ANALYZING FEMALE LABOR FORCE PARTICIPATION IN Identifying the Key Barriers that Prevent Women from Entering the Labor Force By Ishani Desai and Li Li, under advisor Professor Michael Callen

Ishani is a program ofcer at Results for Development, where she leads the M&E eforts for a global health program. Ishani previously worked with the World Bank assessing household risks and coping mechanisms in Afghanistan. She graduated from Harvard Kennedy School’s Master in Public Administration in International Development program, where she spent a year examining female labor force participation in Afghanistan. Li graduated from the Master in Public Administration in International Development program from Harvard Kennedy School in 2016. She currently works at Fred Hutch as a systems analyst and researcher on cancer outcomes. Her previous work focused on and women’s economic empowerment, such as gender-based violence and CCT programs for women entrepreneurs in Ghana.

I. INTRODUCTION outside the home.4 Since 2012, Afghan women he current female labor force participa- have been stating “lack of job opportunities” T tion (FLFP) rate in Afghanistan is around as one of the top three largest problems facing 16 percent—one of the lowest in the world.1 women. In the broader South Asian context, Afghanistan ranked 171st among 188 countries traditional norms have played a large role in in the UN’s and, in dictating gender roles; nevertheless, indica- 2011, it was the most dangerous country in tors for women in neighboring countries have the world for women.2 In 2015, UN Women improved. estimated that only 5 percent of Afghan busi- We believe there is a role for well-designed nesses are female owned. These dismal statis- policy to improve the status of women in tics have serious implications for the coun- Afghanistan.5 To date, there has not been a try—in terms of socioeconomic inclusivity, comprehensive analysis to study the issues poverty reduction, and for overall growth and Afghan women face to join the labor force. productivity, as female engagement in eco- The World Bank reports that policymakers nomic activity does not only benefit economic should “support agricultural value chains, development, but also promotes sustainable where women’s contribution is strongest and long-term growth.3 most valued, strengthen quotas and incen- A low FLFP in Afghanistan implies that tives for women to participate in the public there are other underlying factors that pre- sector, and target key barriers to women’s vent women from working. These include employment,”6 but there is limited clarity on limited mobility, gender inequality through what these barriers may be.7 In order to bring legal frameworks, lack of economic opportu- sustainable long-term improvement to FLFP, nities, and low household bargaining power. it is critical to determine the key barriers Evidence suggests that Afghan women are will- women face and use policy to effectively tar- ing to work. For example, in the 2015 Survey of get these barriers. Afghan People, 74.5 percent of Afghan women In our research, we aimed to, first, deter- state that women should be allowed to work mine the underlying barrier(s) that prevent

34 ∙ Women’s Policy Journal from entering the actions that favor women at the expense of labor force, and second, find effective actions men, possibly for a long time.”11 that the Government of Afghanistan (GoA) can take to affect reform and monitor prog- The Afghan Context ress in the Afghan context. The presented While there is a clear need for policy to analysis is a short excerpt of a full research enforce equality, promoting equality has been paper completed in March 2016 and can be challenging in countries with long-stand- found in the Harvard Kennedy School library ing traditions that sustain discrimination reserve.8 in all facets of life.12 Even with political will, these policies are often insufficient to change II. CONTEXTUALIZING THE PROBLEM deeply embedded gender stereotypes—such is Despite dramatic changes across the world in the reality for Afghanistan. women’s access to employment, education, and political participation, progress has been EVEN WITH POLITICAL WILL, uneven.9 We started by analyzing why this gap THESE POLICIES ARE OFTEN exists. INSUFFICIENT TO CHANGE DEEPLY EMBEDDED GENDER Is Government Policy a Part of the STEREOTYPES—SUCH IS THE Solution? REALITY FOR AFGHANISTAN. FLFP in developing countries is quite complex to study since there are many interrelated For over a decade, the international com- factors to consider. Researchers have high- munity has spent millions of dollars on top- lighted a few dimensions that affect women’s down programs in Afghanistan, yet Afghan ability to engage in the labor market: level of women engagement indicators remain economic development, educational attain- among the worst in the world. Some may even ment, social dimensions (e.g., norms influ- argue that these programs have provoked encing marriage, fertility, and women’s role Afghan men into taking more conservative outside the household), institutional setting positions.13 This is clearer as Bahri suggests (e.g., laws, protection, and benefits), access to that efforts to promote gender equality could credit and other inputs, and household/spou- be bolstered by including men in programs, sal characteristics.10 and addressing the issue through an Islamic There is strong evidence suggesting a need perspective as opposed to an ethnocentric for policy action. Policy instruments such as approach. In addition, he suggests that pro- top-down options like quotas, institutional grams to encourage women’s participation reform, and conditional transfer programs in society should focus on the benefits to the have been common, and have had some pos- family instead of economic reasons.14 itive outcomes. Even if these top-down policy There is some hope: evidence suggests that measures do not bring about radical changes certain policies can work to some degree, even in the way women are perceived in society, in environments where women are subjected evidence presented suggests that properly to very high levels of discrimination. A ran- designed policies targeted towards women domized experiment finds that gender quotas can have immediate consequences. Duflo in Afghan village councils improves outcomes takes an extreme position when she states specific to female participation in some areas, that “equity between men and women is only such as increased mobility and income gener- likely to be achieved by continuing policy ation; however, quotas produce no change in

Women in Post-Confict Governance ∙ 35 the short term to deeply entrenched attitudes influences FLFP. For women, the higher the towards the role of women in society.15 Along level of education is, the more likely a with the deeply entrenched cultural practices, participates in the labor force. On the other the heightened insecurity imposes further hand, men’s education has a reverse effect on constraints on female mobility in the Afghan FLFP. As average years of education for men at context.16 the district level increase, women within that district are less likely to enter the labor force, III. METHODOLOGY holding other factors constant. This suggests Participation of women in the labor force is that the quality and content of education mat- an outcome of various micro and macro-eco- ters: the fact that men in an area are educated nomic factors. To design effective policy, we does not imply greater mobility for women. need to understand how these factors interact Third, exposure to the outside world is an with FLFP. As part of our research, we devel- important indicator for FLFP. Watching tele- oped a robust predictive model to understand vision and obtaining information from com- the characteristics of the women in the labor munity shuras—councils held to discuss par- force and identify the major determinants of ticular issues including governance, security, women’s decision to enter the labor force in and development—is positively correlated Afghanistan. Using the 2015 Survey of Afghan with FLFP. This may imply that television and People, the model allowed us to study how community shura, as channels for receiving FLFP correlates with female characteristics, information, may provide space for policy instances of security and violence, and infor- interventions. In addition, evidence suggests mation sources, as well as men and women’s a causal link between the perception of wom- perceptions of gender roles. Even though it en’s mobility potential and women joining the does not directly provide causal linkages, the labor force. This not only includes the idea of model provides strong evidence of the rela- women being empowered and expected to tionships between our factors of interest and be mobile, but also men’s perception toward FLFP. Building on the predictive model, we women’s work. The model suggests that deployed a diagnostic framework to deter- women are more likely to be in the labor force mine the key barrier(s) that prevent women when men in the region are more supportive from entering the labor force and prioritize of equal education opportunities and are less policy reforms.17 strict on where women can work. Using the results of the model, we exam- IV. ANALYSIS AND FINDINGS ined key critical barriers: security, education, We start by acknowledging that the low FLFP caregiving, religion, norms, and the legal in Afghanistan is due to supply-side barri- framework. A detailed diagnosis (see the full ers, or factors that affect the supply of female research paper) of these barriers led us to labor, as well as demand-side barriers, factors conclude the following: affecting the demand for female labor. Due • Security is an underlying bar- to data limitations, we focused solely on sup- rier limiting women’s participation ply-side factors. in the economy. High violence not First, the model shows that security is a only diminishes women’s willing- common concern, negatively associated with ness to leave their home, but it also FLFP, across different ethnicities and dis- raises concern in men, which further tricts in Afghanistan. Second, men and wom- restricts women’s mobility. en’s significantly • Norms produce another key barrier

36 ∙ Women’s Policy Journal that limits FLFP. Prevailing norms, We recommended the GoA dedicate including men’s biased attitude resources to meet with relevant internal and towards the role of women in the external actors and stakeholders, and design family and in society, have resulted in a new regional strategy to alleviate the esca- a weak legal framework, low auton- lating security tensions. This is by no means omy, and low mobility for women. a simple task, but the approach to ensur- • A supporting environment for ing security needs to be revisited. However, women’s education is instrumental, given the tense security situation, the GoA especially men’s support for equal can work to encourage and promote home- education opportunities. Women based women’s work, such as handicrafts and self-report that limited or complete textiles. While this does not solve the mobility lack of education and professional issue, it allows women to work and sheds light trainings are among the largest chal- on an important industry in Afghanistan. lenges they face. • Information matters, and the channel Increasing Access to Information in which women receive information Through Television matters. Obtaining information from Second, we recommended policymakers the outside world provides exposure develop a television program, leveraging and can update women’s prior beliefs. channels like television media programming • Given the state’s capabilities, pro- and the increase in viewership to improve grams that seek to be effective have to access to information and exposure to the take both top-down and bottom-up outside world. This directly addresses norms, approaches. Afghanistan is a weak another main barrier affecting gender roles in state, but has a long history of strong Afghan society. informal ruling through community FLFP is highly correlated with various shuras and jirgas—ad-hoc meetings modes of information-sharing, and watching held to resolve disputes, which have television is one of the most positively cor- power to exercise local control. related indicators: 72 percent of women in the labor force obtain information via television, V. RECOMMENDATIONS whereas about 60 percent of those not in the labor force use television. Cable television is Security thought to be an effective form of persuasion In the Afghan context, security is a challeng- because people emulate what they perceive to ing issue and requires bold leadership from be desirable behaviors and attitudes; there- several actors in the international arena. Our fore, increased access to information through analysis led us to believe that security tensions television allows us to effectively target the arising from acts of violence are the main perceptions that prevent women from enter- underlying barrier preventing women from ing the labor force. entering the public space. Our main recom- From an administrative feasibility dimen- mendations did not delve into security as the sion, a television program would be sim- topic is beyond the scope of this analysis; how- ple and cost-effective. With roughly half of ever, we felt this is a first-order concern that the population with access to a television, will need to be addressed at a regional level to the long-term benefits of initiating this cen- ensure the empowerment not only of Afghan tralized program would potentially be pro- women, but the Afghan people in general. found as it can reach women in urban and

Women in Post-Confict Governance ∙ 37 rural areas at a low variable cost, despite the handicraft work, etc.; high up-front infrastructure cost. In addi- • Capture the sector of employment; tion, with evidence suggesting the interna- • Directly ask labor force involve- tional community has relatively high interest ment questions instead of indirectly in supporting a targeted women’s program, attempting to identify a source of international stakeholders can be engaged to income for female participants by provide financial assistance to a sound pro- asking, “In the last week, did you do gram design. any work for pay, for profit, or for From a political perspective, this program family gain?” would be disseminated like any other tele- vision program, without the explicit knowl- VI. CONCLUDING REMARKS edge of it being a women’s empowerment We know from literature that gender equality program. While religious scholars or opposi- matters for development, and we believe that tion groups may try to prevent women from greater female participation in the economy watching television, over time, as televisions is a major opportunity for the Afghanistan’s become widespread, it will be difficult for reconstruction. We acknowledge this will opposition groups to ban television viewer- require addressing obstacles at all levels, ship. Also, the goal of the program would be to including shifts in social customs, greater start a discourse, and watching the show is not accommodation of women in the market, and required for this to occur. Furthermore, the inclusive policy that addresses the key barri- program will be centrally managed and dis- ers that women face to join the labor force. seminated through local partner networks. Security is a critical barrier that dispropor- This would also minimize the political resis- tionately affects women—this issue needs to tance from religious scholars, who have the be addressed by the GoA in tandem with a strength to exercise local control. The central- program that aims to affect FLFP. Prevailing ized structure of the policy also limits the neg- norms are another underlying barrier that ative influence from provincial governments prevents women from entering the labor and threats from the . force, but we believe that a targeted interven- tion, such as increasing access to information, Timely and Comprehensive Data can help to lessen this barrier. Finally, we recommended the GoA revisit its In sum, while progress has been made, national surveys to enable the precise mea- there is a long way to go for women in surement of the magnitude of challenges facing Afghanistan. We hope the GoA can start a women. There are commendable data col- long-term dialogue to empower not only lection efforts in Afghanistan, but currently, women, but all of Afghanistan. datasets are missing a few critical elements with respect to FLFP. GoA could aim for the national survey to: • Include questions capturing the activity of all the women in the house- hold and not only married women; • Probe women who self-identify as to see if they partic- ipate in the informal labor market through agriculture, domestic labor,

38 ∙ Women’s Policy Journal NOTES 1 World Bank, World Development Indicators 2015 (Washington, DC: World Bank, 2015), https://openknowledge.worldbank.org/ bitstream/handle/10986/21634/9781464804403.pdf.

2 Richard Ghaisy et al., Afghanistan’s Private Sector: Status and Ways Forward (Stockholm: SIPRI, 2015), 38-42, https://www.sipri. org/sites/default/files/2016-04/SIPRI_Afghanistans-Private-Sec- tor_Report-2015.pdf.

3 Esther Duflo, “Women Empowerment and Economic Devel- opment,” Journal of Economic Literature 50, no. 4 (December 2012): 1051-1079.

4 Afghanistan in 2015: A Survey of the Afghan People (Washington, DC: Asia Foundation, 2015), https://asiafoundation.org/resources/ pdfs/Afghanistanin2015.pdf.

5 Ghaisy et al., Afghanistan’s Private Sector, 38-42.

6 World Bank, Afghanistan: Women’s Role in Afghanistan’s Future—Taking Stock of Achievements and Continued Challenges, report no. AC34474 (Washington DC: World Bank, 2013), http:// documents.worldbank.org/curated/en/813391468184173017/ Womens-role-in-Afghanistans-future-taking-stock-of-achieve- ments-and-continued-challenges.

7 World Bank, Afghanistan: Women’s Role in Afghanistan’s Future.

8 Ishani Desai and Li Li, “Analyzing Female Labor Force Par- ticipation in Afghanistan,” Second-Year Policy Analysis, Harvard Kennedy School, March 2016, https://wappp.hks.harvard.edu/ files/wappp/files/desai_li_sypa_final_0314.pdf.

9 World Bank, World Development Report 2012: Gender Equality and Development (Washington, DC: World Bank, 2012), https:// openknowledge.worldbank.org/handle/10986/4391.

10 Isis Gaddis and Stephan Klasen, “Economic Development, Structural Change, and Women’s Labor Force Participation: A reexamination of the feminization U hypothesis,” Journal of Popu- lation Economics 27, no. 3 (2014): 639-681.

11 Duflo, “Women Empowerment and Economic Development.”

12 Andrew Beath et al., “Empowering Women through Devel- opment Aid: Evidence from a Field Experiment in Afghanistan,” American Political Science Review 107, no. 3 (August 2013).

13 Javed Bahri, “Western Gender Policies in Afghanistan: Failing Women and Provoking Men,” Gender, Technology and Development 18, no. 2 (2014): 163-185.

14 Bahri, “Western Gender Policies in Afghanistan.”

15 Beath et al., “Empowering Women.”

16 Beath et al., “Empowering Women.”

17 Desai and Li, “Analyzing Female Labor Force Participation.”

Women in Post-Confict Governance ∙ 39