I am interested in understanding how culture shapes political and development outcomes. By culture, I mean shared beliefs and preferences of groups, in particular, which are shaped by social norms, religion, and social identity. Culture shapes political and economic outcomes through the tastes and wants of respective group members. One implication is that understanding this process is fundamental to comprehend the causes of certain outcomes and to inform the design and implementation of various policies in a way that addresses, or even takes advantage, of this knowledge. While paying special attention to causal identification, my work applied a broad range of methods, including advanced econometrics, survey experiments, and social statistics.
Coverage : World Bank Development Impact
Abstract
We show that new economic opportunities that compel women to abandon their domestic roles have driven the adoption of religious veiling. We measure the prevalence of veiling among young women across Indonesia's districts for more than two decades by human-coding around a quarter million photographs attached to Indonesia's public high school yearbooks. We exploit exogenous variations generated by the interaction between international demand for Indonesia's products and sectoral and gender composition of local industries to show that the relationship is causal. This study suggests that veiling represents an effort by young women to reconcile their desire to join formal labor markets and the prevailing social norms in society.
Wearing your religion: Evidence on employers’ preferences in Indonesia
(with Eliana La Ferrara – Harvard Kennedy School and Stefano Fiorin – Bocconi University)
Abstract
In collaboration with one of the largest job portals in Indonesia and over 1,200 employers, we examine the impact of displaying gendered religious identities—headscarf for females and beard for males—on labor market outcomes. Using a modified version of the Incentivized Resume Rating (IRR) method, we find that displaying religious identity negatively affects hiring outcomes, including the likelihood of being hired, the probability of being interviewed, and the salary offered to candidates. The penalties associated with headscarves and beards are similar in magnitude and do not vary significantly across different job levels, job types, job gender requirements, or the gender of employers. However, the patterns differ between urban and rural areas and are less pronounced for pious Muslim employers. In a beliefs-elicitation exercise, we find that candidates displaying religious attributes are perceived as more religious but are rated lower by employers in terms of quality and personality traits deemed important in the workplace.
Accountability vs. Identity: Revealed Preferences of Indonesian Voters
(with Piero Stanig – Bocconi University)
Abstract
Do voters support incompetent candidates that share their religion or ethnicity? Can voters be encouraged to prioritize a candidate's track record? We run a multi-wave survey experiment on a large representative sample of Indonesian citizens. Respondents chose between candidate pairs differing in track record, policy stances, and ascriptive features like ethnicity and religion. Track records, especially those related to corruption, are the most important for voters; ethno-religious congruence and ideological proximity follow; importantly, voters are willing to support out-group candidates with good track records. An informational intervention—explaining in simple terms the logic of elections as instruments of accountability—significantly increases the weight voters place on track records, especially among those initially less informed about the idea that elected officials are agents (“employees”) of the public. Overall, while identity politics matters, voters prioritize candidate quality. A simple and scalable informational intervention further increases the importance that voters assign to performance in office.
Working Papers
Abstract
This study postulates that the degree of competition between factions within the politically influential (majority) group in society could explain the severity of ethnic conflicts. Competing factions engage in various activism, including violence against "others", to signal their commitment towards the group, to gain sympathy and support. I devise a new index of factional competition to capture this phenomenon. Using the context of Indonesia’s consolidating democracy and its religious organizational dynamics, I show that this index is strongly correlated with the severity of religious conflicts, measured by the number of fatalities from religious violence. This relationship is robust to controlling for widely used fragmentation and polarization indexes as well as a wide range of confounders. In placebo tests, I show that this index has no relationship with other types of violence, such as crimes or domestic violence, or even other conflicts that are not religiously motivated.
Working Papers
Money or ideology: The motivation of politicians in a young democracy
Abstract
Which factors shape the behavior of politicians? Using a novel, hand-coded dataset of over 1500 reports of wealth declaration by Indonesia’s Member of Parliament, I attempt to shed light on this question. Multiple reports submitted by individual MPs at a different point in time allow me to estimate the effect of sitting in a parliamentary position on personal wealth accumulation by holding individual characteristics constant. I find that politicians from ideological (Islamist) parties significantly differ from their non-ideological MP fellows in characteristics that are in line with their ideological tenets: They are poorer, have more children and are less likely to be a woman. Importantly, compared to non-ideological MPs, politicians from Islamist parties have a significantly lower growth of personal wealth. Further analysis suggests that different mechanism of vote gathering is the driver for this gap in behavior. The findings suggest that ideology could play as a disciplining mechanism for politicians in a clientelistic political environment.