My research interests focus on three main questions: How do religious and ethnic identities affect political and economic behavior? How do politics impact identity? And what do these processes imply for how we measure identity, across space and time? My interest in these questions is deeply rooted in the Turkish case, where I have conducted extensive fieldwork. 

Throughout my research, I focus on the twin challenges of conceptualization and measurement, especially as they relate to religious and ethnic identity.


It was recently named a 2022 Best Book by APSA's MENA Politics section.

In this book, I challenge existing explanations for the success of Islamic-based political and economic movements in Turkey and across the Muslim world. In contrast to theories that emphasize the personal piety of these movements' supporters, using a combination of qualitative observation and quantitative analysis, I demonstrate that the Islamic advantage is rooted in feelings of trust among individuals with a shared, religious group-identity.

Read the introduction here. More information, including the Online Appendix, can be found here.

Read reviews of the book here, here, here, and here.

Purchase a copy directly from CUP or through Amazon. 

Social scientists have long been interested in how religious beliefs and practices impact and are impacted by socio-political and economic processes. Most recently, scholarly attention has focused on the interplay between religiosity and local actors, events, and institutions. Until now, measures of religiosity have relied heavily on self-reports in surveys, but these cannot always be safely collected and tend to be costly. Even where available, survey-based measures may be too obtrusive and are rarely representative of sub-national units. Here, I propose an inexpensive method that uses satellite imagery to unobtrusively estimate religiosity across small geographic units. I hypothesize that night-lights are affected by the behavior of fasting Muslims during Ramadan, especially in places where daytime activities are otherwise unchanged (i.e., where this is no “day-night inversion”). I explore and confirm the validity of this measurement strategy in the Turkish case, using a series of high-quality surveys and electoral results, representing 973 administrative districts. I conclude with a discussion of the external validity of this method and an overview of the ethical concerns raised by the use of remote sensing to estimate religiosity, in the Muslim world and elsewhere. (Replication Materials)

Causal mediation analysis (CMA) requires measurement of an outcome variable (O) with and without treatment, plus a set of mediator variables (M) that constitute possible pathways for the treatment effect. There is no consensus on whether surveys should measure potentially mediating variables before or after the outcome variables – “MO” or “OM.” We use a replication exercise to demonstrate how the order of mediator and outcome items can be consequential for the results from CMA. Order can affect mediation conclusions, even if the treatment effect is similar across designs. As such, randomizing order is usually prudent, though best practice depends on the researcher’s contextual knowledge about her particular application.

See the Appendix here.

Before the Arab Spring, the Middle East and North Africa was cited for its low levels of political mobilization, but recent events call for a more systematic study of “protest potential” in MENA and the broader Muslim world. Combining surveys from 141 countries, I identify a significant participation gap in Muslim-plurality countries: individuals living there are less likely to have participated in politics and are also less likely to self-report being willing to do so, a difference that holds across all demographic categories. After confirming the validity of retrospective and prospective self-reports using a survey of participants in Istanbul’s Gezi Park protests, I explore the role of religion in the Muslim world’s participation gap: while Islam, as a doctrine, does not inhibit activism, certain forms of religiosity do undermine participation. In particular, private beliefs and practices limit protest potential in the Muslim world, while collective religious engagement bolsters it. The latter, I argue, supports activism by helping address the collective action problem underlying mass politics.

Now available online as part of the Oxford Handbook of Politics in Muslim Societies (2020), edited by Melani Cammett and Pauline Jones.


I am PI on a multi-year project updating estimates of ethnic, religious, and linguistic diversity around the globe. Our approach combines self-identification from surveys and a database of census results with machine learning to identify and correct systematic measurement error.

Find the full project description here. More information about the NSF award available here.

Working Paper

"The Cultural Legacy of Tolerance: The Case of Ottoman Turkey"

with Mark Westcott

After centuries of inter-mingling across religious lines, Ottoman-Turkey was transformed, through a series of state-sponsored interventions, into a religiously homogenous republic, with lines dividing the Muslim population by ethnicity. We explore the legacy of Ottoman-era tolerance in contemporary Turkey, combining georeferenced historical censuses with novel surveys and making use of an instrumental variable to support causal identification. Ultimately, we find that a positive legacy persists: individuals from once-diverse districts are today more trusting of non-coreligionists and more tolerant of non-coethnics. Leveraging variation in internal migration, we show that the Ottoman legacy is a cultural artifact, passed down across generations. In so doing, we demonstrate that tolerance – and not just intolerance – can survive over time, even without formal institutions to support it and in the face of concerted efforts by the state to reverse it. 

Public Scholarship