This is the 13th in this year’s series of posts by PhD students on the job market.
The practice of veiling, or wearing a headscarf, has dramatically increased in the past decades. In Indonesia, for instance, very few women wore a veil in the early 1990s, but a recent survey reveals that more than 70 percent of Indonesian Muslim women wear a veil (Fossati, et.al., 2017). While much of the popular debates interpret this phenomenon as religiously motivated, a growing body of literature shows that veiling, especially since the 1970s, was not initiated by older, religious women in rural areas, but instead was championed by young, educated, urban middle-classes. Smith-Hefner (2007) reports that university students who opt to wear a veil tend to study in the medical and engineering program. They are those who would benefit the most from new, lucrative jobs created by economic development. Why do they choose to veil?
In my job market paper, I attempt to shed some light on this conundrum by showing that a key driver of veiling is, strikingly, the emergence of new economic opportunities that compel young women to abandon their domestic roles. The veil signals certain attributes, e.g. feminine, rule-abiding, faithful, etc. Such signaling helps young women to counter-balance traits associated with the act of taking employment outside the home, such as autonomy, ambition, or disregard of traditional morality, which may conflict with social norms. As a result, veiling increases when there are more job opportunities, as there are more incentives for women to leave their homes, but at the same time, they want to retain their reputation in society (Carvalho, 2013). Thus, veiling is a negotiation tool that functions well in a Muslim majority context like Indonesia; it is a voluntary choice by women who happen to be embedded in a culture that specifies certain ideal attributes of men and women.
How do you measure veiling?
A reliable measure of veiling is hard to come by. Previous studies rely on surveys to elicit information about veiling by directly asking female respondents whether they wear a veil. Alternatively, veiling is measured by simply observing if the respondent is wearing it during the interview. A potential problem with these methods is that they might suffer from reporting bias. For instance, if the surveyors are female or if they conduct the survey at home, then a respondent who regularly wears a veil in public may not put the headscarf on during the survey. Alternatively, if the surveyor is perceived by the respondents as having a certain degree of religious commitment (e.g. if the surveyor is wearing a headscarf herself), then this may also lead to biases in respondent answers. Another major limitation of currently available data is that no source systematically traces the evolution of this cultural practice over time. Most of the surveys are ad-hoc and conducted at different times by different institutions.
I propose an innovative way to measure veiling by coding the yearbooks of Indonesian public schools, which contain photographs of registered pupils. Indonesian public high schools commonly keep a long series of these registers, so I could count the number of female pupils who wear and not wear the veil across districts for over more than two decades. No comparable data on veiling of this scope is available to date.
Figure 1. The evolution of veiling among public high school students
More economic opportunities, higher veiling
To show that economic opportunities cause the adoption of veiling, I exploit the fact that export is one of the main drivers of the country’s economy. I build a shift-share instrumental variable (SSIV), which I call ‘export shock’, by combining the variation in international demand for Indonesian products with the gender and sectoral composition of local industries. The idea is that the combination of these three elements may generate plausibly random shocks in the availability of job opportunities for women by assigning higher dosage of “treatment” in areas that specialize in female-intensive industries and lower dosage in others.
To illustrate, let’s say there is a shock in the international demand for textile products in year x, then districts that specialize in textiles would experience shocks in the availability of jobs for women because this industry employs mostly women. On the contrary, districts that specialize in mining would not receive shocks in year x, because mining industries employ mostly men.
Using a dataset of 93 schools in 49 randomly selected districts over a 22-year period, I find that a one percentage point increase in formal employment for females has caused a two percentage point increase in the veiling rate. This number is equivalent to about ten percent of the mean value of the veiling rate in the whole sample.
Is this about social norms, or something else?
Among other explanations consistent with these results is that the increase in veiling is driven by general income increases. As the economy develops and women are getting richer, they might then think of other needs beyond necessity, such as fashion, so more veiling may simply capture this phenomenon.
To rule out this alternative mechanism, I do two things: 1) I show that employment shocks in the informal sector, where women are less likely to be required to leave their home, do not have a similar effect on veiling. 2) I also show that economic shocks to male-intensive industries, which might increase family disposable-income, do not produce a similar effect on veiling.
An alternative story is that veiling could be a type of cultural backlash. One could imagine that in a conservative society, some women may adopt the veil as a protest against the fact that more women leave their homes to work. I argue that this is not likely to be the case. Using survey data, I show that more educated young women have a higher probability of wearing a headscarf. This suggests that veil adopters are those who could benefit the most from joining the labor market. I also show that women who adopt the veil are those who take advantage of the employment opportunity themselves, and not those who stay at home.
The more room for negotiation, the stronger the effect
Finally, I examine how employment shocks interact with the strength of initial gender norms, measured by the fraction of males who declare doing house chores (typical ideal role for females in the Indonesian context) by district in the base year . I find that the effect on veiling follows an inverted U-shape: The effects are small when the initial gender norms are the strongest or weakest, and peak when the norms are more negotiable. This implies that the effect of economic shocks on veiling is highest when there is more room for negotiation about gender roles in the household. Meanwhile, the effect is weaker when everybody agrees about these gender norms or when gender norms are most egalitarian, which means a veil is not needed in the first place.
The results of my study imply that the religious headscarf could act as a liberating device that enables young women to navigate through certain gender norms while pursuing their aspiration to benefit from economic development. Hence, the rise of veiling in contemporary Indonesia might, in fact, be seen as a sign of economic modernization, which signifies the expansion of female participation in the formal sector but, at the same time, is shaped by the prevailing culture of gender relations. This study demonstrates how culture–i.e. group shared beliefs and preferences–shapes economic development in a way that seems counterintuitive at a first glance.
Naila Shofia is a Ph.D. Candidate at Bocconi University, Milan.