public goods can increase political uncertainty. In particular, I show that citizens experience the decline of successful claim-making to the state for privatized public goods, I find that this learnt behavior has a spillover effect on other, state-provided public goods. Ethnographic data from over ten months of fieldwork, and a unique dataset of over 25,000 service delivery clusters, are used to examine the effect of varying quality of service delivery on political outcomes. Finally, I test these claims using an original survey of 1000 households. I suggest that instead of being viewed as discrete outcomes, two or more services can interact to inform lived experiences and behavior in urban centers. In particular, positive learnt experiences from privatized public goods can spillover to state-provided public goods, and negative experiences can cause voters to exit.

Disempowered: Political trust and service delivery

Why are some households more likely to blame their neighbors for low-quality public services, and not the state or a service provider? States across the developing world seek to improve revenue collection for services such as electricity: often, by contracting to private firms. A key focus of these reforms is to differentiate between paying and non-paying consumers. Using original qualitative and survey data from Karachi, Pakistan, I examine the implementation of these reforms and find that many paying households get ‘stuck’ in non-paying neighborhoods, and are subject to collective punishment. I leverage a dataset of 25,000 service delivery clusters to suggest that uneven electricity distribution coincides with low political trust. Using a survey experiment, I find that messages informing citizens of theft have a causal impact, and lower reported trust in neighbors. I suggest that transitions to privatized utility regimes have unintended adverse consequences for political trust and participation. Working Paper Version