Adi Greif

Advisor:  Frances Rosenbluth


Does religiosity vary in the Middle East? Religiosity is often considered important to political outcomes such as state gender policies. Existing measures of religiosity use self-reported survey data or the percentage of Muslim citizens registered in a country. These measures do not accurately convey differences in levels of religiosity. Self-reported data may be biased due to incentives for individuals to exaggerate or underplay their religiosity; moreover, these data are not available for all countries. The percent of Muslims in a country assumes that all Muslims observe Islam in the same way.

This paper will develop a new proxy for religiosity in the Middle East: the extent of fasting during Ramadan. Fasting during Ramadan is one of the five pillars of faith for Islam and is widely acknowledged as required for all Muslims for whom it is medically safe to fast. The extent of fasting will be estimated using the prevalence of lights on at dawn in households before  and after Ramadan. This paper conjectures that there will be more lights on at dawn during Ramadan as compared to before Ramadan because many people who observe Ramadan wake up immediately before dawn to pray and to eat before the fast begins.

Data on lights at dawn is available using satellite data available from the Defense Meteorological Satellite Program (DMSP). The extent of fasting is not a perfect proxy for religiosity because there are individuals who do not fast or skip the dawn prayer and still consider themselves religious Muslims, and some countries observe policies that promote fasting. Despite these problems, a comprehensive and dispassionate measure of religiosity will provide a useful alternative proxy to those currently existing.