Against the Wind: Labor Force Participation of Women and Economic Instability in Iran

ABSTRACT

In the last three decades, Iranian women's educational attainment has continuously increased while their fertility rate has fallen rapidly. Yet in spite of these developments, which in many countries have a positive effect on women's labor force participation, female labor force participation (FLFP) rates have remained at low levels. This paper argues that despite its overall static trend, FLFP of some Iranian women responded to economic pressures induced by macroeconomic instabilities. Looking at the Iranian economic crisis of 1994–5, the study shows that, controlling for individual fixed effects, married women in rural areas and never-married women in urban areas increased their participation rate by as much as 38 percent. No change in hours worked was found for any group of women. The differences in responses and their underlying reasons have policy implications for many developing countries.

KEYWORDS:

• Female labor force participation
• economic instability
• Iran

JEL Codes:

• J22
• O53

ACKNOWLEDGMENTS

I am grateful to the anonymous reviewers, associate editor and editors, and John Strauss, Jeffery B. Nugent, and Djavad Salehi-Isfahani for their valuable comments, suggestions, and advice. I also wish to thank the Wallis Annenberg Foundation Fellowship for Research on Women and Families and the College of Letters, Arts, and Sciences at the University of Southern California for financial support.

NOTES ON CONTRIBUTOR

Mahdi Majbouri is Assistant Professor of Economics at Babson College. He received his PhD in Economics from the University of Southern California in August 2010. He holds a BS in Mechanical Engineering and an MBA from Sharif University of Technology in Tehran, Iran. His expertise is broadly in the microeconomics of development and economics of gender with a focus on the Middle East and North Africa (MENA) region.

Notes

1 For the trend in women's education in Iran, see Djavad Salehi-Isfahani (2005a) and Mahdi Majbouri (2010).

2 The fertility rates were relatively stable before 1984 and after 2000. They were around 6 and 6.5 children per woman between 1970 and 1984, respectively, and a bit less than 2 children per woman after 2000 (Abbasi-Shavazi, McDonald, and Hosseini-Chavoshi 2009).

3 Looking at the impact of the same crisis and using a novel approach, Seçil A. Kaya Bahçe and Emel Memiş (2013) found that, in Turkey, a 1 percentage point rise in the unemployment risk of a spouse increases women's and men's work time by 5 and 1 percent, respectively.

4 For example, Mahdi Majbouri (2015b) uses first-born twins as an instrument for the number of children and finds that an exogenous increase in the number of children reduces FLFP for women ages 18 to 44.

5 Author's calculation based on annual household expenditure and income surveys of 1990–2004. In these datasets, one is considered a participant in the labor force if she is looking for work or worked at least two days in the week preceding the survey. The definition of participation changed after 2005, and numbers are not comparable later on.

6 Similar arguments are offered in the context of the Middle East. For example, Jennifer C. Olmsted (2005a) discusses the factors that strengthen the “patriarchal social contract” of the male-breadwinner and explains the role of the state through the legal and social framework of safety nets. İpek İlkkaracan (2012) shows that the male-breadwinner mindset on the demand side of the labor market has resulted in low female participation in Turkey. Massoud Karshenas and Valentine M. Moghadam (2000) and Ragui Assaad (2005) attribute low FLFP to the interrelation of oil revenues and demand- and supply-side discriminatory institutions. Ross (2008) argues that oil and gas profits lower FLFP, and Majbouri (2014) shows that the impact of oil and gas profits is larger where family laws give men the right to veto women's decisions to work.

7 Rural men were about five times less likely to work as unpaid family members relative to rural women, and in urban areas they were ten times less likely to work as unpaid family members.

8 In 1979, a revolution changed the constitutional monarchy into an Islamic Republic, a system of government that has a religious figure at the head of the state overseeing the military forces, as well as the judicial, executive, and legislative branches. This figure, known as the Supreme Leader, is selected by an assembly of religious experts. The president, members of the parliament, and the assembly of experts are chosen in elections, but candidates are vetted in advance by a committee of twelve appointed clerics and lawyers, known as the Guardian Council (the six clerics on the Guardian Council are appointed by the Supreme Leader; the six lawyers are recommended by the judiciary and officially appointed by Parliament).

9 The Iran–Iraq War started in September 1980 and lasted until August 1988. It is the longest conventional interstate war in the twentieth century. Estimates of casualties vary greatly depending on the source. Some point to nearly half a million Iraqi and Iranian soldiers dead (Dilip Hiro 1991).

10 The five-year development plans set mostly economic and social objectives for the government to achieve in a period of five years. The first five-year plan was implemented between 1989 and 1993, and its main goal was post-war reconstruction. Since then, these plans are designed and implemented every five years.

11 Fixed-effect probit and logit models give similar results. Coefficients of linear probability models are presented as they are easier to interpret as marginal effects.

12 There are a few households whose composition and members completely changed in the later rounds. It is unlikely that they were the same households but they had been given the same household identification number over time. Interestingly, they were all in the same location and presumably interviewed by one interviewer. Supposedly, the original household, which was given that identification number, moved out and a new household came in. The interviewer did not compare the new household to the old one and gave the new household the same identification number. If the data is sorted based on household identification number, these households can be identified with identification numbers between 11505803 and 11505821.

13 See http://www.irimet.net/.

14 The full results, such as those in Table 3, are available upon request.

15 Iran was divided into twenty-five provinces in the 1992–5 period. The country is approximately 17.2 percent of the size of the United States and 17.7 percent of the size of China. Since the provinces are not large, using rainfall at the province level is a good way to capture variation in precipitation.

16 P-values are 32 percent, 91 percent, and 67 percent for rural women, urban women, and rural men, respectively.

17 Never-married women are still controlled by patriarchal norms and do not have much freedom or control over various aspects of their lives, especially compared to their male counterparts.