Report No: ACS25170 . Hashemite Kingdom of Jordan Understanding How Gender Norms in MNA Impact Female Employment Outcomes . June 1, 2018 . GSU11 MIDDLE EAST AND NORTH AFRICA . Standard Disclaimer: This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Copyright Statement: The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. MNA GSURR & eMBeD TTLs: Mariana T. Felicio & Varun Gauri Team Members: Iman Kalyan Sen, Tasmia Rahman, Anoud Allouzi, Samantha Constant EXECUTIVE SUMMARY 1. Do women in Jordan want to work? How do men feel about working women in their family? To what extent do personal beliefs and societal expectations influence a woman’s decision to work and why should this matter in development interventions? The Jordanian government and development partners have invested heavily in promoting women’s economic inclusion. However, Jordan has the lowest female labor force participation (FLFP) in the world of a country not at war. As development practitioners working on issues related to social and economic inclusion in the Middle East and North Africa (MNA) region, we ask ourselves these questions to help us understand binding constraints that prevent excluded groups, such as women and youth, from having equal opportunity to improve their quality of life. We also ask these questions to distinguish between our own and others’ perceived notion of inclusiveness. Building evidence from the field is key to enable development practitioners design more effective interventions to support female labor force participation. 2. The study’s1 objective was to measure the extent to which social norms and beliefs concerning gender influence women’s access to and participation in the labor market. It aimed at producing insights on what barriers women could be facing, and the extent to which individual beliefs versus social norms play a role. Results of this study contribute measurable evidence to gaps in social norms literature. There are many studies that have looked at structural and institutional barriers as well as identified social norms as a barrier to labor market participation, however many of these do not delve into mechanisms through which norms influence behavior nor do they explore intra-household dynamics in a measurable way. Similarly, others have talked about masculinity, without looking more closely at how it plays out at the household level. In contrast, the Jordan study also examines intra- household dynamics and how expectations of women’s spouse or male relative’s beliefs could be a constraint as well. 3. The study’s sample confirmed that 60% of non-working women actually want to work.2 The large majority of Jordanians are in favor of women working outside the home, and almost two-thirds of non-working women would like to work.3 Women’s preferences and personal beliefs are not a major obstacle to participation, therefore would likely respond favorably to policies that address some of their binding constraints. 4. The study also confirmed some of the most important binding constraints for women. These are: childcare, flexibility, part-time work, hiring and wage discrimination, limited job growth, lack of 1 The study was co-led and co-financed by GSURR MNA and eMBed teams. 2 More than 24% of these had in fact inquired about working within the last six months. Some 17% of non-working women had worked at some point in the past. Among all respondents, only 3% said they were opposed to women working; and only 15% were opposed to women working outside the home. 3 More than 60% of non-working women in our sample expressed a desire to work. More than 24% of these had in fact inquired about working within the last six months. Some 17% of non-working women had worked at some point in the past. Among all respondents, only 3% said they were opposed to women working; and only 15% were opposed to women working outside the home. adequate public transportation, particularly in rural areas, and scarcity of attractive job opportunities, lack of jobs in the public sector. 5. It is increasingly clear that intra-household expectations (i.e. expectations of respondents about their counterpart’s beliefs) matter in female labor force participation (FLFP) decisions. Both men and women agreed that men are the ultimate decision makers in the household; including deciding for women on whether to accept a job offer or not. More than 90% of women respondents said that, in the decision to work, the views of male household members (mostly husbands) are important or extremely important. Women seemed to prefer to seek advice from fathers, because they believed that they are more experienced.4 6. While women’s actions are more aligned with men’s preferences, women tend to underestimate the position of their counterparts on issues related to preference around women’s work. For example, working married women underestimate how disapproving people – particularly men – are of married working women returning home after 5pm. • 35% of women believe their counterparts will disapprove of them working in mixed-gender environments, however 60% of men admit they disapprove. • 42% of women believe their counterparts will disapprove of married women returning after 5pm; in reality 70% of men disapprove. • 42% of men believe that women prefer to stay at home to take care of children and household, versus 37% of women. • One of the most common reasons for not working cited by women is discouragement from the husband (19%); men slightly underestimate this (14%) • Occupational preferences of women are very different from what they actually do and tend to be more closely aligned to occupational preferences that male relatives hold for them.5 For example, a third of all working female respondents are currently employed in the education sector, however only 19% indicate these to be “ideal jobsâ€?. 7. Beliefs and expectations about women working in mixed gender environments pose certain barriers for women. This may be partially attributed to fears about harassment, as 44% respondents believe that working women are exposing themselves to harassment.6 One-third of respondents also believe that working women are risking their reputation by working. Together, these beliefs may not only restrict the types of jobs women can take up (i.e. jobs in mixed-gender workplaces), but also give rise to additional concerns about personal safety and reputation when deciding whether to work. Many working women (58%) believe that it is okay for women to work in environments where most other employees are men; non-working women are slightly more conservative (46%); and men with non- working counterparts are even more conservative (22%). 4 Only 25% agreed that a woman has the right to work if it makes the men in the household uncomfortable. 5 In contrast, 30% of male respondents find jobs in the education sector to be ideal for their female relatives. Similarly, while only 8% of working women are in the sale and service industry, 14% consider jobs in this sector to be ideal, compared to 8% of men. 6 The estimate is slightly lower for working women and their male counterparts (34%-36%), and higher for non- working women and their counterparts (around 47%). 8. Men seem to underestimate the demands placed on women once they are married. For women, getting married is one of the key reasons for exiting the labor market. According to men, child bearing is the key factor for women to leave the labor market. Nevertheless, the main reason expressed during focus group discussions for women not working or disengaging from the work force was due to having children. The long working hours made it difficult for some of the women in the discussion to juggle work, taking care of a newborn, and household responsibilities. Surprisingly, working women returning after 5pm seems to be a binding constraint: only 26% of men and women think it is okay for working women to return after 5pm. Societal expectations are that women should take care of all the household chores (cooking, cleaning, caring for children, taking them to school, etc.). Men’s contribution to household chores is negligible, regardless of women’s work status, (except when it comes to feeding children).7 9. Having younger children can restrict labor force participation for married women as well. According to respondents, a child should be at least 4.5 years old before the mother can leave him or her at daycare (or elsewhere) and go to work. Working women think the appropriate age in around 3.4 years on average, while for men with non-working spouses, the average is even higher at 5.2 years. 10. A key driver for women’s employment in Jordan is financial need. Whether the question was about the circumstances that would make women consider working or lead them to joining the workforce, the majority of participants answered that the high cost of living would force them to find a job. This was highlighted mostly by women participants in Amman, who believed that it is their husbands responsibility to provide for the family, not their responsibility. 11. Irrespective of work status, but much more so among non-working female respondents, women’s financial independence is relatively low in our sample. Seventy-six percent of women do not have bank accounts: this includes 44% of working women and 90% of non-working women. Over 70% of women, irrespective of work status, do receive monthly allowances from their spouses for personal or household/children’s expenses, though not all receive allowances for both. Greater access to assets, financial inclusion and autonomy are key aspects of women’s voice and agency, and as such 7 Share of household chores performed by men, according to women, and similar to men’s reporting: taking children to school (16%); cleaning/laundry (12%), cooking/dishes (21%), and feeding children (41%). increase a woman’s bargaining power and positively influence intra-household dynamics to her advantage. 12. Overall, men’s views are more conservative than women with regards to labor force participation. However, there also seems to be some relationship between women’s and men’s views being more liberal when women work. When men’s spouses are working, they are more liberal about women working, women working outside the home, working in mixed environments, returning after 5pm and leaving their children with relatives. This is also true for women; however, it is still unclear if there is a causal effect. None working men had the most conservative views related to preference around women’s work in our sample. 13. Jordanians overestimate the percentage of working women in Jordan to be 70%, when in fact, the actual percentage of working women is only 14%. Surprisingly, when participants were asked about percentage of women working, the majority of those interviewed gave a very high percentage of 70% on average. Societal norms and expectations in this case do not seem to be the defining barriers for women’s low level of economic participation in Jordan. In addition, respondents highlighted the changes in society that led to the change of women’s role but changes as negative due to their concern that such changes will lead to women changing their behavior and becoming “arrogantâ€? or “bossyâ€?. 14. Role models for women, specifically non-working women, were mostly family members. When women in the focus group were asked the reason behind their choices, the answers focused mostly on the hardships that these female family members had to overcome to successfully manage the family and the household. Success for women was described as “women who work and are able to manage their householdsâ€?; women’s participation in the labor force cannot be untied to the nee d to continue to take care of household chores and children. For men, their role models were male family members due to their ability to provide for the house and certain characteristics such as honesty. 15. In terms of aspirations for the next generation and advice for daughters, gaining a good education was the most repeated advice during the FDGs. The reasons for achieving a good education varied however depending on who was asked: for older women, the key motivator was for women to have virtuous qualities, while younger women were more outward looking highlighting their wish for their daughters to achieve their goals and dreams and gain the skills needed for them. On the other hand, non-working men were more conservative in their responses, and focused on wanting women to “know their boundariesâ€?; given that men are the providers. Similar trends were observed in the quantitative data- while women see education as a means to improve employment opportunities, financial independence, and confidence for themselves, men also value education (for women) for its positive impact on child rearing and improving the family status. 16. Together, these results suggest that social norms, particularly expectations of what others do, are important influencers of FLFP when it comes to key areas of decision-making, such as general acceptability of women working and gender roles of married women. Household status and publicness and gender-mixing as they relate to women working, however, seem to be more internal, and is influenced primarily by intra-household beliefs and expectations. Lastly, people’s own beliefs are important predictors of work across everything, except when asked about status. 17. At the intrahousehold level, men and women are not very different in their perceptions of each other when it comes to women working, though working women tend to have more liberal expectations than their male counterparts. Non-working women expect their male counterparts to have much more conservative views about mixing with men and being outside of the home for work and vice versa. While women do not perceive marriage to be a major issue for their counterparts when it comes women’s work decisions, long hours and having young children are expected to restrict their approval of working women. 18. A major discrepancy in expectation and reality is observed when it comes to estimates around FLFP. Even though female labor participation is only 14%, on average, all respondent groups estimate it to be closer to 50%. This is a huge overestimation, and is puzzling given what we know to be the case in Jordan. On the other hand, social normative expectations about whether other people are okay with women working or married women working suggest that respondents, particularly non-working women and men with non-working counterparts, overestimate the level of conservatism around them. Respondents, however, underestimate the extent to which people disapprove of married working women returning home after 5 PM. 19. The following table summarizes the views and expectations respondents have of society and their own personal views regarding FLFP. Summary of Findings: Personal Beliefs and Social Norms Social Empirical Expectations Personal Beliefs Social Normative Expectations Themes Estimations of what others do* What respondents believe people should do Expectations of what others think people should do** Women working Both men and women highly overestimate the Over 95% believe it is okay for women to On average, respondents expect a third of (should women work; level of female labor force participation work, though slightly fewer believe it okay to other people to find it unacceptable for should they work outside (within and outside the home), both within work outside the home- lowest for men with women to work, both in an outside the the home; male relatives their communities and across Jordan, non-working (NW) counterparts (63%). home. Very few people (2/10) are not comfortable) (estimates are close to 50% or 5/10). Working is believed to be contingent on how expected to find it acceptable for a comfortable husbands/fathers are with their woman to work if it makes her wives/daughters working. husband/father uncomfortable. Publicness & mixing Working women (WW) are generally not Only 38% believe it is okay for women to 5/10 people are expected to find it (mixed-gender work estimated to work in male-dominated work in mixed-gender workplaces, with men unacceptable for women to work in mixed environments; workplaces (3/10). They are estimated to being more conservative than women. 44% gender environments. Perceptions about harassment; reputation) experience some harassment (2.6/10), but believe that working women are exposing harassment and reputational risk are also their reputation is not estimated to be themselves to harassment, and 35% believe not favorable: 4/10 are expected to think affected by their decision to work. that WW are risking their reputation by that WW are exposing themselves to working. harassment, and over a third are expected to think that women risk their reputation by working. Gender roles Estimations of married working women is Marriage lowers acceptability of women Marriage does not change expectations of (married women working; lower (4/10). 8/10 are estimated to return working (72% believe it is okay), especially for the acceptability of women working, but returning home after 5 PM; home by 5 PM. Accordingly to estimations, male respondents. Work hours and younger 5/10 and 4/10 people are expected to find leaving children with working mothers do not leave their child and children pose barriers. Less than 40% believe it unacceptable for married WW to return relatives; appropriate age go to work until child is at least 2, with men it is okay for married WW to return home home after 5 PM and leave children with to leave child) having slightly higher estimates. A quarter of after 5 PM. 54% believe it is okay to leave relatives, respectively. Appropriate age to working mothers are estimated to leave child child with relative, but they believe that the leave child, according to others, is with relatives. child needs to be at least 4.5 years old before expected to be 3.5-4 years (for women) mother can work. and around 4.5 years (for men). Status of family Families with working women are estimated to Aside from financial status, women’s decision 75% expect most other people to think (financial need; husband have lower financial status (84%). 52% to work has little influence on what that WW’s families have financial need. not in charge; husband estimate that some or most of these women respondents believe about the status of their Around a quarter expect most people to unable to provide; less have husbands who cannot provide for them families and husbands. 70% percent believe think of WW’s families as less traditional traditional) and 31% estimate some or most to have families of WW have financial need. and their husband’s to not be in charge or husbands not in charge. Some third estimate not be able to provide (32%). Women households with WW to be less traditional. have slightly more conservative expectations of what others think, relative to counterparts. Men with NW counterparts have the most Men, especially those with NW counterparts, Relatively less variation observed between conservative empirical expectations of have much more conservative and patriarchal the four subgroups. Across most Who is most working women in their communities. NW beliefs on issues related to women’s labor responses, NW women generally expect conservative? women have more conservative estimates force participation compared to women. others in their reference groups to have when it comes to publicness & mixing. the most conservative beliefs. Men with working counterparts expect the least conservative beliefs. * Question asks how many women out of 10, where respondent lives, do these/fit this description (e.g. 2/10 means that on avg, 2/10 women are expected to do this). For status, respondents (only women) were asked how many WW- some, most, few, none- fall into this category. ** Question asks how many women out of 10, where respondent lives, think or speak badly about women who do these/think it is okay for women to do these. For status questions, respondents were asked if most people think this about working women and their families. Methodology 20. This is the region’s first ever systematic, representative, quantitative study of social norms (across four areas of factual beliefs, personal normative beliefs, empirical expectations and normative expectations) at the national level, and perhaps the first in developing countries. It provides new techniques for measuring and validating social norms related to gender, a model that could be applied to other countries, as well as innovative mixed methods research, including the analysis of narratives. 21. The study developed techniques to measure the extent to which social norms and beliefs concerning gender influence women’s access to and participation in the labor market. The measurement of social norms requires questions carefully targeted at a) personal beliefs, preferences, and behaviors; b) expectations regarding how others behave (social empirical expectations); c) expectations regarding the normative beliefs of others and their sanctioning behavior (social normative expectations). In addition to these, we also target questions at the mutual expectations of male-female pairs in the household (intra-household expectations). We examined normative influences across four key dimensions of FLFP: i) general views on whether women should work; ii) beliefs about gender roles as they relate to women’s ability and availability to work; iii) beliefs about publicness and mixing of genders in the work environment; and iv) beliefs about the link between women working and family status. 22. Five key activities were carried out: a qualitative round comprising of 14 focus group discussions (FGDs) in Amman and Mafraq, pre-test and full rollout of the quantitative social norms survey in Amman, Mafraq and Zarqa, a lab-in-the-field experiment conducted concurrently with the survey, and, lastly, a quantitative follow-up phone survey to assess if responses to social norms and beliefs related questions stay consistent over time. The quantitative survey was conducted with a sample of 2007 respondents in urban areas of Amman, Zarqa, and Mafraq. The sample was split evenly between male (any age) and female (working age, i.e. 20-55 years) respondents, with female respondents further split by their work status. 23. Since the study focuses on beliefs and norms around women’s labor force participation, it was necessary to include sufficient working women in the sample in order to learn more about them and gage how they are different from their non-working counterparts. Given the low rate of female labor force participation in Jordan, working women needed to be deliberately oversampled at each location to obtain a sufficiently large sample size for comparison and analysis. 24. To include at least 30% working women, the sampling methodology was a stratified cluster sample based off an enumeration exercise. With Amman, comprising a very large proportion of the frame (70%), the governorate was under-sampled by 15 clusters which were transferred to the Mafraq strata. Localities within each governorate served as the primary sampling unit; grids within each locality as the secondary sampling unit; and within each randomly selected grid, census blocks acted as the tertiary sampling unit. Census blocks were enumerated to locate working women, and priority of selection was given to working women within each household. A stratified (by work status and gender) simple random sample was drawn from eligible households (fourth sampling unit) in order to oversample working women. Syrian Refugees 25. Despite efforts to examine norms among Syrian refugees as well as Jordanians, the team shifted direction mid-course to focus only on host communities. The original intent of the study was to measure social norms and beliefs that influence both Jordanian and Syrian refugee women’s labor force participation. The goal was to eventually use the findings to improve their employment opportunities in Jordan. Examining behavior as well as how shifting power structures in times of crisis influence decision making is critical to ease constraints facing women and girls in their pursuit for quality education and decent work. As such, a focus on Jordanian as well as Syrian refugees could bring to light the impact of conflict on traditional gender roles among refugees and then spill-over effects to host communities. However, despite the intended focus, the component of incorporating refugees in the study was removed after qualitative field work targeting Syrian refugee women and men revealed that day-to-day humanitarian challenges faced by refugees far outweighed broader normative issues. While such issues still need to be addressed, it ought to be done with more time allotted and through a more tailored fragility and conflict lens. Recommendations 26. What can the World Bank do to be more effective in supporting female labor force participation? What has this study taught us and how can we support the GoJ and sectoral teams to design interventions that have meaningful and sustainable impact? If the GoJ’s is fully committed to promote female economic empowerment, then it is necessary that proposed structural policy changes be accompanied by efforts that contribute to strengthening women’s bargaining position with their male counterparts, and invest in ‘changing mindsets’ activities. 27. Furthermore, packaging the study’s key findings and recommendations will be equally important. Understanding the conceptual underpinnings of measuring social norms work is complex therefore it is important that the study be translated into an easily digestible format. Based on lessons from similar studies, it is recommended that the study be converted into a short brief, an infographic, and a 15-20 pages publication to be translated into Arabic for wider dissemination and reach. 28. This section provides three-tiered recommendations specific to Jordan structured around: (i) policy level (for both the World Bank teams and the GoJ); (ii) sector-specific recommendations, (including working with IFC and Role of media campaign interventions) coupled with (iii) broader recommendations to be considered for other countries in the MNA region. Policy-level Recommendations 29. The policy level recommendations are meant to address structural barriers impeding women’s active labor force participation and to do so within the context of social norms. While many of these were acknowledged prior to the study, the study has confirmed that in fact, these factors do 2 contribute more significantly than expected in preventing women’s participation. The study has however provided a greater depth to some of these issues (ie., childcare: it is insufficient to provide affordable and reliable childcares if not accompanied by efforts to change stigma around mothers who leave their children in childcare before the age of 4.5). • Childcare: Having younger children can restrict labor force participation for married women as well. According to respondents, a child should be at least 4.5 years old before the mother can leave him or her at daycare (or elsewhere) and go to work. Working women think the appropriate age in around 3.4 years on average, while for men with non-working spouses, the average is even higher at 5.2 years. A study by Sadaqa demonstrated that vouchers for childcare were ineffective and that what was important for working women was proximity to their kids. Recommendation: increase access to affordable child care by complying with article 72 in the labor law to establish day care centers at the workplace and include men in the number of employees counted for the establishment of a daycare center at the workplace. Support the work of local organizations such as Sadaqa on achieving this goal. Currently, if a company has 19 women or more with children under the age of four, the company has to establish daycare center. Subsequent tranches of the First Equitable Growth and Job Creation Programmatic Development Policy Financing (DPF) could require an amendment to this law that is gender neutral, (ie. Require that 19 employees who have children, rather than women). • Moreover, additional information and role models regarding how women currently handle work and children care duties could also help. Child care is important for working women, but people do not appear to understand that Jordanian working women are able to overcome this hurdle nor how they manage to do so. Most women leave their children with close relatives, but most people think that working women are most likely to leave their children in daycare. Most respondents, men and women, believe that working women should return to work, and do return to work, after their children are 4.5 years old. These restrictive personal and social normative beliefs can potentially delay a woman’s re-entry into the labor force (if she was already working), and can do so to a point where she does not return to work at all. Recommendation: Provide community based peer to peer mentorship of working or prospective working mothers within a network of family and neighbors/community members to learn from each other how to best manage the double burden, share resources and information about quality day care services, best commutes, provide support and create new role model especially for young women. • Family Friendly Policies: The most strongly reinforced social norm in the study was not opposition to women working outside the home, but that women should return home by 5 pm. Personal beliefs, social normative expectations, social empirical expectations, and intrahousehold expectations all mutually support the idea that working women should be home by 5 pm. The DPF introduces a Flexible Work System which is likely to support women’s economic empowerment. The Borrower’s Council of Ministers has approved a Bylaw on Flexible Work that benchmarks the minimum wage rate by unit of hours. The DPF is expecting an increase in growth of formal, private, part-time female workers from 31.6 percent (2014-2016) to 35 percent during (2018-2019). Recommendation: Subsequent tranches could further advance flexibility and part- time arrangements. It could introduce measures that facilitate and encourage employers - 3 particularly the private sector – to create jobs that can be done from home, such as technology- based activities which could be developed and ideal for working women. Communication about the Flexible Work System would be important to ensure that workers and prospective workers are aware of these options and minimum wage requirements, and that employers include clear and gender-sensitive guidelines about flexible arrangements in their employment policies. • Information About Jobs Available: Information about women’s work appears to be inaccurate, leaving people with uncertainty about where women should look for work and how to go about the process. For instance, respondents vastly over-estimated the share of women and married women who are working. We interpret this to be an awareness on the part of respondents that “things are changingâ€? in Jordan. But this finding also shows Jordanian households do not have specific information regarding women’s labor force participation, and are instead relying on vague generalities when thinking and talking about women’s work options. Moreover, although 72% of women in our sample worked in the private sector, and less than 1/3 worked in traditional fields for women (teaching or nursing), respondents thought that only about 30% of women work in mixed gender environments. Recommendation: Develop plans to support both public and private sector entities to strengthen information availability and dissemination about share of women in the labor force and employed in the private sector, types of jobs available and composition of employees, etc. Sector-specific Recommendations • Education Sector: 31% of men associate women education with improved family status and better parenting. Changing the narrative to associate education with better careers that would result in improving family status and better parenting could shift perceptions (see social media campaigns section). Education and sustainable job creation linkages have to be better articulated through World Bank operations. o Female teachers face wage discrimination, discouraging them to work. We also know that employers have the practice of paying teachers in cash, as many do not have bank accounts. The Ministry of Education, through the Education Reform Support Program for Results could provide incentives for female teachers to open bank accounts. o Efforts to change mindsets about gender roles and stereotypes that are currently being propagated through the curricula and teaching methods employed by the Ministry of Education could be supported. Invest in changing the teaching materials and in building capacity on gender sensitive approaches to teaching. Whenever possible, work through parent teacher associations (PTA)s, targeting fathers in particular, to support girls’ education and its benefits that go beyond being more “eligible for marriageâ€?. Recommendation: change the narrative to associate education with better careers that would ultimately result in improved family status and better parenting. • Social Protection & Jobs: In terms of job opportunities, propose and integrate solutions and global lessons from the IFC into job creation that fosters an environment that can sustain women’s productive engagement in the long-term, (not just short-term jobs, such as public works). In terms of social protection, international best practices show that giving benefits to women will 4 result in better outcomes – particularly for children and may contribute to women’s economic empowerment. Through the DPF, the GoJ has approved a Plan to improve and expand the coverage of the National Aid Fund (NAF) cash transfer program to at least 85,000 additional households in 2019 and 2020 (42,000, and 43,000 households in 2019, and 2020 respectively). The program gives preferences to female beneficiaries e.g. prioritizing women-headed households, families with children, and working poor (including working women) and can extend its coverage in subsequent tranches. Recommendation: For future tranches, learning from other DPF’s such as Brazil’s Development Policy targeting vulnerable groups Loan Bahia Strengthening Fiscal Management and Promoting Better and More Inclusive Service Delivery (DPL) may be considered. http://www.worldbank.org/en/results/2017/12/04/improving-access-public- services-vulnerable-groups-bahia-brazil. Lastly, the pipeline operation Safety Nets and Youth Employment Project may also be a venue to promote these approaches. • Transport Sector: Affordable and reliable transportation would contribute to women’s labor force participation however it is not enough. The majority (82%) of women are willing to take public transportation; however, women believe that the longest acceptable commute time is 32 minutes on average. Results from the survey also show that concerns around harassment can be prohibitive: a large share of respondents (44%) think that working women expose themselves to harassment either commuting to and from work or at work. Recommendation: Design a public transportation system that facilitates safe and reliable transportations needs for women and can provide shorter commutes. Concerns about harassment could be accompanied by efforts, such as that supported in Mexico through a World Bank financed pilot project, “Hazme el Paroâ€? (Have My Back), which empowered 46 male bus drivers to question ‘machista’ attitudes around harassment that had become normalized. The project gave drivers the tools to detect different types of harassment and respond unwaveringly fast according to an agreed protocol. • Financial Inclusion of Women: Provide global best practices to the GoJ on regulatory frameworks and procedures to facilitate access of women to bank accounts. Over 90% of women in our sample, and 46% of working women, did not possess their own bank account. The working women who do not possess a bank account are likely being paid in cash. Not having a bank account may limit the kinds of jobs they pursue and are able to obtain. At the same time, status concerns are an issue for Jordanian households. Almost 70% believe that women work because of financial need (not to make something out of their lives or some other reason), more than half estimate some or most of working women have husbands who cannot provide, and about a third estimate that households with working women have husbands who are not in charge. Several projects in the financial sector could support some of these initiatives. Possible entry points may include: (i) MSME Development Project for Inclusive Growth, (ii) Innovative Start Up Fund Project, (iii) Strengthening the Regulatory and Institutional Framework for MSME Development, and (iv) Promoting Financial Inclusion Policies in Jordan. • Infrastructure (Energy and Extractives, Water) and Urban/Municipal Planning: Draw from the results of the survey to generate a deeper understanding of how intrahousehold dynamics affects decision-making processes around women’s participation in various economic activities related 5 to infrastructure. This will ultimately promote cross-sectoral linkages with infrastructure, calling attention for the need to identify gender entry points that incorporate a family centered approach when engaging women in activities related to hard sectors and project planning. • Cross Cutting Themes (Gender, Social Development, Communications): Leverage social development and gender entry points in all operations to integrate awareness raising campaigns that can accompany proposed structural (policy) and design (project level) issues, and that specifically target constraints highlighted in report and policy recommendations section. Working with the International Finance Corporation (IFC) & Private Sector 30. Building on IFC’s ongoing efforts to strengthen opportunities for young women and men in private employment while leveraging its regional experience on financial inclusion can open doors for stronger private sector engagement in World Bank operational projects. • On employment: Linking to and learning from IFC’s education for employment program in Jordan could be an important entry point to widen the benefits of market-relevant programs especially to women. This can be done by exploring the potential of incorporating behavioral aspects to skills building or employability programs as well as keeping in mind normative constraints revealed through the study such as on travel time, mixed workplaces (or training places, expected time to be home, etc. What can be implemented to avoid additional burdens on companies? How can we ensure that the companies don’t end up refusing to hire women because they have to be flexible? • A second entry point is through a focus on financial inclusion: According to results of the survey, 76 percent of female respondents did not have bank accounts with a large gap between working women (44 percent) and non-working women (90 percent). Lack of access to and control of assets can be a major barrier for women’s economic participation, especially in countries where women are not granted equal inheritance and land rights. Such inequities also reduce women’s bargaining power in the home, which ultimately impacts her voice and agency with regard to work. Working with IFC on exploring the potential of adapting its financial inclusion programs, such as the Banking on Women program in Lebanon and the Palestinian territories, can provide greater opportunities for women especially entrepreneurs to access financial services. It can also enable banks to rethink their client market. For example, the Bank of Palestine (BoP) in partnership with the IFC has made a concerted effort to changing its banking culture across the Palestinian territories by training its BoP staff on gender-responsive customer service to ensure financial information and resources are shared equally with both female and male clients. The BoP has also been working on bank-wide policy changes that would overcome gender-based legal restrictions to financial access, such as allowing women to open an account for their children without a male guardian consent. 8 The potential of such a program could be explored in Jordan, within the context of strengthening financial access to women and strengthening their household bargaining power. Strengthening access to women and girls, and incorporating financial literacy in early 8 Hillis, Samira Ahmed; Constant, Samantha M. 2018. Second Country Gender Action Plan (C-GAP II) for Palestinian Territories (FY2018-2021) (English). Washington, D.C.: World Bank Group. 6 education may also help increase women’s agency in financial inclusion. Tax incentives to companies that provide flexible working hours for women could also be explored. Role of media campaign interventions 31. The GoJ could implement a multi-pronged effort that invests in shifting mindsets of men, women, and boys from an early age. For example, actual male beliefs are more liberal than men imagine them to be. This opens the door to information campaigns and narratives that change their understanding of what others believe. For instance, if mixed gender work environments were normalized through discussion, discourse, and role models (e.g., educational attainment could serve this role), men would learn that other men know about these environments, and become less concerned about them. For example, families of working women are not perceived as less traditional nor are they accused of the husband’s inability to provide for the familyâ€?. Only 19% of respondents associate married women working with their husband’s inability to provide for her (slightly higher for non-working women and their male counterparts); and less than 10% believe that the husband is not in charge in families with working women (slightly higher for women compared to men). However, the false narrative of the statements above make Jordanians believe that most people hold such views. 32. There is evidence that media campaigns, entertainment education, and aspirational interventions can, under the right circumstances, change behavior (World Bank 2014). Following are some media campaigns that could be implemented to encourage higher female labor force participation in Jordan: • Aspirational videos: Aspirations are shaped by exposure to and observations of outcomes of individuals whom one can observe and identify with. Aspirational videos, featuring working women from relatable backgrounds who have managed to find a way around structural and intra- household barriers, can be shown to raise Jordanian women’s aspirations and broaden the scope of livelihood options they consider viable for themselves. Aspirational videos shown in rural Ethiopia, featuring Ethiopians from similar backgrounds who were successful in agriculture and business on their own, were more effective at raising villagers’ aspirations when people in the community were exposed to the content (Bernard and others 2014). • Media campaigns targeting men: Given the critical role men play in women’s labor force participation decisions, media interventions need to focus on and target men as well. More than 90% of women respondents said that, in the decision to work, the views of male household members (mostly husbands) are important or extremely important. Only 25% agreed that a woman has the right to work if it makes the men in the household uncomfortable. For example, a Promundo type campaign can be developed to engage men as active partners in women’s empowerment and transform harmful gender attitudes and norms and impede women’s labor force participation. Advocacy campaigns, community outreach, and workshops targeting men and boys in India, Brazil and Chile lead to statistically significant changes in attitudes towards gender- based violence, and decrease in intimate partner violence (Instituto Promundo, 2012). 7 • Soap operas: Story-telling and media exposure can have a powerful influence on real life behavior. Soap operas can be used to deliver entertainment education across a society to change prevailing beliefs and mental models around the role of women in society and within the household. For example, a soap opera could be produced in Jordan with a storyline featuring a working woman, her family, and their struggles and experiences, to normalize the concept of families with working women and potentially even provide a role model for women. Soap operas could address the disconnect between what Jordanian men observe and what they assume other men think and believe regarding women working in mixed gender environments (40% believe it is sometimes okay, men estimate that half of all other husbands would not be okay if their wives worked in mixed gender settings). In Brazil, access to the TV Globo network—which was dominated by soap operas with independent female characters with few, or even no children—has been linked to the country’s rapid drop in fertility. Viewing the soap operas had an effect equal to 1.6 years of additional education (La Ferrara, Chong, and Duryea 2012). 33. Approaches like these could be used to stimulate women’s labor market participation in Jordan. Soap operas, aspirational videos, social media campaigns, or a combination, could be used to broadcast information about how and where to find work, correct misperceptions regarding aspects of Jordanian society that estimated to be more restrictive than they really are, empower women by lifting aspirations, and provide new mental models and role models for family life. Recommendations for other MNA Countries 34. The research study in Jordan both confirmed and demystified some of the constraints believed were preventing women’s labor force participation. It demonstrated that, in the case of Jordan, personal beliefs and intra-household bargaining are more influential than societal norms and expectations, though expectations of what others do can also be influential. While we cannot ascertain that is the case in other MNA countries, some lessons from this study that could be taken into account for others are: • It is important that Project Teams that aim to promote favorable labor market policies are cognizant that proposing changes to structural issues alone may not be sufficient to achieve the expected outcomes. It is paramount that they be coupled with “softâ€? interventions such as those that can influence intra-household decision-making. • Linked to this, employability and skills building (i.e. financial inclusion, vocational, etc) programs that target women and youth, especially young women, should also consider a behavioral component to help women navigate intrahousehold dynamics that may disadvantage them from fully benefiting from such programs. This component could also be an opportunity to engage men and boys in order to ensure male members of a household are not alienated during the process of women’s economic empowerment. • Invest in interventions that support the changing of attitudes towards shared household responsibilities, (chores, etc.) emphasizing messages that it is ‘manly’ to help around the house. • Support activities – wherever possible – to enable women to open their own bank accounts and gain financial literacy. Women’s economic empowerment is core to the WBG’s and MNA specific 8 gender strategies, and this should be promoted whenever possible (via MFM, PSD but not limited to these sectors). • Invest in interventions that encourage women to aspire for a career path beyond obtaining a college degree. Focus on messages highlighting women with career paths in various sectors and the impact of having careers and two incomes on the living standards of their homes and the aspirations of their children. • When promoting women’s labor force participation, consider integrating global lessons to promote longer-term employment and job growth for women. What processes, procedures, laws, and/or incentives can be included to encourage women to stay in the job, so that they can envision professional growth opportunities? • When promoting female job creation is one of the project’s objectives, ensure that flexible jobs (flexibility in hours and location) is part of the design. For example, a positive – unintended – externality of the Egypt Inclusive Housing Program for Results, was that the Program produced home-based jobs for women through online loan application processes and responding to grievances and queries. 9 Table of Contents EXECUTIVE SUMMARY .......................................................................................................................................... 4 INTRODUCTION ................................................................................................................................................... 11 1.1. Background .............................................................................................................................................. 12 1.2. Barriers to Women’s Labor Force Participation....................................................................................... 17 1.3. How are we different? ............................................................................................................................. 23 SOCIAL NORMS ................................................................................................................................................... 25 2. What are Social Norms?.............................................................................................................................. 26 METHODOLOGY .................................................................................................................................................. 30 3.1. Sequence of Activities .............................................................................................................................. 31 3.2. Qualitative Findings ................................................................................................................................. 34 3.3. Quantitative Survey Instrument .............................................................................................................. 42 3.4. Sampling Methodology ............................................................................................................................ 46 RESULTS .............................................................................................................................................................. 50 4.1. Sample: Who are our respondents? ........................................................................................................ 51 4.2 Results ....................................................................................................................................................... 53 4.3 Women’s preferences and decision making towards work...................................................................... 81 RECOMMENDATIONS ......................................................................................................................................... 92 5. Recommendations ...................................................................................................................................... 93 5.1. Policy-level Recommendations ................................................................................................................ 93 5.2. Sector-specific Recommendations........................................................................................................... 94 5.3. Recommendations for other MNA Countries .......................................................................................... 98 10 Jordan: Understanding How Social Norms in MNA Impact Employment Outcomes (P161633) “Being capable and independent financially makes a woman bossy.â€? – Nonworking Jordanian woman, age 31-59, focus group discussant “If my wife has a job and she helps me out with the house expenses then this is a positive thing since life circumstances are becoming very difficult. However, there is only one negative aspect to it and that is if there is any differentiation between the husband and wife and what they bring to the house – saying that this is from her and she is the provider of the family, for example.â€? -- Nonworking Jordanian male (age 22-40), focus group discussant “My daughter worked at a pharmacy here in Mafraq and when she had her [son] she could not leave him. She told the pharmacy owner and she began taking her child with her.â€? – Working Jordanian female, age 31-59, focus group discussant INTRODUCTION Building evidence from the field is key to designing smart “I will advise any woman to work interventions. Do women want to work? How do men feel about for the sake of herself and also to working women in their family? To what extent do personal beliefs be productive because it is very and societal expectations influence a woman’s decision to work and important to be productive in order to become an active why should this matter in development interventions? As member in society.â€? development practitioners working on issues related to social and economic inclusion in the Middle East and North Africa (MNA) – Working Jordanian female, age 31-59, region, we ask ourselves these questions to help us understand focus group discussant binding constraints that prevent excluded groups, such as women and youth, from having equal opportunity to improve their quality of life. We also ask these questions to distinguish between our own and others’ perceived notion of inclusiveness. How do we define inclusion and empowerment? How do others undergoing daily realities in a specific context define it? Does their context affect their behavior? Are their beliefs and behavior influenced by underlying cultural and social norms, and if so, how, and at what level- household or societal? Answers to these questions matter because interventions are more effective when they are designed in a way that resonate with people’s needs and understanding of the world they live in. There is a general understanding that restrictive attitudes toward gender equality are at play in MNA but a limited sense of how this is impacting development outcomes. There is a growing body of evidence around perceptions and attitudes of women and men in MNA about how they feel about issues related to personal status, education, access to work and overall life challenges. However, there is little evidence from the region that shows the extent to which one’s own beliefs and decisions around female employment are influenced by what others within their household or community believe or is perceived 11 to believe. There is also little experimentation to show how belief systems translate to actual behavior. In fact, while the role of norms is a growing topic in much of the development discourse to support women and their participation in the formal labor force, there are very few MNA country or regional studies that measure norms in a rigorous way. Why has promoting economic empowerment, especially among women, become such a key focus for the World Bank in the MNA? While variations exist across the region, participation in the formal labor force is particularly low in MNA averaging 22 percent – a figure substantially lower than all other regions. When one half of the population is not engaged in the labor force, countries lose out, and families struggle harder to make ends meet. On the social side, it has been proven that when women work, there are benefits that pass down to children. For example, studies have shown that increases in maternal income impacts a girl’s schooling much more so than increases in paternal income. 9 On the economic side, statistics show that the region loses up to USD 575 billion in regional income due to the current levels of gender-based discrimination in laws, social norms and practices that constrain women’s rights and opportunities.10 Increasing female labor-force participation to the levels of men could boost regional GDP by 47 percent.11 The study is divided into five chapters. This first chapter introduces Jordan’s socio-economic context and provides a situational analysis of gender gaps in the labor market, constraints facing Jordanian women’s economic participation and the motivation behind the study. The second chapter delves into the theory behind measuring social norms and review of its literature. Subsequently, the report presents the methodology covering sequence of activities, qualitative findings, quantitative instruments, pilot findings and adjustments. Integrated into this chapter and the next is the description of the field experiments and some of the preliminary results emerging from the calls to the Job Advisor and the call-backs. The fifth chapter presents a snapshot of the results of the survey. Finally, the research concludes with additional research areas to look into using the data from the field and proposes recommendations on policy reform and interventions, taking into consideration the cross-cutting nature of behavioral study in all dimensions of development work. 1.1. Background Low levels of engagement among women in MNA’s labor force vary from country to country yet female unemployment is consistently high across the board. Female labor force participation is as high as 58 percent in Qatar, 49 in Djibouti, 47 in Kuwait and as low as 14 percent in Jordan, 12 percent in Syria and 6 percent in Yemen.12 However, Jordan’s female labor force participation is the lowest in the world of a country not at war. Female unemployment, however, in nearly all countries in MNA is double that of their male peers. While some point to conservative attitudes enshrined in Islam, that explanation is insufficient. 9 WDR 2012 10 Mckinsey Report 11 Ibid 12 WDI 2017, retrieved in April 2017. 12 Countries with large Muslim majorities such as Malaysia and Indonesia experience a much higher participation level.13 Despite progress made in human development outcomes across the region, there are large gender disparities in key dimensions of economic and political empowerment. While the region has seen improvements in its Global Gender Gap ranking for the past two consecutive years, closing more than 60% of its overall gender gap, it continues to rank behind all other regions in the overall Gender Gap Index14. Ranking ahead of Sub-Saharan Africa and South Asia on educational attainment and ahead of East Asia Pacific and South Asia on health outcomes, MNA is lagging on political empowerment having only closed less than 9% of its gender gap from the previous year. With regards to economic participation and opportunity, MNA countries (Iran, Yemen, Saudi Arabia and Syria) represent four out of the five countries ranked at the bottom of that sub-index.15 Ongoing conflict affects women and men in different ways, and threatens progress made in human development indicators. There are active conflicts in Syria, Libya, West Bank and Gaza, Iraq, and Yemen in which women and men are affected differently. For women and girls, conflict increases risks to maternal and reproductive health, exposes them to gender based violence, and can lead to increases in child marriage, whereas men and boys are at an increased risk of being killed in battle or in the outdoors. Progress made in human and economic development indicators is already regressing in several countries such as in Syria and Yemen, where signs of educational attainment and labor force participation is dropping. For example, in Syria primary completion rates dropped from 100 percent in 2011 to 69 percent in 2013; in Yemen, female labor force participation has been decreasing over the years and to date stands at 6 percent (rates for men have also continues to decrease in Syria and Yemen). Women can be also impacted in different ways than men by other types of conflicts in their respective countries, such as by neighboring conflicts (in the case of Lebanon and Jordan); and civil unrest (in Egypt and Algeria). Studies have also shown that women are more likely to face constraints related to their mobility when there are demographic shifts in their community16. Jordan has been and continues to be particularly susceptible to external shocks and geo-political developments resulting from regional turmoil. As recent as 2017, Jordan was classified as a low middle- income country (from high middle-income country) due to deterioration of the human development index (HDI) and other development indicators. Conflicts in Syria, Iraq and Palestinian territories have had a fairly significant effect on Jordanian demographics. Fifty to sixty-five percent of the current Jordanian population is of Palestinian origin, comprised mostly of Palestinians who entered as refugees in 1948, 1967, and after the Iraq invasion of Kuwait in 199017. To date, there are more than two million Palestinians who live in Jordan as refugees in addition to 700,000 Iraqis and 635,000 registered Syrian refugees. After rebounding to an average of 8 percent between 2004-2008, the country experienced a major dip in its 13 Oxford Analytica, 2018. 14 See Annex 1 for description of Global Gender Gap Index structure and data sources. 15 Global Gender Gap, 2017. 16 IRC – to be confirmed. 17 WDI figures, SCD. 13 GDP in 2009 to 5.477 due to the financial crisis, where it dropped even further to 2.311 in 2010. The Arab Spring and ensuing conflicts in surrounding countries, particularly Syria, has made it more difficult to bounce back – economically. During 2004-2008, Jordan averaged a growth of 8 percent compared to the period between 2010-2016 where it averaged 2.55 percent (well below its potential). Internal pressures resulting in increased feelings of exclusion, unequal access to opportunities, poor accountability, constrained freedoms of expression and outcomes of rapid demographic shifts also contribute to the fragility in Jordan. Unemployment is high while majority of informal and low wage workers are among the country’s poorest, generating a sense of exclusion and undermining opportunities for ensuring shared prosperity.18 In a poll that was conducted in 2017, 52 percent of Jordanians felt that the politicians did not listen to the needs of ideas of young people, a substantial rise from 26 percent in 2015. In addition, lack of public accountability coupled with constrained freedoms add to social tensions and concerns for stability. The lack of job opportunities, low quality of public services exasperated by the influx of refugees, and unpopular economic reforms (e.g. bread subsidy and tax reform) are translating to rising social tensions. According to Jordan’s 2015 Census, Syrians make up 13.2% of the country’s total population, which makes Jordan the country with the second-highest number of Syrian refugees in the world. Jordan has made progress toward achieving gender equality in the area of human development such as in maternal and reproductive health, though of late is experiencing setbacks in terms of education attainment. Life expectancy for women has increased over the years and now surpasses men. In terms of reproductive and maternal health, women have almost universal access to prenatal care at 99 percent and maternal mortality ratio has decreased from 77 (per 100,000) in 2000 to 58 in 2015, higher than the MNA average of 81 that same year. Teenage pregnancy has also been dropping and stood at 23 percent in 2015 compared to a regional average of almost 40 percent. Still, there been some regressions. Primary completion rates have dropped from universal coverage in 2008 for both girls and boys to 91 percent in 2014. During the four-year period between 2010-2014, school enrollment at secondary school dropped at a higher rate for girls (from 82 to 71 percent) than for boys (79 to 69 for boys.) Figure 1. Female labor force participation rate 2017 18 Ibid. 14 While Jordanian women are among the most educated in the MNA region, their success in education attainment is not translating into the labor force.19 Jordanian women comprise 53 percent of university graduates yet female labor force participation rates in Jordan is the lowest in the region and among the lowest in the world, (135 out of 144) 20. In 2017, female labor force participation (FLFP) stands at 14 percent compared to 60 percent among men and only 13 percent among women nationals. And despite a variety of programs to support increased female labor force participation, between 2009 to 2016, women’s participation in the labor force has dropped from 14.9 percent to 13 percent compared to MNA regional average of 22 percent.21 Figure 2. Global Gender Gap - Education The more educated a woman in Jordan is the more likely she will be unemployed. Table 1 shows the Jordanian population by level of educational attainment, economic activity and gender. The gender divide 19 Global Gender Index, 2017. 20 Ibid. 21 Jordan Department of Statistics, 2016. 15 is striking: of the economically active population (those who are employed or in search of employment) only 13.3% are women, compared to approximately 60% for men. When disaggregated by level of education, economic inactivity among Jordanian women is highly concentrated among those with lower levels of education. Government data on women’s employment is incomplete however as many work in the informal sector (such as agriculture), therefore are excluded from the data. Fifty-nine percent of women in the labor force have a bachelor degree or above, followed by 28 percent of women with intermediate diploma (or associate degree), 5 percent with secondary, and 3 percent with less than secondary. Highly educated women are also more likely to be unemployed. Overall female unemployment is twice as high as men (22 percent versus 11 percent men, respectively) and for those with university degrees, approximately one in every two woman who is participating in in the labor force is unemployed versus one in every seven men participating. Challenges that exist in the Jordanian labor force lead many women of working age to give up looking for a job, or discourage them from even searching. As a result, the average inactivity rate of women with a secondary education or lower is almost 97 percent 22 . Approximately 46 percent of women graduates studied education and arts and humanities. Sixteen percent studied business, 13.6 percent have studied law and administration, natural science, math and statistics. For women who are in the labor force, unemployment is high --- particularly among youth. While overall unemployment is high among women in Jordan, there is variance across Governorates. According to the 2015 Department of Statistics data, female unemployment stands at 6.4% in Zarqa and 18.8 in Karak. 23 There does not seem to be a marked difference between urban and rural areas, 10.2% and 10.8% respectively.24 The very low employment rate among Jordanians is particularly striking among women but is also visible among men. Economic activity among women is of 13 percent compared to rate among 22 Ibid. 23 Jordan Department of Statistics, 2015 Census. 24 Ibid. 16 Jordanian men (at 60 percent). These rates are very low even compared to other MNA countries, for which the average employment rates are 81 percent for men and 27 per cent for women. 1.2. Barriers to Women’s Labor Force Participation Low labor force participation and high unemployment among Jordanian women are a result of a combination of different institutional, structural, legal and behavioral factors. Barriers to women’s employment are both multi-dimensional and inter-linked. Some of the key contributing factors include: “The distance [between work and • Low wages: In many cases, women are paid below home] is very important since the the legal minimum wage. A study shows a 12.3% gap woman is the pillar of the house…her in average wages between men and women in working hours should be from 8am-2pm since she will have time to go back Jordan, rising to 17% in the private sector. The home and prepare the food in addition average monthly wage for men is JD 403 to picking up the children. She must do [USD$568.01], compared to JD 359 [US$505.99] for all those tasks and get back home women. 25 before her husband.â€? • Limited job growth: There is also inequality in annual -- Working Jordanian mann (age 21-52), focus raises and promotions, and women do not receive group discussants some allowances. 26 This is particularly true in the private sector where men will earn as much as 40 percent more than women for the same job. 27 While the gender-based pay gap is less in the public sector it is still high at 28 percent.28 • Lack of a safe transport: Unsafe public transport is a deterrent for many women. Public transportation is often inadequate and unsafe, particularly for women traveling alone in rural areas. Women tend to spend much less time traveling than men: only five percent travel more than 60 minutes to work compared to 15 percent of men.29 • Lack of child care facilities: Lack of a friendly work environment to mothers, with access to child care facilities, also affects labor participation. Affordability of such facilities makes it necessary for households to assess benefits and costs of the different options at hand. Moreover, pre-school care is costly and domestic helpers are expensive. These issues combined with a minimum wage of $245 Jordanian dinars [USD$ 345.32] aggravate the options available. • Gender discrimination in hiring practices: Hiring a woman would entail taking into consideration the potential of time off for maternity leave, maternal benefits, provision of child care. Though efforts have been made to address some of these employer issues (i.e., social security reform) there is evidence of discriminatory behavior among employers. 25 Gender Diversity in Jordan, IFC. 26 Ibid. 27 ILO: http://www.ilo.org/global/about-the-ilo/newsroom/features/WCMS_213754/lang--en/index.htm 28 Ibid. 29 World Bank 2013: http://documents.worldbank.org/curated/en/503361468038992583/pdf/ACS51580WP0P130ox0379850B00PUBLI C0.pdf 17 • Lack of public sector jobs: Women still prefer to work in the public sector because of the shorter working hours, job security, and benefits. However, availability of public sector jobs tend to be limited and the waiting time for a position could be years, not months. • Intra-household barriers: Family pressure or perceived conflict between domestic and professional responsibilities as a result of uneven power dynamics to the disadvantage of women within the household further exasperate lack of access and participation in public life. • Societal expectations and norms: Perceptions and customs linked to young women and men’s transitions to adulthood, adolescent marriage, masculinity and the role of women and men after marriage, women’s responsibilities towards the husband, childcare, household work, and other normative constraints that contribute to women’s reduced participation. Similarly, the absence of social norms also has effects on the aforementioned barriers. For example, without restrictive social norms in issues related to work, there would be more opportunities for child care services, reduced discrimination, creation of new private sector jobs, etc. However, in spite of these constraints, many women aspire to be employed, irrespective of marital status. According to a 2010 World Bank survey of Jordanian female community college graduates entering the workforce30, 92 percent said they plan to work after graduation and 76 percent said they expect to be working full time. This is consistent with attitudes reflected in the Jordan Labor Market Panel Survey (2010), where 80 percent of women in Jordan disagreed or strongly disagreed with the statements that “A woman with a full-time job cannot be a good mother,â€? and that “Having a full-time job interferes with a woman’s ability to have a good life with her husband.â€? Over 80 percent of 54 women agreed with the statement that the “place of a woman should not only be the house, she should also be allowed to work.â€?31 There are yet other factors that impact women’s labor force participation. These include: access to finance/credit for entrepreneurs, education and training to acquire the skills/competencies needed to start a new business or join the labor force; incentives to provide the right skills and qualifications for high productivity occupations – and particularly representation of women in STEM (science, technology, engineering and math); long-term support measures for entrepreneurs – coaching, mentorship advice after initial training; low quality of women’s existing jobs (either wage employment or small business); and lack of women-friendly business regulatory environment. 32 Lastly, higher wages and better conditions in the public sector discourage women to pursue private sector jobs.33 Again, social norms are closely inter-linked to the aforementioned barriers. Table 2: Underlying constraints shared by women in Jordan • Family care responsibilities as well as time and mobility constraints that make it difficult for women to participate freely in the public domain. • Norms that pose obstacles to access and effective participation because women are required to get approval by male guardian to work, travel, or start a business. 30 Jordan New Work Opportunities for Women Pilot, 2010. 31 JLMPS, 2010. 32 Ibid. 33 Oxford Analytica, 2018. 18 • Under-representation in policy-making institutions and in decision-making processes. • Norms that discourage mixed work environments making it burdensome for women to participate in the private sector even if it were their preferred option. • Lack of sensitivity or awareness of women’s issues in schools or in the workplace. • Lack of affordable and quality child care • Informational constraints that limit women’s access to job opportunities and prevent companies from reaching out to available supply of female labor. • Limited opportunities for career advancement and wage equality. • Discriminatory recruitment practices Note: Author’s analysis based on desk research. 34 Balancing employment with household responsibilities makes it challenging for women to envision a long-term career. Women in Jordan are expected to manage a lot of their family-related responsibilities, which creates pressure on them because it is hard to balance professional and family responsibilities35. Studies have shown that marriage and family plays a major role in women’s decision-making process about her career aspirations and the type of job she will accept. Many Jordanian women have little expectation of a long-term career. 36 Female labor force participation decreases significantly after marriage as a result of perceived difficulties in balancing employment with household responsibilities37. Household responsibilities mean most women are unable to spend long hours away from home: in their calculations of the benefits of a job, time spent commuting is added to time spent working. Personal safety particularly in the context of work in hotels, manufacturing and domestic work is concerning. In all three contexts, there is a strong concern for women’s vulnerability in any environment where they are alone with men as well as a perception that employers in domestic work and manufacturing are harsh and exploitative. Research indicates that it is more important for women that there be other women at the workplace, more than full gender segregation. The “publicnessâ€? of the job, and the working environment - whether there are a lot of men, seem to influence women’s decisions. More generally, work places that do not require long commutes and which are “female friendlyâ€? are considered desirable.38 The study delves into: (i) views on working from inside or outside the home, (ii) gender roles (role of women after marriage, responsibilities towards the husband, childcare, and household), (iii) publicness and mixing (harassment going to/at work and environments with a lot of men), and (iv) status (financial need of a family, religiosity, and husbands failing to provide). 1.2.1. Legal barriers Normative and legal constraints for women make it challenging for them to get to and from work, choose the work of their choice, balance work and personal life, and/or cope with social expectations 34 GCC Engagement Note #3, Paving the Way for Women’s Economic Inclusion in the Gulf Cooperation Council, 2016. 35 Jordan Department of Statistics 2013b. 36 IFC 2015. 37 ILO Report 2017. 38 Ibid. 19 and family pressures about economic role and public participation. While personal status and labor laws are changing to open up opportunities for women in economic and political spheres, cultural barriers remain strong and need to relax over time for countries to fully harness and capitalize on women’s potential. 39 The extent of gender based legal restrictions that exist in Jordan related to women and work is mixed when compared to the overall regional performance. Differences in laws based on gender and marital status can affect women’s ability to access opportunities and participate in the labor force. Figure 3 shows that laws that provide incentives for women to work exist much more so in Jordan than MNA on average. For example, women are guaranteed their position or an equivalent position upon returning from maternity leave, primary education is free and compulsory, and tax deductions and credits are not gender based. However, while the government supports or provides childcare services, payments for childcare are not tax deductible. Further, laws around accessing institutions, in terms of getting a job or pursuing a trade or profession or traveling outside the home, traveling outside the home, being formally considered as head of household, or choosing where to live is different for women and men – and these restrictions are higher in Jordan than in overall MNA. Yet despite the fact that there are laws that govern freedom of movement and type of job a woman should accept, enforcement of the rule of law tends to be weak therefore the extent to which restrictions matter in practice is still unclear. Figure 3. Legal Barriers to Women’s Employment – Jordan versus MNA A Comparative Perspective: Performance of Legal Measures that Support Gender Equality Acoss Key Dimensions Affecting Women's Employment 90 80 70 60 50 40 30 20 10 0 Accessing institutions Providing Incentives to Work Getting a Job Protecting Women from Violence Jordan MNA Region Source WBL, Author calculations from 0-100 with 100 being the highest score. 1.2.2. Cultural barriers – masculinity, preferences, expectations (individual beliefs) 39 Women, Business and the Law, 2018. Covering 189 economies, Women, Business and the Law provides comparable data across seven topics. The methodology was designed as an easily replicable benchmark of the legal environment for women as entrepreneurs and employees. The methodology aims to understand where laws facilitate or hinder gender equality and women’s economic participation, Women, Business and the Law 2018 provides scores for the first time for each of its seven topics. Scores are obtained by calculating the unweighted average of the scored questions within that topic, and scaling the result to 100. 20 Since 2009, when countries in the region began to intensify their efforts to get more women to join the workforce, an array of studies have shown that cultural barriers to female employment persist. Adhering to traditions and customs when making decisions about personal or work-related matters is widespread in MNA. This is further reinforced by the fact that almost all personal status and family laws are religious based. For example, religious communities in Lebanon apply their own family codes and these can vary significantly depending on the denomination.40 In Jordan, laws governing personal status are based on affiliation.41 Yet they remaine unaddressed by policy solutions that are focused on solving the more objective barriers to employment, such as day care centers and transportation. As a result, female workforce participation rates in Jordan have remained dismal. In the first quarter of 2017, female unemployment rose 8.2 percentage points from 2016 to 33 percent42. Family is also central in young women’s decision-making processes when it comes to the world of work. Statistics show that some families put enormous pressure on young professional women, preventing them from working long hours, encouraging them not to take evening shifts, and so on, which hinders them from taking on more responsibilities and reaching higher management levels. Some professional women who reach middle management sometimes have to leave their posts because of family or social pressure43 A recently published 2017 Promundo and UN Women study helps bring to light powerful evidence of how gender roles—especially social, family, and cultural restrictions placed on women—may be the main factor hindering women’s participation in the workplace in the MENA region. Using data from the International Men and Gender Equality Survey (IMAGES) conducted in Egypt, Morocco, Lebanon, and Palestine, the study reveals that two-thirds to more than three quarters of men support the notion that a woman’s most important role is to take care of the household. In addition, the majority of men believe that it is their role to monitor and control women’s movement. In some countries, majorities of women appear to accept male guardianship. With women being defined as wives and mothers first—rather than as professionals—it is no wonder that the MENA region has low female workforce participation rates.44 1.2.3. Regulatory Framework and Policies Women’s participation in economic, social and political life are straight jacketed due to gaps in both the policy and institutional frameworks necessary for addressing gender equality. The quasi-governmental agency mandated to improve women’s empowerment is hosted by the Jordan National Commission for Women (JNCW). The agency has developed broad policies on gender, such as the National Strategy on Women and has been responsible for monitoring its implementation. The agenda has been primarily donor-driven rather than government funded, undermining government commitment and sustainability. Several of the line ministries, such as the Ministry of Planning and International Cooperation, (MoPIC), 40 Lundvall et.al. 2017. 41 Ibid. 42 Brookings Institute. 43 Jordan Department of Statistics, 2013b. 44 Promundo Study, 2017. 21 Ministry of Education (MoE), Ministry of Health (MoH), and others host “Gender Unitsâ€?, however coordination and clear vision is lacking to enable impact. For example, basic data on gender from the public sector institutions is mostly lacking.45 1.2.4. Institutional and Policy Changes Some progress has been achieved with regards to policy, legislation and procedures adopted to promote gender equality and protect women’s rights in general, but improvements are still needed. In the last 10 years, there has been an increase in women’s representation in decision making and leadership positions in the political arena. In the amended Electoral Law (2010), the quota for women in parliaments increased from 6 to 12 (out of 20). In addition, women have been appointed to various leadership positions. However, a government report noted that, in the absence of a gender equality law and strategy, the progress has been uneven and slow and in some cases there have been some setbacks. In the civil service specifically, women account for approximately 50 percent of employees in Jordan. Despite advances in women’s participation in public administration at lower level positions, representation of women in leadership positions continues to be of concern.46 In 2003, Jordan amended the legal age of marriage to 18 years, although a gap remains in the law, where the marriage of girls under 18 can be left to the judge to decide. In the Personal Status Law (2010) of Jordan, further restrictions apply in cases of early marriage. The Law considers wife-abuse (whether physical or emotional) to be grounds for divorce, and Article 62 provides women with a legal basis to bring a divorce case. Jordan is one of the few Arab countries to have issued a domestic violence law (2008). The National Task Force for Family Protection, led by the National Council for Family Affairs, is drafting a new proposal for a domestic violence law that would take into consideration the definition of violence against women as well as a clear procedures and bylaws. The regulatory framework remains disabling for women to balance work and household obligations. In terms of special measures related to creating a friendly environment at the workplace for women, the Civil Service Bylaw includes some measures such as that of flexible working hours, breast-feeding hour/room, and making it obligatory for ministries and government departments to provide childcare centers in certain cases. However, the lack of reinforcement of these measures remains an issue in the country. These measures are usually determined by employers, and hence are totally dependent on the personalities of the managers and their willingness to support female employees within their departments. Different studies show that there has been limited recourse for reporting sexual harassment in the workplace due to limited protection measures and enforcement of law, and women who are exposed prefer to leave their work rather than experience possible victimization and the social stigma of being harassed.47 45 Jordan Country Gender Assessment, 2013. 46 UNDP Democratic Governance Paper. 47 ILO Working Paper (2012), Response to Classic Fashion Apparel Industries Allegations of Sexual Assault and Better Work - Jordan Follow-up on the Issue of Sexual Harassment. 22 1.2.5. Role of Conflict and Fragility Conflict and fragility has also played a role in outcomes related to women’s economic empowerment. Incipient research has revealed how laws and attitudes restricting women’s mobility, household care burdens, occupational segregation, and son preference constrain women’s economic participation in communities. However, the effects of conflict and displacement also exasperate existing disparities among refugees or host communities, especially women. Conflict also impacts gender relations and dynamics within the household. Examining behavior as well as how shifting power structures in times of crisis influence decision making is critical to ease constraints facing women and girls in their pursuit for quality education and decent work. 1.2.6. Social Norms and Cultural Beliefs Norms and beliefs play a key role when it comes to women’s economic empowerment in MNA. In order for the MNA region to increase women’s economic participation, it is important that we acknowledge and begin to break down these cultural barriers. Informal institutions, such as social and cultural norms, have been identified as key constraints to achieving gender equality in the region. Social norms and beliefs dictate behavior in ways that affect important development outcomes for women, including their participation in public life, access to economic opportunities, educational attainment, ability to choose their own health care, socialize or inherit and control land and financial assets. However, social norms and beliefs are often conceptualized inadequately, often using terms such as “social normsâ€? and “cultureâ€? and “conventionsâ€? interchangeably. Understanding social and cultural norms and the expectations of women’s behavior will help the World Bank projects and institutional partners design appropriate interventions to increase women’s economic participation and in turn to broader economic development and growth. Why is understanding social norms important? The World Bank Development Report 2012 on Gender and Development and the ensuing Opening Doors: Gender Equality and Development in MNA (2013) reports were instrumental in identifying informal institutions (social and cultural norms) as a key binding constraint on the achievement of gender equality in MNA. Qualitative research on gender in MNA brings to life the diverse social and cultural norms that influence women in many dimensions of their lives. For instance, incipient qualitative research reveals how gender norms surrounding women’s restricted mobility, household care burdens, occupational segregation, and son preference constrain women’s economic participation in communities in and around Amman, Jordan. Yet there is very scarce literature of the role for development interventions to influence gender norms and relations, especially in MNA. 1.3. How are we different? Results of the Jordan study contribute measurable evidence to gaps in social norms literature. Over the last five years, the concepts surrounding social norms have garnered a lot of attention yet it remains a nascent field in practice with little evidence of the role it plays in affecting development outcomes. There are many studies that have looked at structural and institutional barriers as well as identified social norms as a barrier to labor market participation, however many of these do not delve into mechanisms through which norms influence behavior nor do they explore intra-household dynamics in a measurable way. 23 Similarly, others have talked about masculinity, without looking more closely at how it plays out at the household level. In contrast, the Jordan study also examines intra-household dynamics and how expectations of women’s spouse or male relative’s beliefs could be a constraint as well. Overall, the study developed techniques to measure the extent to which social norms and beliefs concerning gender influence women’s access to and participation in the labor market. It is the region’s first ever systematic, representative, quantitative study of social norms (across four areas of factual beliefs, personal normative beliefs, empirical expectations and normative expectations) at the national level, and perhaps the first in developing countries. It also provides new techniques for measuring and validating social norms related to gender, a model that could be applied to other countries, as well as innovative mixed methods research, including the analysis of narratives. 24 SOCIAL NORMS 25 2. What are Social Norms? Social norms related to gender are pervasive in every human society. They prescribe a wide range of practices, behaviors, and even thoughts and feelings that are considered appropriate for men and women to have and exhibit. These include inter-personal behaviors (e.g., when and how many children to have), social practices (e.g., dress, speech, dominance, child rearing), political actions (e.g., holding and exercising public office, voting), and economic decisions (e.g., participating in the labor force, opening a bank account, starting a business, employing others). Although these various behaviors usually draw on a common set of mental models48 of the world and human relationships (including the very idea that society is composed of people of either one or another gender), they are somewhat separable. There can be changes in social norms in one area but not in others. For example, in many societies it has become more acceptable for women to work and hold office, but there has been less change in the division of household labor and responsibility for child-rearing.49 These behaviors change when social norms, which are shared mutual expectations in a social group, themselves change. In this section, we develop an account of social norms related to women’s labor force participation, drawing on work by Bicchieri, et al. (2014), Eriksson (2015), etc., and explain how it informs our measurement approach. We used this theoretical framework in the design of our survey instrument and in our analyses, which we present in the subsequent sections. This account also helps explain how social norms are different from “culture,â€? how in certain circumstances social norms exhibit tipping points that can unravel quickly, how social norms can be measured and analyzed, and how development practitioners and other agents of social change can use this framework to develop interventions to induce change. Generally speaking, if within a society, people believe that (1) a large number of women will not join the work force, and (2) they also believe that most people think women should not join the work force, then there exists an expected behavior supported by a normative belief. These are the constituent parts of a social norm. These two sorts of belief are called, respectively, empirical expectations and normative expectations. These two beliefs, plus a conditional preference to conform to the norms that they support, constitute Bicchieri’s operational definition of a social norm50: • Empirical expectations: individuals believe that a sufficiently large part of the relevant group/population conforms to the norm. This is a first-order expectation: a belief about what others do. • Normative expectation: individuals believe that a sufficiently large part of the relevant group/population believes they ought to conform to the norm and may sanction behavior. This is a second-order expectation: a belief about the beliefs of others. • Conditional preference: individuals will prefer to conform to a social norm on the condition of holding the relevant empirical and normative expectations. 48 Women’s economic participation may also be influenced by deeply internalized beliefs or mental models about how the world works, also known as cultural schema or beliefs. These cultural schema shape perceptions and behavior by filtering the “factsâ€? that people believe and are able to understand. 49 This section draws on Raj Patel, “social norms literature review,â€? unpublished review conducted for eMBeD, 2016. 50 Bicchieri (2006, 2015). 26 To understand how social norms work, it is important to distinguish them from other kinds of normative and empirical grounded behaviors. Not all collective practices are supported by social norms. Consider the following examples: (1) A moral rule: A moral rule is usually unconditional on shared social beliefs. If I believe it is morally impermissible to kill for sport, then it is irrelevant whether my reference network (or relevant group) share this moral belief. (2) A descriptive norm: Descriptive norms are conventions and are usually solutions to coordination problems. Descriptive norms are based purely on empirical expectations – individuals believe that a sufficiently large part of the relevant group will conform to a particular behavior, and so they will do so too. Norms about standing in line at the airportis an example of descriptive norms. (3) A custom: A custom is a tradition. Customs usually have a reason for their emergence, but many customary practices are supported simply because that is what people have been taught to do. (4) Rational responses: Behaviors that can be explained as rational responses are usually agents within a given incentive structure maximizing a particular good for themselves. For example, if marrying a child earlier is cheaper because it reduces costs to the family, then child-marriage may be a rational response in the service of cost-reduction. These distinctions are important because to change behavior, one needs to understand the cause of the behavior. For example, if an undesirable social practice is rationally motivated then incentives need to be structured differently. If a pernicious practice is a custom, then a community could be informed about why the reasons supporting a traditional practice no longer hold in order to abandon that custom. But if a particular practice is a social norm, then both empirical expectations and normative expectations need to be changed. This means changing collective beliefs about a particular norm, since social norms are supported by what others think and do (what I refer to as the relevant group or relevant reference network). Surveys aimed at uncovering social norms ought to be constructed to appropriately delineate between practices that are independent and interdependent. Interdependent practices are supported by what others believe and think. In that sense, they are conditional: I will drive on the left side of the road if others drive on the left side of the road. Independent practices are not supported by what others believe and think. In that sense, they are unconditional: I will brush my teeth regardless of whether other people brush their teeth. 27 Observe a collective pattern of behavior People prefer to People prefer to follow it follow it if they irrespective of believe others what others do follow it Empirical Normative Collective custom, expectations expectations are shared moral rule suffice to motivate also needed to or legal injunction action motivate action Descriptive norm Social norm Figure 4. Identifying different behaviors.51 It should be noted that even in the case that a particular behavioral regularity is not supported by a social norm, that is, in the case that it is driven instead by incentive structures or moral rules, social norms might still be created to change the behavioral regularity. Examples in this domain include the recent creation, in some countries, of social norms against smoking, drunk driving, and the disparagement of LGBT individuals. The #metoo movement may be in the process of generating new social norms regarding sexual harassment. Legal scholars have touted the creation of social norms as an efficient mechanism to induce desirable behavior (since enforcement of legal rules is costly).52 Women’s economic participation may also be influenced by deeply internalized beliefs about how the world works, also known as cultural schema or beliefs. These cultural schema shape perceptions and filter the “factsâ€? that people believe and are able to understand (World Development Report, 2015). For example, women may believe (correctly or incorrectly) that some industries are women friendly, and may not apply to others. They may have other beliefs that prevent them from participating, such as beliefs about safety, transportation, and the level of skills required. This careful and systematic measurement of different components that comprise social norms, and eliciting cultural beliefs will allow us to decide whether one or the other (or both) may explain behavior, and thereby arrive at relevant and appropriate designed interventions. For example, if we find that men or women severely underestimate the number of women that work (for a given demographic, or region or sector) or that more women are more interested in working than people expect, this may suggest an advertising campaign intervention that corrects these misconceptions. Similarly, a campaign may try to 51 Bicchieri, Social Norms, Social Change. Penn-UNICEF, July 2012 52 Posner (1997). 28 re-brand the role of women in society highlighting female employment in a positive light – if cultural beliefs are the binding constraint. The measurement of social norms requires questions carefully targeted at a) personal beliefs, preferences, and behaviors; b) expectations regarding how others behave (e.g., how many other women work, and what they sacrifice when they work); c) expectations regarding the normative beliefs of others and their sanctioning behavior (e.g., how many other people disapprove of women working, and whether they speak badly of women who do). In addition to these, we also target questions at the mutual expectations of male-female pairs in the household (primarily husbands and wives, but also fathers and daughter, and brothers and sister). In focusing on the behavior of male-female couples, we explore the hypotheses that a) the social norms are enforced primarily in households, rather than through generalized disapproval, gossip, or harassment; b) intra-household power dynamics may constrain women’s economic participation even independently of social norms. 29 METHODOLOGY 30 3.1. Sequence of Activities The study develops techniques to measure social norms and beliefs around female labor force participation in Jordan to understand whether and how they influence women’s behavior and decisions regarding whether or not to work. Eliciting the influence of social norms and beliefs can provide a more nuanced understanding of the binding constraints that prevent women from working, and subsequently, inform the design of interventions aimed at improving female labor force participation and employment opportunities for women in Jordan. Our study includes five key activities outlined in Figure 5 below. Figure 5. Timeline of activities 3.1.1. Focus Group Discussions Measuring the influence of social norms and beliefs on choices and behavior related to labor force participation requires, first and foremost, an understanding of the prevailing norms and beliefs that are commonly expressed and observed in a specific context. It is evident from the current literature on women’s labor force participation in Jordan and the MNA region that structural barriers, such as lack of good transportation, childcare, and limited supply of job opportunities, etc., do impede women’s participation. However, several studies also point to non-structural, social and cultural barriers that appear to play a key role in decision-making. Another key insight that emerges from these studies is that women are not independent decision-makers when it comes to their own employment. To inform the development of our quantitative survey instrument measuring the influence of social norms, beliefs, and intra-household dynamics as they relate women’s labor force participation and decision-making, we first carried out a series of Focus Group Discussions (FGDs) with six distinct groups of respondents from urban areas in Amman and Mafraq. The FGD instrument was developed based on review of the existing literature, and was designed to both understand social norms and beliefs that influence labor force participation decisions as well as to fill in gaps identified in the existing literature on decision-making dynamics. Our implementation partners, the local survey firm in Jordan and D3 Systems (USA), were consulted in the process to refine the instrument 31 and adapt it to the local context. The instrument was subsequently translated into Arabic, and further refined based on feedback from the survey firm’s field staff, which was pre-tested. The FGDs explored four key themes that affect decision-making: 1) role models and aspirations; 2) vignettes on employment decisions across various scenarios; 3) decision-making and employment related behavior; and 4) norms, expectations, and barriers to work. The instrument was further adapted for the four different respondent groups- working women, non-working women, working men, and non- working men (irrespective of age)- keeping the four themes consistent across all instruments. Fourteen FGDs were conducted across Amman and Mafraq in May 2017 with groups of 8-10 individuals from low to middle income households within each group. Data from the FGDs were recorded, transcribed, translated, and carefully analyzed to draw key themes to be explored in the quantitative survey. 3.1.2. Quantitative Survey The quantitative survey involved two phases of data collection- i) pre-test of the draft survey; ii) roll out of the final survey from March to April 2017. The draft quantitative survey instruments (separate for men and women) were developed based on findings from the FGDs, and included narrative questions (qualitative) asking men and women to discuss their thoughts and opinions on related topics, to elicit their true beliefs. The draft surveys were field tested in in November 2017 to determine how effectively the survey questions elicited social norms and beliefs, improve the language and instructions provided to enumerators on questions, and utilize feedback from the field and survey data to further refine the final instruments. Once again, both the survey firm and D3 Systems as well as external social norms measurement experts were consulted in developing the draft surveys and refining the translated instrument in Arabic. The pre-test was conducted with a small number of households in Amman, and included households where both the husband (usually the household head) and wife were interviewed. A total of 50 individuals were interviewed. The purpose of interviewing both spouses within a given household was to test the feasibility of interviewing couples in a household separately, which allow us to delve deeper into intra-household dynamics as they relate to women’s decision to work. Insights from the pre-test were incorporated to finalize the quantitative survey instruments for male and female participants, and refine the guidance provided to the interviewers on how to administer the social norms questions and elicit informative but concise narratives from respondents. The final version of the survey was administered with 2007 individuals across three governorates- Amman, Mafraq, and Zarqa- between March and April 2018. The sample included 828 households where both a male and a female respondent was interviewed (pairs included either husband and wife, brother and sister, or father and daughter). Section 3.4 discusses the sampling frame and key demographic characteristics of households and respondents in more detail. 3.1.3. Lab-in-the-Field While originally not part of the survey design, a short lab-in-the-field experiment was integrated into the research and conducted concurrently with the quantitative survey. Lab in the field experiments can 32 be useful in measuring how features of human behavior manifest under natural conditions and in a randomized fashion. The experiment involved showing a short 90 second video featuring a working woman alongside a supportive husband and her family in Amman to 50 percent of the survey respondents (a random sample of women and men) to assess whether light-touch interventions such as this have any impact on increasing women’s motivation to look for jobs. Several options for scripts were discussed and local film-makers were hired to identify real life characters who have an inspiring story to tell. The video also included a positive message at the beginning of the story stating data about the growing number of women entering the labor force. All respondents, irrespective of whether they saw the video, were provided with contact information of a job search advisor at the end of the interview (male respondents were asked to pass the information on to their female relatives in the household), whom they could call for career advice and guidance. Information was provided in the form of a flyer announcing that respondents or their female family members interested in information about jobs for women in Jordan could contact a jobs advisor at the listed number. A disclaimer was also included in the flyer that no jobs will be offered. Further, in handing out the flyers, enumerators communicated that the job advisor could be contacted for career advice and guidance. Those who would call would be offered a phone credit cover the cost of the call. A male career advisor was hired based on his local experience in the human resources field and was available from 9:00 am – 5:00 pm on set days of the week for calls.53 3.1.4. Follow-up survey Lastly, a quantitative follow-up survey was conducted with a random subset of respondents, some of whom may have watched the video, two weeks after the completion of the quantitative survey. The purpose of this survey was two-fold: the first was to assess if responses to social norms and beliefs related questions stay consistent over time; the second was to elicit if exposure to the video experiment had any impact on women’s efforts to look for work. 3.1.5. Data Quantitative data from the pre-test, quantitative survey, and follow-up survey were collected on tablets. The short video was shown at the end of the quantitative survey using the same platform. Data were analyzed using Stata to generate descriptive tables and figures and to conduct regression analysis investigating the extent to which expectations and norms predict women’s employment status and preferences. 53 The team preferred to identify a female jobs advisor however it was difficult to find candidates due to the nature of the experiment and time and budget constraints. 33 3.2. Qualitative Findings 3.2.1. Sampling To inform the development of the quantitative instrument for this study, we conducted a series of focus groups in Jordan in order to gain a deeper understanding of the role of social norms, beliefs, and intra- household dynamics in women’s decisions regarding labor force participation. 14 focus group discussions were conducted across Amman and Mafraq, each with 9 to 12 participants from low to middle- income households. The following table outlines different groups that participated in the focus groups: Table 3: Focus Groups Recruitment Criteria City Gender Age Group Employment Refugee Status Status Group 1 Amman Women 21-30 Working No Group 2 Amman Women 31-59 Working No Group 3 Amman Women 21-30 Not working No Group 4 Amman Women 31-59 Not working No Group 5 Amman Women Any Any Yes (5 years or less in Jordan) Group 6 Amman Men Any Working No Group 7 Amman Men Any Not working No Group 8 Amman Men Any Any Yes (5 years or less in Jordan) Group 9 Mafraq Women 21-30 Working No Group 10 Mafraq Women 31-59 Working No Group 11 Mafraq Women 21-30 Not working No Group 12 Mafraq Women 31-59 Not working No Group 13 Mafraq Men Any Working No Group 14 Mafraq Men Any Not working No To recruit participants, the local survey firm used a 3-degrees of separation approach, where they asked people their field teams know to recommend people, who would then also recommend participants who fit the specified criteria. Discussions were facilitated by a trained moderator, and took approximately 70 minutes on average. Data from the FGDs were recorded, transcribed, translated, and carefully analyzed to draw key themes to be explored in the quantitative survey. 3.2.2. Themes 34 The focus group discussion guides were designed to understand the following: (i) Role Models and Aspirations- goals, gender roles and family influence (ii) Decision-making process and landscape- Why do women work? What do they look for? Who are their reference groups? Who are the influencers and decision makers? (iii) Attitudes and social norms- What is considered appropriate? How does society view working women? How many women work? (iv) Differences between men and women, as well as differences between working and non- working people. The questions that were asked aimed to understand the following: i) Role models and aspirations Participants were asked about their role models and aspirations in life. The importance of this question was to understand who do the participants look up to in their lives and what qualities and characteristics make them their role models. Following this, participants were asked about their aspirations for the next generation and what advice they would provide for their daughters. Given that the participants emphasized on the importance of education, a probing question was designed to better understand the importance of education and its linkage to joining the workforce. ii) Women’s role in society The goal of this section was to understand how participants perceive the changes in the role of women in society and the ways it has changed. iii) Vignettes The facilitator of the focus groups presented two short stories about fictional female characters in the Jordanian society. The purpose of the vignettes was to help understand the extent of the participants acceptance of these changes and what changing factors lead them to accept or decline the concept of a working woman. The first story focused on changing certain conditions in the character’s life, such as having children, husband being offered a job with a comfortable salary, the character getting paid a salary higher than her husband, or a change that includes having a job offer that requires regular interaction with men. In the second vignette, the focus was more geared towards changes in the employment status of women in the character’s community. The respondents were presented with a scenario where the women in the character’s community are mostly working women and the second scenario if mostly non-working women. The purpose was to assess how social norms impact the decision-making process for individuals. iv) Decision Making 35 The goal of this section of the study is to have a better understanding of the decision-making process when it comes to employment and the intra-household dynamics related to the decision. Questions in this section were divided into two sets: one targeting working men and women and another targeting non-working women and men. For the non-working group, the questions focused on the reason why they are not economically active; what stopped them from working if they worked in the past; the circumstances they would consider working, and who would they consult to make a decision regarding joining the workforce. For the working group, the questions focused on why did they choose to work, the circumstances that would make them stop working, and the people they seek for advice or guidance when making a decision related to working. v) Norms and Attitudes This section aims to learn about the impact of social norms on women’s participation in the workforce. Understanding how society’s norms and expectations could trickle down and influence decisions on an individual level is crucial for when designing interventions. The questions in this section capture the normative expectations, empirical expectations, and individual behavior of the participants in the focus groups. 3.2.3. Major Findings i) Role models and aspirations Role models for women, specifically non working women, were mostly family members. When women in the focus group were asked the reason behind their choices, the answers focused mostly on the hardships that these female family members had to overcome to successfully manage the family and the household. For men, their role models were male family members due to their ability to provide for the house and certain characteristics such as honesty. Success - When the facilitator asked what does make a woman successful in life, answers describing women who work and are able to manage their households came up frequently. Having a “strong personalityâ€? was another response highlighted by the participants. Valuing social skills was highlighted by non working women. Aspirations and Advice- When the focus of the questions shifted to aspirations for the next generation and advice for their daughters, gaining good education was the most repeated advice. The response of older women in the focus group tends to focus more on their desire for their daughters to have virtuous qualities, while younger women were more outward looking in their responses; highlighting their wish for their daughters to achieve their goals and dreams and gain the skills needed for them. On the other hand, non-working men were more conservative in their responses, and focused on wanting women to “know their boundariesâ€?; given that men are the providers. 36 Education – the added value of education to women’s lives according to the participants focused more on empowering women by building and strengthening their character, enhancing self-confidence, and improving marriage and social prospects. While the importance of education came up frequently in participants’ responses, its value was more of a social net and a “sword to rely on in the event of divorce or death of the male providerâ€?. Men and young women in the focus groups brought up self-reliance as one of the main values gained from education. However, the economic return of education and its importance to build a career was only mentioned a few times. ii) Women’s role in society It was widely agreed by women and men in the focus groups that women’s role in society is advancing. The participants highlighted examples such as women are now “more visible in societyâ€?, have a stronger role in supporting the household financially, and are more comfortable in working in mixed work environments. In addition, respondents highlighted the changes in society that led to the change of women’s role; such as society becoming more accepting of working women, more opportunities are opening for women and more space is provided to them by men, who are now becoming less traditional “prefer committing to educated and working womenâ€?. While the majority perceived the change as positive, some of the respondents perceived these changes as negative due to their concern that such changes will lead to women changing their behavior and becoming “arrogantâ€? or “bossyâ€?, and could look down on their husbands. Others said that women often do not share their income with their husbands and having a job means being less attentive to household duties. Some of the respondents also highlighted that even with changes in the role of women in Jordanian society, some expectations remained unchanged. Raising children and taking care of household responsibilities continue to fall mainly on women. iii) Vignettes When most of the structural barriers such as the provision of daycares, a reliable transportation system and employers offering equal pay are taken care of, the respondents believed that the character in the vignette should accept the job offer. Both men and women believed that given a good opportunity, the character should accept the offer, as it would contribute positively to her personality, financial and social status and the family’s living standards. However, when the change in the character’s life was having a child, some of the respondents believed that it was unacceptable for a mother to leave her infant for work, even if childcare is provided. Men and women, mostly non-working, believed that raising a child should be the woman’s priority, and she could join the workforce at a later stage. 37 In the scenario when the character’s husband is offered a well-paying job, some of the respondents believed that the character will decline a good job opportunity, as the husband will be able to provide for the family. In the case of gender mixed work places, the husband’s trust and permission was the key determining factor for the woman’s decision to either accept or decline the job offer. The impact of social norms in the decision-making process did not seem to play a significant role. Whether women in the community are working or not, if provided with the right opportunity and suitable conditions, respondents believed that the character in the vignette should accept the job. iv) Decision-Making The main reason that repeatedly came up during the focus group for women not working or exiting the work force was due to having children. The long working hours made it difficult for some of the women in the discussion to juggle work, taking care of a newborn, and household responsibilities. Other structural barriers such as the lack of adequate public transportation or child care facilities, and the scarcity of attractive job opportunities also came up. The key driver for women’s employment in Jordan was financial need. Whether the question was about the circumstances that would make women consider working or lead them to joining the workforce, the majority of participants answered that the high cost of living would force them (or forced them) to find a job. This was highlighted mostly by women participants in Amman, who believed that it is their husbands responsibility to provide for the family, not their responsibility. Women in the focus group agreed that the most important factor for them to decide whether to accept the job or decline it was the suitability of working hours, followed by the appropriate work environment and good salary. However, almost none of the participants mentioned the profession or career trajectory as a decisive factor. When asked about the decision-making process, both men and women agreed that men are the ultimate decision makers in the household; including deciding for women on whether to accept a job offer or not. Women tend to seek either their husband’s or father’s opinion or advice when deciding to work. However, women seemed to prefer to seek advice from fathers, because they believed that they are more experienced. They also preferred consulting their fathers because they consider them to be more lenient than their husbands and could convince their husbands to change his mind if opposing the wife’s decision. Nonetheless, young working women in the focus groups believed that their personal decision to be the final decision. Male participants seemed more conservative in comparison to female participants in the group. They were more opposing to late working hours, mixed gender workplaces and in many cases, they feared that women would neglect their household responsibilities if they work. 38 v) Norms and Attitudes Men and women in the group reflected that society respects working women and perceive them as hard workers, as they support their husbands and enhance their families’ living conditions. The participants were asked to give an estimate of the percentage of working women in Jordan; surprisingly, the majority of the participants gave a very high percentage of 70% on average. Societal norms and expectations in this case do not seem to be the defining barriers for women’s low level of economic participation in Jordan. Yet, when household responsibilities came up, men in the group emphasized on the importance for women to not neglect their household duties for their jobs. Overall, working men in the group were supportive of the idea of their wives working; however, they still expect their wives to be the caregivers and meet the household responsibilities as they do not have a job. The situation was different for non-working men, even with the financial hardships, many preferred that their wives stay out of the workforce and focus on taking care of the family instead. With regards to the discussion of what jobs seemed acceptable and non-acceptable for the group, non-working women seemed to focus more on the type of the job: whether it included mixing with the other gender or requiring working long hours. For working women, the focus was more on the experience that they would gain from the job and the level of independence the job would provide them with. Nonetheless, both groups seem to care about distance to the job, salary, working hours and age of their kids. 3.2.4. Takeaways Constructed gender roles are still a major obstacle for women joining the labor force in Jordan. Although social expectations may not be a binding constraint impeding women from joining the workforce - (whether women in the community were working or not had no impact on individual’s decision to work), women are still expected to be the caregivers and handle the household responsibilities and chores alone. Women’s income is viewed as a secondary source; the main breadwinner is expected to be the man. There is a low level of appreciation of the benefit of having two incomes instead of one on the household living standards. For women to work, it mostly meant the family is going through financial crises or the women themselves chose to “fulfill their timeâ€?. The topic of a woman’s career in Jordan did not come up, which could be a reflection to the lack of career path for women in the country. Upward mobility in the job did not seem to be an option in many cases. Nonetheless, participants in the focus groups conducted in Mafraq had a higher appreciation on the impact of two incomes on the household. The men interviewed in Mafraq also seemed to be more willing to help their wives with household chores than those in Amman. Decisions are made within the household rather than influenced by society’s actions . In general, the decision to work or not for women was expected to be a decision made within a household, and not influenced by the community, and based on the preferences of the woman and male head of the household. Fathers and husbands have a heavy weight on a women’s decision-making process of various aspects of their lives. The team at the World Bank modified the quantitative instrument to include 39 interviews targeting fathers and daughters, and husbands and wives, and brothers and sisters, to understand the decision-making process within the household. This could help understand the gap in society’s acceptance of working women in Jordan versus the reality of the low percentage of women in the labor force. Role models for many Jordanians interviewed seem to be other family members who happen to be caregivers, providers and people who manage to balance between excelling at their jobs and taking care of their homes. The lack of adequate job opportunities, low salaries, unreliable public transportations, inflexible working and lack of child care facilities at the workplace were all issues discussed by participants in the focus groups. Female level of education in Jordan is high and continues to be the primary aspiration that women have for their daughters. However, the purpose of getting a good education does not necessarily correlate with the desire for finding a decent job. A good education was associated with better social status and a social net to lean on for women in case of unfortunate circumstances. The data collected from the FGDs was used to develop an updated quantitative instrument to examine the impact of social norms on female employment outcomes. One of the main takeaways was modifying the quantitative survey to focus on interviewing respondents in pairs (husband and wife, father and daughter, and brother and sister) to look into the magnitude of interhousehold dynamics in women’s labor force participation in Jordan. The quantitative survey was modified to focus more on the decision-making process in the household and the level of support provided to women from their male relatives. Although in general, decisions appeared to be made at the household level, we were not unsure as to how generalizable these findings were from the focus groups, and incorporated both social norms questions and perceptions of a male relative within in the household into the quantitative instrument. Lastly, while focus group discussions were conducted with both female and male refugees in Amman, it was decided to not conduct the quantitative survey with refugees, see below box for details. Adapting to FCV Contexts: The Challenge of Studying Social Norms Among Syrian Refugees in Jordan In general, the design and implementation of survey field work requires a great deal of coordination and planning. In FCV contexts, there is an added complexity in managing uncertainty and the disproportionate impacts of conflict among various population groups. While Jordan is not considered an FCV country, it has been hugely affected by surrounding conflicts over the decades and much of its current demography is characterized by large population influxes of Palestinians, Iraqis and, most recently, Syrian refugees who also face many challenges with securing livelihoods among other needs. The original intent of the study was to measure social norms and beliefs that influence Jordanian and Syrian refugee women’s labor force participation with the aim to eventually use the findings to improve employment opportunities for women in Jordan. The focus on both Jordanian and Syrian refugees was an important aspect of the study as it would bring to light impact of conflict on traditional gender roles among refugees as well as any spillover effects among host communities. Examining behavior as well as how shifting power structures in times of crisis influence decision making is critical to ease 40 constraints facing women and girls in their pursuit for quality education and decent work. In addition, the survey work would complement World Bank Group emergency response to address the growing humanitarian crisis in the region through interlinkages with host communities, with a focus on economic participation and improved livelihoods. Despite the intended focus, the component of incorporating refugees in the study was removed after conducting qualitative field work targeting Syrian refugee women and men. During several focus group discussions with refugees, it became evident that the challenges they faced were not only very different in terms of access and opportunities but also in terms of urgency combined with great uncertainty and other urgent priorities that made it challenging for them to respond to questions around preferences, beliefs, or societal expectations. The team would have needed to adapt the social norms framework in order to adequately reflect these issues and this was not possible to do so within the project timeframe. Therefore, after much consideration, the team decided to focus on Jordanian women and men for this current round while leaving open the possibility of readapting the social norms framework to cover refugees in the Mashreq/region. 41 3.3. Quantitative Survey Instrument According to Bicchieri, Lindemans, & Jiang (2014), the identification of a social norm requires measuring at least five components: a) individual behavior; b) empirical expectations about the behavior of others; c) personal normative beliefs; d) normative expectations concerning the moral beliefs of others; e) cost- benefit calculations on the part of individuals. It is only by measuring all five components, specifying the reference group to which they apply, and weighting them adequately, can we diagnose the existence of a social norm, as well as its strength. It is also important to measure the reasons people act in the ways that they do – some people may be acting in ways that are consistent with a social norm but for rational, or self-interested reasons. If that is the case, changing behavior may require not only changing the social norm but also changing incentives. Women’s labor force participation may also be influenced by deeply internalized beliefs about how the world works, also known as cultural schema or beliefs. These cultural schema shape perceptions and filter the “factsâ€? that people believe and are able to understand. For example, women may believe (correctly or incorrectly) that only poor women work or that women are only suited for some professions, such as teaching. They may have other beliefs that prevent them from participating in the labor force, such as beliefs about safety, transportation, and required skills. In contrast, social norms do not involve beliefs but preferences, and specifically conditional preferences: people are willing to engage in a behavior if they believe that others are engaging in it or believe that others think they should engage in it. So how do we measure the strength of social norms in influencing behavior? If normative and empirical expectations are consistently reported in a social group, there is strong prima facie evidence that a social norm exists. Similarly, if a sufficient number of respondents express mutually consistent views about what others should do, normative expectations are strong. Normative expectations are necessary, but not sufficient, for the existence of a strong social norm. In order for a social norm to be strong, the normative expectations need to cause conformity, or change behavior, in sufficiently large numbers. To demonstrate the causal efficacy of the norm and see if individuals, in fact, do conform to the social norm, it is necessary to measure individual behavior as well54 (see Figure 6 below). Figure 6. Non-social beliefs Social beliefs Non-normative Empirical beliefs Factual beliefs expectations Personal Normative Normative beliefs normative expectations beliefs 54 Source: Bicchieri et al., 2014 42 Our study in Jordan roughly follows this framework, and explores five key aspects of women’s decision- making vis-à-vis labor force participation- current work status and nature of work, own beliefs and perceptions regarding work and working women, expectations of work-related behavior and workplace characteristics of other women in their reference group, expectations of normative beliefs of other people in their reference group (also male family members, in the case of female respondents) regarding work and working women, and lastly, personal preferences and practical considerations vis-à-vis work. In measuring the influence of social norms on individual behavior, we examine normative influences across four related dimensions: i) general views on whether women should work; ii) beliefs about gender roles as they relate to women’s ability and availability to work; iii) beliefs about publicness and mixing of genders in the work environment; and iv) beliefs about the link between women working and family status. These themes were identified based on barriers identified in the literature, findings from our focus group discussions, analysis of the pilot data. All four thematic groups are explored using a sub-group of questions, which were adapted as needed for male and female respondents (see Figure 7 for more details). Figure 7. 43 For each sub-theme, we first elicit the respondent’s own behavior and direct beliefs, followed by their corresponding social empirical and normative expectations. Table 4. provides an illustrative example of the set of questions we explore under each sub-theme. To understand the extent to which perceptions of what others do and believe affect women’s own (and in the case of men, female relative’s) decisions to work outside the home, we ask men and women direct questions about whether they or their wife/ sister/ daughter work outside the home and whether they think it is okay for women to do so. To measure respondent’s social expectations, we ask them to think about people around them who are most likely to be in their reference group. They are subsequently asked to think about ten such people who are female, married, and working, and estimate how many, out of these women, work outside their homes. Similarly, to measure social normative expectations, we ask respondents to estimate how many people, out of ten, would criticize married working women who work outside. This structure allows us to examine how closely associated our main outcome of interest- women working outside their homes- is to first order expectations of whether others in one’s social group also work outside their homes, and second order expectations of acceptability of working outside the home. Table 4. Personal behavior Personal normative beliefs Do you/your spouse work? Is it okay for women to work outside of their homes? Social empirical expectations Social normative expectations Take a moment to think about the adult women Take a moment to think about all the people where you live. These could include your family where you live. These could include your family members, friends, neighbors, and others. Out of 10 members, friends, neighbors, and others. Out of such women, how many work outside their home? ten such people how many would think or speak badly about married women who, because of work, return home after 5pm in the evening? For women only: Think now for a moment about your husband/ father/ brother, and his views Does he think or speak badly about women who work outside their homes? A key insight from the FGDs and pre-test of the quantitative survey was that there are substantial differences in beliefs and empirical expectations between working women and men. Our qualitative findings indicate that male relatives (particularly fathers, brothers, and husbands) can be particularly influential when it comes to women’s decisions about whether to work. To examine these intra-household dynamics, female respondents are also asked an additional set of normative expectation questions about views held by their male relatives. Given that a subset of our survey sample includes both male and female members from the same household, this will allow us to examine: a) how views and beliefs about various 44 aspects of women’s work differ between male and female respondents within the same household; b) how accurate women are in their perceptions of the views held by their male relatives; and c) how influential women’s expectations of male relative’s normative beliefs are in their own decision-making. To better understand decision-making processes and influencers, respondents are also asked about people they (in the case of men, their female relative) consult for work-related advice and level of support from male relatives. In order to gage physical support from spouses, both male and female respondents are also asked to indicate the share of household chores performed by men. In addition to social expectations, decision-making in this context can be substantially, or at least partially, based on rational considerations. The literature on women’s labor force participation points to various structural and practical barriers to women’s work, including lack of child care options, low wages, limited supply of jobs, etc. To account for these non-social- and mostly structural- considerations women need to take into account when weighing the costs and benefits of whether to work or accept a job, the survey includes modules on current job characteristics (if working), job preferences (ideal job, reservation wage, public vs private jobs, preferred work hours, etc.), reasons for working or not working, and job search effort and intensity, in addition to basic demographic characteristics of respondents, their spouses and their households. In the case of male respondents, these questions primarily focus on a female relative (either wife, daughter, or sister) and her work-related decisions or preferences. Lastly, the survey also includes questions about the value of education for women both in terms of the goals of educational attainment and the status associated with it, women’s perceptions of their own skills, indicators of personal and financial independence (e.g. having driving license, bank account), and a module to measure happiness for all respondents. To examine the cultural mental models and “schemataâ€? that may be shaping behavior, we asked for permission of all respondents in our sample to record open-ended questions regarding their self- understanding and their mental models of themselves and their society. These questions are taken from a classical account of how narratives shape behavior (Burner). Narrative analysis is an implicit way of examining how people experience and interpret the world around them. Narrative psychology posits that human beings deal with experiences by interpreting them through stories- both of their own lives as well as the lives of others. They are inculcated from an early age to make sense of the world by connecting people, which, fundamentally, gives meaning and morality to what otherwise would be a sequence of isolated events (Crossley, 2002). Narratives, in that sense, are thus tied to people’s behavior by acting as the organizing principle for their choices and actions (Sarbin, 1986). Each respondent is asked a random set of narrative questions that explore their descriptions of how typical or atypical they perceive themselves or their relationship with a specified male or female relative to be, how they explain their or their spouse’s motivation to work to others, how they talk to others about work related decisions, and why they think education is important for women’s future. In measuring sensitive beliefs and social norms, it is important to be wary of the possibility of social desirability bias- the tendency of respondents to answer questions in a way that is “appropriateâ€? or “correctâ€?, rather than a true reflection of their beliefs. In our case, respondents may guess that the enumerators would prefer to hear answers that support women’s economic empowerment, or they have an internalized judgment about women who do not work. To overcome this bias and elicit accurate 45 responses, we incentivized responses to some questions by offering rewards if they are able to estimate correctly the number of respondents who express certain beliefs or behave in certain ways. Our enumerators were also all Jordanians, which may go some way toward mitigating social desirability bias. If the behaviors they are being asked about are exemplary, then questions about normative expectations of the relevant community may not illicit realistic assessments about what the respondent thinks that others think and believe. The thought here is that incentivizing accuracy will lead to a more realistic assessment of normative expectations on the respondent’s part.55 Another measurement challenge involves the identification of the relevant reference group. Because social norms are enforced by disapproval (or in extreme cases by socially sanctioned violence, harassment, or ostracism), the views or behavior of distant or irrelevant individuals do not affect the conditional choices of respondents. For instance, the views of Danish men regarding women’s labor force participation would not drive the behavior of Jordanian women. But would the views of people in Amman affect the behavior and social norms of men and women in Zarqa? In the pilot study, we explored various ways to name the relevant reference group for survey respondents. After exploring terms like “in your community,â€? or “in Jordan,â€? we settled on the term “where you liveâ€? when referring the reference group for respondents. A somewhat flexible term is preferred, we believe, because the actual reference group varies, and this allows individuals to imagine the group that matters to them. During the pre-test, we also explored additional variables but decided not to include the following in the final questionnaire: a) implicit attitudes, which were excluded because the behavior in question are likely driven by explicit views and because it was costly to administer an implicit association test; b) vignettes examining the conditionality of work preferences. The latter turned out to involve too many conditions to be useful. For example, social norms related to women’s work in Jordan are not unconditionally strong. The pilot and the focus groups showed that many believe it is appropriate for women to work under certain (often restrictive conditions). Vignettes could not easily capture these. 3.4. Sampling Methodology The quantitative study is a survey of respondents between 20-55 years of age in the urban areas of Amman, Zarqa, and Mafraq. The study was overseen by D3 Systems, and fieldwork was conducted by a local survey firm based in Amman, Jordan. The target sample size was 2,000 with an equal split between men and women. Due to the complexity of the study, and the targeting of a rare population (working women), the sampling methodology was a stratified cluster sample based off an enumeration exercise. The total sample for the survey was allocated disproportionally to the target population of Amman, Mafraq, and Zarqa’s urban areas. With Amman, comprising a very large proportion of the frame (70%), the city was under-sampled by 15 clusters which were transferred to the Mafraq strata, which only comprised 4% of the total sample in the original proportional distribution. 55 Bicchieri (2015: chapter 2). 46 Table 5. Population by Governorate and Sample Distribution Location Total Population Sample Allocation Amman Urban 2,451,792 125 Mafraq Urban 146,785 25 Zarqa Urban 872,443 50 Total 3,471,010 200 Each governorate was divided into localities of which the precise population is known. As the locality is the lowest administrative unit with a known measure of size, it acted as the primary sampling unite (PSU) for the study. To select the localities for inclusion in the sample, a probability proportional to size (PPS) draw was conducted within each governorate. It should be noted that some of localities were grouped together due to their proximity (which is how they are organized by the Jordanian government). Of the 52 possible localities, 43 were selected with certainty due to large PSU size and PPS systematic sampling being implemented. Za’atari was the only locality removed from the frame due to inaccessibility. The secondary sampling unit (SSU) was selected via a gridding exercise where a grid of approximately equal-sized rectangles was placed on the maps of the selected localities to number the grid cells within an area. A number is then generated using a random number generator to randomly select a cell within the grid. To avoid visiting the same households and to expand the geographic area sampled within a locality, the grid cell sampling process is repeated for every time a certain locality is sampled, if it is sampled multiple times. This was often the case due to the large size of some PSU. Following the selection of grid cells, census blocks within those grids acted as the tertiary sampling unit (TSU). Census blocks are a collection of commercial or residential buildings that are marked by the Jordanian Department of Statistics when a census is taking place. Those blocks are often used as bases for sampling and other enumeration activities. Within the selected grid cells, multiple census blocks can be found. The right-hand rule is used to select the first census block on the right within a grid cell. The selected census block is where the field teams goes to conduct the enumeration. One of the main components of this project is to interview a disproportionately high number of working women which is a rare population. Due to the difficulty in reaching this population with a random walk/route procedure, census blocks were enumerated to locate working women. For the enumeration, households were contacted once, and if the contact attempt was successful, basic information was collected to allow for the sampling of enumerated households. The information collected included the following: • Presence of female age 20-55 in the household • Presence of male counterpart age 20-55 in the household (brother, husband, or father) • Working status of female in the household • Availability/willingness to participate in the survey • Address for revisit 47 To shorten the enumeration survey and not ask for the information of all residents of the household (conduct a full household roster), the process of selecting a respondent within the household occurred simultaneously with the enumeration. The following rules were followed to select the man and women to participate. • Women: o Priority of selection was given to the household’s working women as they are considered the rare population. If there was one working woman in the household and one (or multiple) non-working women in the household, the working woman was selected. o If multiple working women were in the household, one was selected randomly. o If there were no working women in the household, but there were multiple non-working women, one non-working woman was selected randomly. • Men: there was no prioritization in the selection of men. In households with multiple eligible men, one was selected randomly. During the enumeration, the field teams attempted to contact 9,664 households. Successful contacts were made with 5,158 of those households. However, to produce a sampling frame from the enumeration, many of those successfully attempted households were removed from the sampling frame to leave only households that fit the project requirements. The filters ran on the data to remove non- eligible households are the following: • Households where no contact was made • Households with refugee residents only • Households with no counterpart (only male or only female households) – this was necessary due to the design of the survey instrument, where a large part of the survey asks women about their male counterparts and vice versa. • Households who refused to participate in any follow-up surveys The table below outlines the results of the enumeration: Table 6. Enumeration Results by Governorate Location Households Households Eligible Households Attempted Contacted Amman Urban 7,041 3,509 1,532 Mafraq Urban 862 440 365 Zarqa Urban 1,761 909 563 Total 9,664 4,858 2,460 Of the eligible households, a stratified (by working status and gender) simple random sample was conducted to allow for the over oversampling of working women. This was the fourth sampling unit (FSU) Replacements were also sent to supplement the sample draw. However, due to the low response rate, 48 the entire draw and replacements were used, resulting in a full census of the eligible households from the enumeration. Those households were revisited a maximum of 4 times to attempt to reach the respondent, given a non-terminal disposition (such as a refusal). 3.4.1. Challenges and Resolutions With a very low women employment rate in Jordan, the main challenge of this study was to oversample working women. During the pilot phase of the project, the household selection process was done using the random walk/route method while implementing quotas to oversample working women. That process was extremely time consuming and inefficient as field teams were spending an extended amount of time trying to find working women. Following that experience, it was clear that oversampling working women via a random walk/route method was not going to be feasible, so it was decided that an enumeration exercise would take place before fieldwork to identify working women and easily oversample them. This method was proven very effective as the final women’s sample contained over 300 working women of 1,000 total women. Another sampling-related challenge was the uncertainty regarding the response rate in the main survey following the enumeration. As sampling from an enumeration was an unprecedented method for the local survey firm, the teams did not have a reliable response rate to help estimate the response rate from the enumeration frame. Coupled with the low total number of enumerated households due to only attempting a household once while enumerating, the field teams had to do a full census of the enumeration frame rather than a sample of the enumeration frame to reach the required sample size of 2,000 respondents. A challenge related to the instrument was the repetitive nature of the questionnaire. Something observed from the pilot study was that respondents were high respondent fatigue as they felt that they were being asked the same questions repeatedly. The differences in the types of questions (personal normative beliefs vs empirical expectations vs intrahousehold bargaining vs social normative beliefs) is minute for a respondent who does not understand the structured approach behind the study. Following the feedback from the pilot, the management teams ensured that the field teams understood the differences between the four main types of questions and offered ways to reassure the respondents that these questions are not the same such as saying, “we understand the next set of questions might be similar to what we have asked before, but they are not, so please listen carefully and answer the specific questions I am askingâ€?. Another strategy used was incentivizing the social normative beliefs section. 49 RESULTS 50 4.1. Sample: Who are our respondents? The quantitative survey was conducted with a sample of 2007 male and female respondents across urban areas of Amman, Mafraq, and Zarqa. The sample was split evenly between male and female respondents, and working women were deliberately oversampled at each location to provide a sufficiently large sample size for analysis. Across all three governorates, 30% of female respondents were working, and 70% of male respondents were working. Of male respondents, 29% had working female relatives (either wife, sister, or daughter), and 71% had non-working female relatives. Nearly 83% of the respondents were interviewed in pairs- father and daughter (7%), brother and sister (17%), or spouses (76%) living in the same household- resulting in 1179 unique households altogether. Table 7. presents a detailed breakdown of the key demographic information of the survey sample, separated by gender and work status of respondents (for women) or respondent’s female counterpart (for men). Table 7. Demographic Information Variables All Women Men N 2007 1061 946 Gender 53% 47% Working 30% 70% M with M w/ non- All Working W Non-working working working W counterpart counterpart N 2007 315 746 275 671 Location Amman 57% 67% 55% 65% 52% Mafraq 13% 13% 13% 13% 14% Zarqa 29% 20% 32% 21% 34% Married (%) 81% 63% 84% 73% 89% % w/ child’s age<6 42% 37% 47% 35% 42% 39.4 34.7 36.0 42.0 44.3 Age (years) (12.34) (8.95) (9.92) (14.48) (13.17) 11.90 13.72 11.32 12.87 11.29 Years of edu. (3.52) (3.25) (3.10) (3.66) (3.63) 1.94 1.57 2.17 1.51 2.02 Avg. no. of children (1.77) (1.67) (1.79) (1.67) (1.79) Avg. HH income 540 717 448 769 463 (JD/month) (499) (731) (316) (733) (329) Note: Standard deviation in parenthesis There are notable demographic differences between working and non-working women, especially when it comes to marital status and children. Across all groups, working women have the highest level of education on average (13.7 years), while men with non-working counterparts have the lowest (11.3 years). While the difference is small, it is worth noting, that educational attainment of working women surpasses that of working men, while non-working women have similar levels of education as their male counterparts. Similarly, marriage and children appear to be important differentiators between working and non-working groups. Sixty-three percent of working women in the sample are married compared to 84% of non-working women. Working women also tend to have lower number of children on average 51 (1.57 compared to 2.17 for non-working women), with less having children below the age of six (37% compared to 47%). Similar trends are also observed among male respondents with working and non- working female counterparts, indicating that both marriage and having younger and more children does depress women’s labor force participation. Male respondents in the sample are older on average (42-44 years) compared to women (around 35 years) because our sample includes father-daughter. Lastly, average household income is significantly higher for working women and men with working female counterparts, but this is most likely due to the effect of having multiple income earners in the household. 52 4.2 Results 4.2.1 Do social norms explain female labor force participation? Decision’s about women’s labor force participation may be influenced by a combination of norms, beliefs, and expectations. Our social norms framework hypothesizes a positive relationship between women’s labor force participation and social norms (both social empirical and social normative expectations) in order to establish a link between women’s work status and social norms. We expect liberal social expectations to be positively linked to women’s labor force p articipation, while conservative expectations are likely to be negatively linked. In addition, it is increasingly clear that intra- household expectations (i.e. expectations of respondents about their counterpart’s beliefs) also matter in women’s labor force participation decisions. Our analysis thus takes into account social norms as well as norms within the household to explore predictors of women’s work-related decisions. For completeness, we also include the respondent’s personal beliefs and attitudes. Equation (i) below regresses women’s work status on a combined standardized index of all beliefs and expectations, including personal, social, and intra-household level variables. This allows to us examine how well liberal beliefs and norms predict the work status of women. Equation (ii) regresses women’s work status (which includes both work status- working or non-working- of female respondents and work status of female counterparts of male respondents) on indices56 of social empirical expectations, social normative expectations, personal beliefs, and normative expectations of counterpart (intra-household expectations): (i) Work status = βo + β1all_indices + βiXi + Æ?i (ii) Work status = βo + β1social_empirical_expectations + β2social_normative_expecations+ β3personal_beliefs + β4intra_household_expecations + βiXi + Æ?i where Xi is gender, age57, years of education, marital status, and whether respondent is from Amman, has a child below the age of 5, and if respondent’s mother ever worked. Equation (ii) allows to us examine the extent to which social and intrahousehold expectations and personal beliefs of respodents predict the work status of women. In addition, separate regressions analysis was done across the four thematic areas: women working, publicness and mixing, gender roles, and status to ascertain which of these matter more when it comes to women’s decision to work. Results (see Table 8 below) of the logistic regression show statistically significant positive relationship between women’s labor force participation and liberal beliefs and expectations (Column 1). This suggests that when society and individuals within the household are more liberal (or perceived as more liberal), women are more likely to work. Results predict that a one standard deviation increase in the index is 56 Indices reflect the average of standardized responses across all relevant variables. Some responses had to be reverse-coded to make sure all variables within each index are measuring the same thing, i.e. either conservative or liberal views. Separate indices were constructed for variables under each of the four themes: women working; publicness and mixing; gender roles; and status. 57 We use age and age-squared as we do not expect age and work status to have a linear relationship. 53 linked to a 1.6 times increase in the odds of women working. Given that female labor force participation in our sample is close to 30%, this represents a fairly large increase. Table 8. Regression results (next page) 54 (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) All WW WW WW PM PM PM GR GR GR ST VARIABLES Working Working Working Working Working Working Working Working Working Working Working All indices 1.598*** (0.107) Social empirical index 1.288*** 1.258*** 1.220*** 1.127* 1.102 1.075 1.343*** 1.245*** 1.265*** 1.004 (0.0644) (0.0635) (0.0701) (0.0620) (0.0613) (0.0672) (0.0678) (0.0660) (0.0786) (0.0644) Social normative index 1.193*** 1.136* 1.028 1.258*** 1.052 0.982 1.358*** 1.160* 1.084 1.027 (0.0599) (0.0582) (0.0617) (0.0679) (0.0638) (0.0703) (0.0711) (0.0673) (0.0743) (0.0591) Personal belief index 1.316*** 1.295*** 1.234*** 1.184* 1.356*** 1.304*** 1.166* (0.0752) (0.0775) (0.0730) (0.0781) (0.0851) (0.0929) (0.0754) Intra-HH norms index 1.098 1.064 1.271*** 1.223** 1.218** 1.202** (0.0554) (0.0589) (0.0769) (0.0788) (0.0740) (0.0811) Female 0.869 0.808** 0.949 0.735*** 0.988 (0.0647) (0.0636) (0.0735) (0.0621) (0.0697) Age 1.078* 1.081* 1.081** 1.090** 1.076* (0.0329) (0.0333) (0.0323) (0.0340) (0.0320) Age-squared 0.999* 0.999* 0.999* 0.999* 0.999 (0.000347) (0.000355) (0.000338) (0.000356) (0.000338) Years of Education 1.140*** 1.175*** 1.163*** 1.163*** 1.174*** (0.0258) (0.0267) (0.0263) (0.0266) (0.0274) Married 0.282*** 0.272*** 0.275*** 0.290*** 0.268*** (0.0641) (0.0616) (0.0615) (0.0663) (0.0596) Have child<6 1.162 1.151 1.173 1.042 1.161 (0.213) (0.210) (0.212) (0.192) (0.209) Amman 1.735*** 1.578** 1.717*** 1.756*** 1.699*** (0.255) (0.231) (0.252) (0.262) (0.250) Mother worked 1.367 1.309 1.454* 1.314 1.374 (0.242) (0.225) (0.257) (0.235) (0.239) Constant 0.0249*** 0.409*** 0.400*** 0.0183*** 0.408*** 0.399*** 0.0173*** 0.401*** 0.401*** 0.0177*** 0.0175*** (0.0174) (0.0204) (0.0203) (0.0129) (0.0204) (0.0204) (0.0120) (0.0204) (0.0214) (0.0126) (0.0121) 55 Clustered errors at HH Yes No No Yes No No Yes No No Yes Yes level Observations 2,003 2,000 1,998 1,994 1,997 1,996 1,992 1,995 1,903 1,901 2,002 seEform in parentheses *** p<0.001, ** p<0.01, * p<0.05 WW=women working; PM=publicness & mixing; GR=gender roles; ST=status 56 Equation (ii) allows us to break down the liberalalism index into its component parts and investigate how their importance varies across the four dimensions of beliefs and expectations we look at. We subsequently show results for the four themes (women working, publicness and mixing, gender roles and status). For each theme, we show the results for the two components that comprise a social norm i.e. social empirical and normative expectations, then we add personal and intrahousehold belief indices, and finally we add controls and correct for correlation from drawing a subset of the respondents from the same household. When it comes to general acceptability of women’s labor force participation, both social empirical expectations and personal beliefs are positive and staistically significant. A one standard deviation increase in the personal belief index is linked to a 1.3 times increase in the odds of women working, while the same for social empirical expectations is linked to a 1.2 times in the odds of women working (See Column 4). This suggests that women’s labor force participation outcomes are associated with what men and women observe around them vis-à-vis working women, as well as their own beliefs about whether women should work. Liberal social norms around working women being out in public and interacting with men in the workplace does not seem to be strongly associated with women’s labor force participation. What does matter, however, are personal beliefs and intra-household normative expectations. Results (Column 7) show that a one standard deviation increase in the personal belief and intrahousehold expectation indices are associated with a 1.2 times increases in the odds of women working. When it comes to gender roles of women within the household, both as a mother and a wife, personal beliefs, spouse’s beliefs as well as social empirical expectations are important predictors of women’s labor force participations . All three variables have statistically significant coefficients, with each associated with an increase in the likelihood of women working by around at a quarter (see Column 10). Personal beliefs are associated with the largest increase in the odds of working (by 1.3 times), while a one standard deviation in intrahousehold expectations index are associated with a 1.2 times increase. Household status is predicted only by intrahousehold expectations (statistically significant at the 10% level), suggesting that women’s decisions to work are linked to their expectations of what their male counterparts (generally the spouse) believe. A one standard deviation increase in the intra-household expectations index is linked to a 1.2 times increase in the likelihood of women working (see Column 11). Together, these results suggest that social norms, particularly expectations of what others do, are likely to be important influencers of women’s labor force participation when it comes to key areas of decision-making, such as general acceptability of women working and gender roles of married women. Both of these have to do with the roles women are traditionally expected to play within society as well as the household. Observing what others do may serve as a cue for both men and women to decipher what roles are acceptable for women to take on, and what behavior is generally endorsed. Household status and publicness and gender-mixing as they relate to women working, on the other hand, seem to be more internal, in the sense of being influenced primarily by intra-household beliefs and expectations. These considerations are closely tied to a man’s role in the 57 household- as a provider, protector, and the head of the household- which provides some explanation as to why intra-household norms appear to be such important predictors in these scenarios. Lastly, people’s own beliefs are important predictors of work across everything, except when asked about status. When personal beliefs and intrahousehold norms are taken out of the equation, however, social normative expectations also become statistically significant predictors. This shows that the respondent’s personal beliefs may be correlated with their belief about others in society. Finally, when women work, they and their spouses may develop more liberal views. In order to overcome the endogeneity issue, we compared the views of those non-working women who want to work, or have looked for work in the past 6 months, and those who have not. We do not expect to see differences since all non-working women we interviewed were randomly selected. However, we find evidence that having more liberal views in the overall index, and in the disaggregated personal belief and intrahousehold expectation indices, leads to increased odds of women wanting to work, or looking for work. This result can be interpreted as the causal impact of these components on women’s intentions and effort towards labor supply. Below we report in greater detail on the questions we asked, and the views that our respondents have of society, for themselves, and of their male counterparts. 4.2.1. Social Empirical Expectations Empirical expectations are informed by first order expectations of what others in one’ reference groups (which, in this case, is either 10 out of all adult men and women, 10 out of all working women, or 10 out of all married working women where one lives) are engaging in, and may be influential if one believes that a sufficient majority of people in in the reference group conform to the norm. The survey el icited respondents’ social empirical expectations across four dimensions related to work, namely, their expectations of others do when it comes: i) women working; ii) publicness and mixing at the workplace; iii) gender role, as they relate to married working women; and iv) status (both financial and social). Table 9. summarizes our findings. 58 Table 9. Social Empirical Expectations Overall N Working N Non- N Men with N Men with N women working working non- women counterpart working counterpart Women Working women work 5.09 1993 6.12 315 5.18 743 5.22 273 4.44 662 ( 2.79) ( 2.75) ( 2.85) ( 2.56) ( 2.66) work outside home 4.65 1991 5.56 314 4.65 742 4.92 273 4.11 662 ( 2.75) ( 2.78) ( 2.81) ( 2.53) ( 2.62) % women work in Jordan 49.27 1991 50.80 312 52.49 742 46.52 271 46.09 666 ( 21.77) ( 22.76) ( 22.82) ( 20.64) ( 19.92) job search months for women 8.28 1963 9.38 311 7.62 738 8.94 270 8.24 644 ( 10.66) ( 9.51) ( 8.49) ( 12.69) ( 12.32) job search months for men 8.17 1982 8.23 310 7.44 742 10.11 272 8.17 658 ( 11.17) ( 12.50) ( 10.82) ( 12.43) ( 10.24) Publicness & Mixing work with men 3.14 1985 3.55 312 2.87 741 3.73 271 3.02 661 ( 2.66) ( 2.69) ( 2.58) ( 2.70) ( 2.67) reputation at risk 1.69 1977 1.60 312 1.57 735 1.51 270 1.95 660 ( 2.43) ( 2.14) ( 2.31) ( 2.26) ( 2.72) harassment 2.62 1974 2.51 310 2.46 734 2.79 273 2.78 657 ( 2.85) ( 2.52) ( 2.73) ( 2.97) ( 3.06) Gender Roles married women work 4.21 1991 4.95 311 4.25 742 4.27 275 3.79 663 ( 2.59) ( 2.51) ( 2.68) ( 2.48) ( 2.51) return after 5pm 1.68 1980 2.19 310 1.61 742 1.97 273 1.40 655 ( 2.03) ( 2.22) ( 1.97) ( 2.12) ( 1.90) leave children relatives 2.62 1973 2.94 311 2.41 737 2.99 271 2.55 654 ( 2.42) ( 2.38) ( 2.37) ( 2.55) ( 2.42) age of child, SE 2.45 1892 1.97 301 2.18 722 2.68 259 2.91 610 ( 2.81) ( 2.32) ( 2.69) ( 2.76) ( 3.11) 59 Status* do not follow traditions Most or some 30.46 322 27.16 85 31.85 237 A few or none 69.54 735 72.84 228 68.15 507 husband cannot provide Most or some 51.52 542 52.40 164 51.15 378 A few or none 48.48 510 47.60 149 48.85 361 husband not in charge Most or some 30.93 326 30.57 96 31.08 230 A few or none 69.07 728 69.43 218 68.92 510 financial need Most or some 83.68 882 85.99 270 82.70 612 A few or none 16.32 172 14.01 44 17.30 128 Notes: Standard deviation in parentheses *Respondents were asked how many working women fall into this category, rather than asking about the number of women out of 10. 60 Women Working Both men and women highly overestimate the level of female labor force participation within their communities and across Jordan. On average, respondents think that 5 out of every 10 adult women where they live are currently working, and generally doing so outside their homes (i.e. not engaged in home-based work or business). Working women have the highest average estimate, 6.12 out of 10, whereas men with non-working counterparts have the lowest (4.44 out of 10). Among the four subgroups, non-working men appear to have the most conservative empirical expectation of women working in their communities. This is also reflected in their expectations about female labor force participation in Jordan in general (49% on average). These responses are somewhat puzzling given the actual labor force participation rate for women in Jordan (14%), but they are nevertheless consistent with our findings from prior qualitative discussions. An important takeaway here is that working women do, in fact, tend to have more liberal expectations of other women in their community, and may even be influenced by it. While our study is not designed to establish causality (in either direction), from a policy standpoint, it is worth noting that for both men and women, liberal social empirical expectations tend to be good predictors of their own behavior vis-à-vis women’s labor force participation. When norms are changing, increasing the visibility and knowledge of these positive changes thus has the potential to induce higher labor force participation for women. Irrespective of gender, empirical expectations about the time it takes for men and women to find jobs is relatively long- 8.3 months on average. Though the difference is relatively small, working women tend to think men find jobs faster than women, while men with working counterparts seem to think the opposite. Publicness and Mixing Though empirical expectations about women’s labor force participation are high, working women are generally not expected to work in environments where most other employees are men. Out of 10 working women around them, respondents expect less than a third to be employed in male-dominated environments. Given that many women work in the education sector, which is predominantly female, this is not altogether surprising. Empirical expectations of non-working women are the lowest (2.87 on average), which implies that non-working women may not expect other women to take up jobs in male- dominated environments. Working women, however, are expected to experience some harassment. On average, respondents expect a fourth of working women to have experienced harassment of some sort either at work or on their way to work. Men’s expectations are marginally higher than women’s irrespective of their counterpart’s work status. Women’s reputation seems to be unaffected by their employment decisions. Respondents do not expect women to be risking their reputation by deciding to work, and on average, expect reputation to be an issue for less than 20% of working women around them. 61 Gender roles The literature suggests that women in Jordan are less likely to participate in the labor force after they are married or have children. We see evidence of this in our data as well as non-working women in our sample are more likely to be married and have more children. Indeed, marriage and children do add additional barriers to labor force participation for women, particularly when it comes to factors like childcare. Empirical expectations about working women change when respondents are asked specifically about married working women. While 5 out of 10 women are expected to work on average, when asked specifically about married women, this estimate drops to 4 out of 10. Working hours are also restricted for married women. More than 8 out of 10 married working women are expected to return home by 5 PM, with estimates especially low for non-working women and men with non-working counterparts. Both male and female respondents think that where there live, working mothers do not leave their child and go to work until the child is at least two years old. Women’s estimates are lower than men’s in this regard by over 6 months, but once again, working women have the lowest estimate (24 months on average) and men with non-working counterparts have the highest estimate (35 months on average). A quarter of working mothers with young children are expected to leave their child with a close relative when they go to work. Status Families with working women tend to have lower status, especially financial, according to our respondents. Eighty-four percent of female respondents 58 expect some or most families of working women to suffer from financial need, while 52% estimate that some or most of these women have husbands who cannot provide for them. Indeed, nearly a third of the respondents also estimate that some of most working women are from families where the husband is not in charge. This suggests that while women joining the labor force maybe considered a solution to mitigating financial need, this does not reflect well on the role of the husband, or lack thereof, in the family. In addition to status, households with working women are also expected to be less traditional . Nearly a third of all female respondents estimate that some or most of the working women where they live tend to be from families that do not abide by traditions. Empirical Expectations of Working Men About Female Co-workers While mixed-gender workplaces are not very common, working men who do work in mixed-gender environments tend to have liberal views on male-female interaction. A quarter of all men in our sample are employed in workplaces which have at least one female employee. On average, 31% of their coworkers (14 people on average) are female. 86% of these men report that their male coworkers interact frequently with female coworkers, and 81% feel that men at their workplace are okay with working alongside women. However, their perceptions about the spouses of their female co-workers is less liberal as they expect around 50% of women to have husbands who are not okay with their wives working in mixed-gender environments. This, however, does not mean that their husbands are opposed to them working altogether. 58 Due to sensitivity concerns, these set of questions were only asked to female survey respondents. 62 Though mixed-gender workplaces involve frequent interaction between men and women, workplace harassment is expected to be low. Over 90% of men who work in such workplaces think that female coworkers are never or rarely harassed by their male coworkers. When asked about childcare options availed by female coworkers, 51% of men estimate that women leave their children at daycare, while 34% estimate that they leave them with a close relative. 63 4.2.2. Personal Beliefs Men, particularly those with non-working counterparts, have much more conservative and patriarchal views and beliefs regarding women’s labor force participation compared to women, and even men with working counterparts. On the opposite end of the spectrum, working women appear to be the most liberal across almost all dimensions of normative beliefs asked about in the survey. Table 10. Personal Beliefs Overall N Working N Non- N Men with N Men with N women working working non-working women counterpart counterpart Women Working Women Work Yes 95.57 1918 96.51 304 97.05 724 94.55 260 93.89 630 Sometimes Y/N, No 4.43 89 3.49 11 2.95 22 5.45 15 6.11 41 Work Outside Home Yes 80.47 1615 95.56 301 88.61 661 84.73 233 62.59 420 Sometimes Y/N, No 19.53 392 4.44 14 11.39 85 15.27 42 37.41 251 Women Right to Work Yes 24.78 497 34.92 110 29.89 223 20.73 57 15.97 107 Sometimes Y/N, No 75.22 1509 65.08 205 70.11 523 79.27 218 84.03 563 Live Comfortably Yes 79.92 1604 89.21 281 84.99 634 81.09 223 69.45 466 Sometimes Y/N, No 20.08 403 10.79 34 15.01 112 18.91 52 30.55 205 Publicness & Mixing Work with Men Yes 38.27 768 57.78 182 45.71 341 35.64 98 21.91 147 Sometimes Y/N, No 61.73 1239 42.22 133 54.29 405 64.36 177 78.09 524 Exposed to Harassment Yes 43.82 876 35.87 113 47.78 355 34.31 94 47.08 314 Sometimes Y/N, No 56.18 1123 64.13 202 52.22 388 65.69 180 52.92 353 Reputations at Risk Yes 34.61 694 29.84 94 35.79 267 23.64 65 40.06 268 64 Sometimes Y/N, No 65.39 1311 70.16 221 64.21 479 76.36 210 59.94 401 Gender Roles Married Women Work Yes 72.35 1452 90.16 284 82.98 619 76.36 210 50.52 339 Sometimes Y/N, No 27.65 555 9.84 31 17.02 127 23.64 65 49.48 332 Return After 5pm Yes 26.21 526 40.63 128 31.90 238 26.18 72 13.11 88 Sometimes Y/N, No 73.79 1481 59.37 187 68.10 508 73.82 203 86.89 583 Leave Children Relatives Yes 53.81 1080 71.75 226 60.05 448 57.09 157 37.11 249 Sometimes Y/N, No 46.19 927 28.25 89 39.95 298 42.91 118 62.89 422 Appropriate Age of Child 4.53 1817 3.44 299 4.42 695 4.55 266 5.24 557 ( 3.03) ( 2.59) ( 2.96) ( 2.85) ( 3.22) Status Do Not Follow Traditions Yes 8.32 167 3.17 10 7.64 57 6.18 17 12.37 83 Sometimes Y/N, No 91.68 1840 96.83 305 92.36 689 93.82 258 87.63 588 Husband Cannot Provide Yes 19.29 387 14.29 45 22.12 165 13.09 36 21.04 141 Sometimes Y/N, No 80.71 1619 85.71 270 77.88 581 86.91 239 78.96 529 Husband Not in Charge Yes 9.77 196 10.48 33 12.60 94 6.55 18 7.60 51 Sometimes Y/N, No 90.23 1811 89.52 282 87.40 652 93.45 257 92.40 620 Financial Need Yes 69.71 1399 68.89 217 70.64 527 68.73 189 69.45 466 Sometimes Y/N, No 30.29 608 31.11 98 29.36 219 31.27 86 30.55 205 Note: Standard deviation in parentheses 65 Women Working Over 95% of survey respondents think it is okay for a woman to work, though fewer consider it acceptable for women to work outside their homes. Women, regardless of work status, are more liberal than men in their views. While nearly all working women, who think it is okay for women to work (97%), are also okay with women working outside their homes, for non-working women, the acceptability drops by 8 percentage points (from 97% to 89%) when asked specifically about working outside the home. For men with working counterparts, this is slightly higher, at around 10 percentage points. However, for men with non-working counterparts, there reduction in acceptability is striking- 31 percentage points (from 94% to 63%)- indicating that a third of these men would only be okay with their female relatives working from home. However, all of this appears to be highly contingent on how comfortable husbands or fathers are with their wives or daughters working. Only a quarter of the survey respondents believe that women should have the right to work even if it makes their husbands or fathers uncomfortable. Though both conservative, there is significant difference in responses between working women and men with non- working counterparts. While 35% of working women think that women should have the right to work irrespective of father or husband’s views, less than half of the men with non-working counterparts (16%) thought the same. Non-working women are also much more liberal compared to their male counterparts, which points to a potential intra-household barrier to labor force participation for women who may be liberal, but have spouses or fathers with more conservative views about women working. Once again, financial considerations appear to be important for women’s labor force participation. Nearly 80% of respondents believe both spouses need to work in order for their families to live comfortably. Once again, men with non-working counterparts are least likely to hold this view as 31% do not agree with this statement. Publicness and Mixing When it comes to working outside the home, publicness of the job and mixed-gender workplaces may pose certain barriers for women. Despite participating in the labor force themselves, not all working women in our sample believe that it is okay for women to work in environments where most other employees are men (58% do so). Non-working women are slightly more conservative (46%), while men are even more so. Only 22% of men with non-working counterparts are okay with women working in mixed-gender environments. This maybe partially attributed to fears about harassment, as 44% respondents believe that working women are exposing themselves to harassment. The estimate is slightly lower for working women and their male counterparts (34%-36%), and higher for non-working women and their counterparts (around 47%). Around one-third of respondents also believe that working women are risking their reputation by working. Surprisingly, while men with working counterparts appear to be most liberal in their views on this (24%), men with non-working counterparts are the least liberal (40%). Together, these beliefs may not only restrict the types of jobs women can take up (i.e. jobs in mixed- gender workplaces), but also give rise to additional concerns about personal safety and reputation when deciding whether to work. 66 Gender roles Working outside the home is not the only factor influencing how okay respondents are with women working. Marriage also significantly lowers the acceptability of women working, especially for male respondents. While 90% of working women and 83% of non-working women believe that it is okay for married women to work, 76% of men with working counterparts and only 51% of men with non-working counterparts find it acceptable. Work hours are also restricted for married women, as less than 40% respondents believe that it is okay for married women to work if they have to return home after 5 PM. Men, once again, are the most conservative in this regard, and only 12% of men with non-working counterparts are okay with married women working late. This further restricts the type of jobs women can apply to, as regular 9-5 jobs would require women to return home after 5 PM. Having younger children can restrict labor force participation for married women as well. According to respondents, a child should be at least 4.5 years old before the mother can leave him or her at daycare (or elsewhere) and go to work. Working women think the appropriate age in around 3.4 years on average, while for men with non-working spouses, the average is even higher at 5.2 years. The same male respondents are also the least okay (37%) with leaving children under 5 years of age with close relatives if the mother has to work. In contrast, 72% of working women and around 60% of non-working women and men with working counterparts find it acceptable. Status How does women’s labor force participation affect perceptions about the family according to respondents? Aside from financial status, women’s decision to work has little influence on how others perceive their families. Families with working women are generally considered to be experiencing financial need by around 70% of respondents. However, this does not necessarily mean that the husband is not contributing sufficiently. Only 19% of respondents associate married women working with their husband’s inability to provide for her (slightly higher for non-working women and their male counterparts); and less than 10% believe that the husband is not in charge in families with working women (slightly higher for women compared to men). Similarly, families of working women are not perceived as less traditional. 67 4.2.3. Social Normative Expectations Social normative expectations are informed by second-order expectations of or beliefs about what others believe, and can be influential if people believe that a sufficient majority of people in their reference group conform to these beliefs and sanction the behavior. Compared to social empirical expectations and personal beliefs, somewhat less variation is observed between the four respondent subgroups when it comes to their social normative expectations. However, across most responses, non-working women seem to expect others in their reference group to have the most conservative beliefs, followed closely by men with non-working counterparts. Men with working counterparts, on the other hand, appear to be have the least conservative social normative expectations across almost all responses. Table 11. summarizes our findings. Table 11. Social Normative Expectations Overall N Working N Non- N Men with N Men w/ non- N women working working working women counterpart counterpart Women Working Women Work 3.68 1997 3.32 313 3.90 743 3.38 275 3.74 666 ( 3.00) ( 2.89) ( 3.09) ( 2.91) ( 2.96) Work Outside Home 3.73 1995 3.42 312 4.01 742 3.24 275 3.78 666 ( 2.99) ( 2.87) ( 3.07) ( 2.91) ( 2.96) Women Right to Work 2.38 1994 2.34 313 2.42 742 2.45 275 2.31 664 ( 2.86) ( 2.79) ( 2.93) ( 2.79) ( 2.86) Publicness & Mixing Work with Men 4.96 1997 4.80 313 5.45 742 3.92 274 4.91 668 ( 3.20) ( 3.15) ( 3.10) ( 2.98) ( 3.30) Exposure to Harassment 4.07 1989 3.73 313 4.29 739 3.41 274 4.27 663 ( 2.95) ( 2.93) ( 2.89) ( 2.77) ( 3.04) Reputations at Risk 3.70 1990 3.38 313 3.92 739 3.17 274 3.82 664 ( 2.89) ( 2.84) ( 2.89) ( 2.66) ( 2.97) Gender Roles Married Women Work 3.66 1995 3.36 313 3.85 739 3.12 275 3.81 668 ( 2.85) ( 2.78) ( 2.88) ( 2.63) ( 2.91) Return After 5pm 5.06 1995 4.64 313 5.20 740 4.28 275 5.43 667 ( 3.22) ( 3.13) ( 3.19) ( 3.10) ( 3.28) 68 Leave Children Relatives 4.19 1996 3.84 313 4.41 742 3.48 275 4.42 666 ( 3.09) ( 2.86) ( 3.13) ( 2.76) ( 3.22) Age of Child, SN 4.17 1855 3.57 298 3.92 680 4.19 262 4.73 615 ( 3.01) ( 2.61) ( 2.95) ( 2.81) ( 3.24) Status* Do Not Follow Traditions Yes (%) 23.58 473 21.90 69 26.04 194 18.18 50 23.85 160 Sometimes Y/N, No (%) 76.42 1533 78.10 246 73.96 551 81.82 225 76.15 511 Husband Cannot Provide Yes (%) 31.72 636 29.52 93 36.38 271 25.45 70 30.15 202 Sometimes Y/N, No (%) 68.28 1369 70.48 222 63.62 474 74.55 205 69.85 468 Husband Not in Charge Yes (%) 24.71 495 24.52 77 29.97 223 19.27 53 21.19 142 Sometimes Y/N, No (%) 75.29 1508 75.48 237 70.03 521 80.73 222 78.81 528 Financial Need Yes (%) 74.98 1504 74.60 235 78.79 587 69.82 192 73.03 490 Sometimes Y/N, No (%) 25.02 502 25.40 80 21.21 158 30.18 83 26.97 181 Note: Standard deviation in parentheses * Respondents were asked if most people think the following about working women and their families, rather than asking how many believe it out of 10 69 Women Working On average, respondents expect a third of the people in their reference groups to hold conservative views about women working. Out of 10 people where they live, respondents expect less than 4 on average to think or speak badly about women who work, with no substantial differences in expectations between women working from home and women working outside of their homes. The difference across the four groups of respondents is very small, with non-working women having the most conservative expectations of people around them. On average, all respondents expect very few people in their references groups (less than a quarter out of 10) to believe that a woman has the right to work even if it makes her husband or father uncomfortable. Publicness and Mixing Non-working women have the most conservative normative expectations of others when it comes to the public and mixed-gender aspects of working for women, while men with working counterparts have the least. On average, respondents expect 5 out 10 people to think or speak badly of women who work in mixed-gender environments, suggesting a perceived bias against mixed-gender workplaces for women. Perceptions about harassment and reputational risk are also not favorable for working women. On average, respondents expect 4 out of 10 people to think that working women are exposing themselves to harassment, and over a third to think that women risk their reputation by working. While the difference across groups is marginal, non-working women have the most conservative expectations about what others think or believe. If these perceptions have any influence on women’s labor force participation decisions, it may provide some insight into understanding women’s work-related decisions. Gender roles While respondents do not perceive that marriage will change others’ views about women’s employment, having younger children is expected to affect this. Expectations about how others perceive married women who work is very similar to their perceived views on women who work in general. What does seem to matter, however, are work hours and younger children. Respondents expect 5 out of 10 people to think or speak badly about a married working woman who returns home after 5 PM, and roughly 4 out of 10 people to do so about women who leave their young children (i.e. under 5 years of age) with close relatives. Men with non-working counterparts expect others to have more conservative views. Men’s social normative expectations about the appropriate age at which a mother can leave her child and go to work is higher than women’s. Women, working or otherwise, estimate that according to others, the appropriate age is between 3.5-4 years. For men, however, the expectation is closer to 4.5 years. Status When families have working women, they are perceived by others to be struggling financially. Seventy- five percent of respondents expect most other people to think that families with married working women are suffering from financial need. 25-30% of working women and their male counterparts expect most other people to also think if a woman works, it means that the husband cannot provide for her. The estimate is slightly higher (between 30-36%) for non-working women and their male counterparts. Lastly, roughly a quarter of respondents expect most other people to think that the husband is not the one in charge of the family if his wife is working. In addition to status, a small proportion of respondents also expect people in their reference group to consider families with working women to be less traditional. Compared to working women and their counterparts, non-working women and men with non-working counterparts tend to expect slightly more people to hold such views. 4.2.4. Intra-Household Beliefs The survey also includes questions to elicit the mutual expectations of male-female pairs in the household. Intra-household power dynamics may potentially constrain women’s economic participation even independently of social norms. Social norms, in such instances, may be primarily enforced with the household, rather than through generalized disapproval, gossip or harassment. If that is the case, male- female pairs’ expectations of each other’s’ beliefs can provide important insights into how liberal or conservative they think their counterparts to be. Table 12. Intra-Household Normative Expectations Working N Non- N Men w/ N Men w/ N Variables women working working non-working women counterpart counterpart Working Women Work Inside Home Yes 7.30 23 8.59 64 . 0 . 0 Sometimes Y/N, No 92.70 292 91.41 681 . 0 . 0 Work Outside Home Yes 9.84 31 24.40 182 . 0 . 0 Sometimes Y/N, No 90.16 284 75.60 564 . 0 . 0 Women Right to Work Yes 23.17 73 20.86 155 24.82 68 20.75 139 Sometimes Y/N, No 76.83 242 79.14 588 75.18 206 79.25 531 Publicness & Mixing Work with Men Yes 26.67 84 41.13 306 . 0 . 0 Sometimes Y/N, No 73.33 231 58.87 438 . 0 . 0 Harassment Yes 33.65 106 50.61 376 26.18 72 38.02 254 Sometimes Y/N, No 66.35 209 49.39 367 73.82 203 61.98 414 Reputations at Risk Yes 27.48 86 44.83 334 17.82 49 31.24 209 Sometimes Y/N, No 72.52 227 55.17 411 82.18 226 68.76 460 Gender Roles Married Women Work Yes 9.84 31 22.28 166 . 0 . 0 Sometimes Y/N, No 90.16 284 77.72 579 . 0 . 0 Return After 5pm Yes 30.16 95 48.05 357 . 0 . 0 Sometimes Y/N, No 69.84 220 51.95 386 . 0 . 0 Leave Child w/ Relatives Yes 47.62 150 47.78 355 . 0 . 0 Sometimes Y/N, No 52.38 165 52.22 388 . 0 . 0 Appropriate Age of Child 3.87 285 5.11 587 4.09 266 5.00 589 (2.70) (3.12) (2.78) (3.11) Status Do Not Follow Traditions Yes 5.40 17 14.96 111 6.55 18 9.87 66 Sometimes Y/N, No 94.60 298 85.04 631 93.45 257 90.13 603 Husband Cannot Provide Yes 13.38 42 25.57 190 10.18 28 18.78 126 Sometimes Y/N, No 86.62 272 74.43 553 89.82 247 81.22 545 Husband Not in Charge Yes 11.15 35 17.74 132 6.91 19 9.69 65 Sometimes Y/N, No 88.85 279 82.26 612 93.09 256 90.31 606 Financial Need Yes 71.75 226 75.13 559 70.55 194 71.54 480 Sometimes Y/N, No 28.25 89 24.87 185 29.45 81 28.46 191 Women Working Men and women are not very different in their perceptions of each other when it comes to women working, though working women tend to have more liberal expectations than their male counterparts. Working women expect men to be very liberal when it comes to women working, both in and outside of the home (which is not surprising given that they, themselves, work, and generally do so outside the home). However, a quarter of the non-working women think that their male counterparts would think or speak badly about women who work outside their homes. Men and women’s expectations of each other when it comes to women’s right to work even if it makes her husband or father uncomfortable are very similar, with around 20% of non-working women and their counterparts and between 23-25% of working women in their counterparts expecting their counterparts to hold such views. Publicness and Mixing Non-working women expect their male counterparts to have much more conservative views about mixing with men and being outside of the home for work and vice versa. More than 40% of non-working women expect their male counterparts to think or speak badly about women who work in mixed-gender workplaces, compared to 27% of women. The same trend is also observed when it comes to exposure to harassment and reputational risks that working women face. While almost half of all non-working indicate that their male counterparts think working women are exposing themselves to harassment and putting their reputation in line, a third of less of working women think so. Men, on the other hand, expect their counterparts to have less stricter views on these issues. Thirty-eight percent and 31% of men with non- working counterparts expect their female counterparts to associate working with exposure to harassment and reputational risks for women, while 26% and 18% of men with working female counterparts think so. Gender roles While women do not perceive marriage to be a major issue for their counterparts when it comes women’s work decisions, long work hours and children are expected to restrict their approval of working women. Less than a quarter of non-working women expect their counterparts to think or speak unfavorably of a married woman who work (as opposed to 10% of working women), but nearly half expect them to do so when a married woman return home after 5 PM. Thirty percent of working women also expect their counterparts to feel the same way. However, when it comes to leaving young children with relatives, nearly half of both working and non-working women expect their counterparts to not to be okay with working mothers doing that. Working women expect their counterparts to consider 4 years to be the appropriate age to leave a child at daycare or elsewhere, while non-working women estimate this to be just over 5 years. Men’s response to the same question about their counterparts is comparable. Status Both men and women, especially working women and their counterparts, do not generally expect their counterparts to think that families with working women are less traditional or that their husband is unable to provide or not in charge. Just over 25% non-working women expect their counterparts to think that working women have husbands who cannot provide, while 15% expect them to think that husbands are not in charge in such families. Men’s estimates of their female counterpart’s views are slightly lower. Financial need, however, is expected to be high for families of working women. Across all four groups, more than 70% of men and women expect their counterparts to think that families of married working women are experiencing financial difficulties. 4.2.5. Expectation vs. Reality Social Norms What does the data tell us about how accurate people’s social expectations are? While intra-household norms may be easier to observe due to proximity and regular interactions between the individuals in question, the same may not be true for everyone else in people’s reference groups. Even when people think about others in their reference networks, e.g. those who live around them, their observations may not be acute enough for them to capture the full spectrum of others’ behaviors and beliefs. To begin with, action tends to be more noticeable than inaction. Certain actions (such as a woman going to work) may be more salient as a result of being more observable. They could stand out more in people’s memories precisely because they are unusual. Availability bias, i.e. the tendency people have to rely on immediate experiences or examples that come to mind when they try to think about a certain topic or incident, may further contribute to people underestimating or overestimating a social norm (this is especially true for empirical expectations). Even though female labor participation is only 14%, on average, all respondent groups estimate it to be closer to 50%. This is a huge overestimation, and is puzzling given what we know to be the case in Jordan. Why then do men and women, irrespective of work status, have such high empirical expectations of women working? One possible explanation could be availability bias. During the focus group discussions, several participants explained their high estimates of women’s labor force participation by saying that “I see women working everywhere these daysâ€?. This, however, may not actually mean that they see women working everywhere. Instead, it is possible that when people do see women working, they take notice because it is unusual. Subsequently, when they are asked if and how many women work in Jordan, their mind drifts back to the instances where they did see a woman working, leading to an overestimation of the extent to which this happens. Another explanation may have to do with perceived intent rather than action. As discussed earlier, around 60% of non-working women in our sample indicate that they want to work. If this knowledge is common, it may lead people to consider the ideal of women working to be more common than it is. On the other hand, social normative expectations about whether other people are okay with women working suggest that respondents, particularly non-working women and men with non-working counterparts, overestimate the level of conservatism around them. While all respondents expect over a third of other people to think or speak badly about women who work, both in and outside the home, in reality, that number is likely to be much lower. Aggregated personal beliefs of respondents (i.e. Actual Belief: All in Figure 8 below) suggest that only 3.1% of all men and women in our sample are opposed to women working, while 14.8% are opposed to women working outside their homes. Figure 8. Social Normative Expectations: Women Working 2.5 2.3 Right to Work 2.4 2.4 2.3 1.5 3.8 Work Outside Home 3.2 4.0 3.4 0.3 3.7 Women Work 3.3 3.9 3.3 0 1 2 3 4 5 6 7 8 9 10 Number of people (out of 10) Actual Belief: All Men without working counterpart Men with working counterpart Non-working women Working women Social normative expectations about acceptability of women working in mixed-gender workplaces as well as harassment and reputational risks are more accurate, as shown in Figure 9 below. Figure 9. Social Normative Expectations: Publicness & Mixing 3.5 3.8 Reputations 3.2 3.9 3.4 4.4 4.3 Harrassment 3.4 4.3 3.7 5.1 4.9 Work with Men 3.9 5.5 4.8 0 1 2 3 4 5 6 7 8 9 10 Number of people (out of 10) Actual Belief: All Men without working counterpart Men with working counterpart Non-working women Working women When it comes to married women working, respondents once again overestimate how conservative other people’s views are. Respondents expect around a third of other people to think or speak badly about married women who work, but 19.4% are actually opposed to the idea of a married woman working. In contrast, respondents underestimate the extent to which people disapprove of married working women returning home after 5 PM. While around 50% of other people are expected to think or speak badly about such women, 64.7% of the respondents report their disapproval for married women working past 5 PM. Figure 10. Social Normative Expectations: Gender Roles 4.5 4.7 Age of Child (years) 4.2 3.9 3.6 3.7 4.4 Leave Children Relatives 3.5 4.4 3.8 6.5 5.4 Return After 5pm 4.3 5.2 4.6 1.9 3.8 Married Women Work 3.1 3.9 3.4 0 1 2 3 4 5 6 7 8 9 10 Number of people (out of 10) Actual Belief: All Men without working counterpart Men with working counterpart Non-working women Working women Though respondents do not generally think families with working women are less traditional or have husbands who are less financially responsible or not in charge, they are much more likely to expect other people to hold such views. For example, 23% of respondents expect others to consider families of working women as less traditional, and 25% expect others to think that husbands in such families are not the ones in charge. In reality, only 8.32% and 9.77% of people (based on respondents’ own beliefs) think so. Figure 11. Social Normative Expectations: Status 69.7 73.0 Financial Need 69.8 78.8 74.6 9.8 21.2 Husband Not in Charge 19.3 30.0 24.5 19.3 30.2 Husbant Can't Provide 25.5 36.4 29.5 8.3 23.9 Don't Follow Traditions 18.2 26.0 21.9 0 10 20 30 40 50 60 70 80 90 100 % of people Actual Belief: All Men without working counterpart Men with working counterpart Non-working women Working women Respondents are not altogether inaccurate in their expectations of what other people believe, but the difference between expectation and reality does appear to be relatively large when it comes to some key beliefs around women’s labor force participation. Even though respondents expect women’s labor force participation to be higher than what it is in reality, they underestimate the strength of social beliefs against women, married or otherwise, working in or outside the home. Similarly, respondents’ expectations of how others view the husband’s status in families that working women exaggerate the extent to which the husband’s role and control are undermined in such households. Given that in the MNA region, men are traditionally seen as the providers and the ones in control of the household, this may pose a barrier to women’s labor force participation if husbands (as well as their wives) feel that their masculinity will threatened if the wives work. Intra-Household Expectations At the household level, how accurate are people’s expectations of what their counterparts think? A comparison, at the household level, of the actual personal beliefs of men and women and their expected beliefs according to their counterparts can tell us how well key decision-makers within the household understand each other’s beliefs and where discrepancies exist. Accuracy is measured by whether the respondent’s expectation of their counterpart’s belief matches exactly with the counterpart’s reported personal beliefs. Since majority of our survey respondents were paired from the same household, our data allow us to make direct comparisons of expectations vs. reality. Table 13. Expectation vs. Reality of Counterpart’s Beliefs Within the Household Women’s Men’s Men’s Men’s Women’s Women’s Variables Beliefs Expectation Accuracy Beliefs Expectations Accuracy Working Women Women Work * Yes 2.05 N/A 4.71 8.34 85.49 Sometimes Y/N, No 97.95 95.29 91.66 Work Outside Home* Yes 6.52 N/A 23.67 18.72 69.32 Sometimes Y/N, No 93.48 76.33 81.28 Women Right to Work Yes 29.95 21.77 63.12 16.57 19.37 67.03 Sometimes Y/N, No 70.05 78.23 83.43 80.63 Publicness & Mixing Work with Men* Yes 42.03 N/A 59.78 34.95 45.34 Sometimes Y/N, No 57.97 40.22 65.05 Harassment Yes 39.47 32.97 57.35 41.50 42.44 54.56 Sometimes Y/N, No 60.53 67.03 58.50 57.56 Reputations at Risk Yes 31.64 26.51 56.17 34.50 37.73 51.39 Sometimes Y/N, No 68.36 73.49 65.50 62.27 Gender Roles Married Women Work* Yes 10.02 N/A 30.19 17.87 62.32 Sometimes Y/N, No 89.98 69.81 82.13 Return After 5pm* Yes 59.78 N/A 70.41 42.15 45.65 Sometimes Y/N, No 40.22 29.59 57.85 Leave Child w/ Relatives Yes 62.80 N/A 41.55 48.43 47.09 Sometimes Y/N, No 37.20 58.45 51.57 Status Do Not Follow Traditions Yes 5.92 8.83 81.74 11.11 11.76 77.46 Sometimes Y/N, No 94.08 91.17 88.89 88.24 Husband Cannot Provide Yes 19.81 16.06 60.87 18.26 19.83 62.1 Sometimes Y/N, No 80.19 83.94 81.74 80.17 Husband Not in Charge Yes 12.32 8.82 72.95 7.49 14.77 73.73 Sometimes Y/N, No 87.68 91.18 92.51 85.23 Financial Need Yes 69.81 71.01 60.26 68.36 73.52 59.73 Sometimes Y/N, No 30.19 28.99 31.64 26.48 *These questions were asked in the reverse order, i.e. they asked if counterpart would think or speak badly of women who do work/work outside/return after 5, etc. The answers to the personal beliefs subsequently had to be revered to match the responses to the intra-household expectation questions. On average, accuracy of expectations is around 60% for both male and female respondents, indicating that 60% of men and women in our sample are more or less aware of how liberal or conservative their counterparts. One notable exception is noted with regards to women’s expectations of their counterpart’s views on women working in mixed-gender environments. While 35% of women expect their counterparts to disapprove of women working in mixed-gender environments, nearly 60% of their male counterparts admit to holding such views. Similarly, 42% of women expect their counterparts to disapprove of a married woman returning from work after 5PM. In reality, 70% of men were against this. Women, in these cases, appear to overestimate how liberal their counterparts are. 4.3 Women’s preferences and decision making towards work In the following sections, we report descriptive summaries on where women in our sample are currently working, what are their reasons for choosing to work or not to work, job preferences and wages for both men and women in our sample, and decision making in the household with respect to women’s labor force participation. These statistics are unweighted, and therefore not representative of men and women in Jordan in general. 4.3.1. Employment: Work Status, Sector, and Type of Job Working women are primarily in paid employment in the private sector. By design, nearly 30% of the female respondents in our sample are currently working. Overall, 25% are in paid employment, 4% are self-employed, and 9% are unemployed, but looking for work. Majority (just over half) are currently home makers, while 3% are not looking to work. Not surprisingly, the employment rates are much higher for men- nearly 70%- and about 13% are looking for work . Eighty-four percent of working women work outside the home, compared to 98% of men. Across both genders, employment is primarily in the private sector, with slightly higher proportion of working men employed in the public sector (30%) compared to working women (28%). This is somewhat surprising given that women largely tend to prefer jobs in the public sector due to better wages (and less wage discrimination) and favorable working conditions. However, given the slow job growth in the public sector, it is likely that demand for jobs in this sector is much higher than the supply. We do see strong evidence of this in Section 4.3.2. There is stark difference between men and women when it comes to occupational choices, as well as salary. While a third of the working women are in the education sector, men are more dominant in the service sector and skilled labor work. For women, the second most frequent occupation are clerical or office work (10%) and unskilled manual labor (10%). Working women earn 306 JD (USD 431.29) on average per month. Men’s average income is much higher at 445 JD (USD 627.21) on average per month. Figure 12. Type of Job Other 1% 11% Unskilled manual labour (Street vendors/ clenars / doormen /… 10% 11% Home and restaurant services/ Housekeepers/ Cooks/ Waiters 3% 3% Drivers/ Small vehicles’ drivers/ Bus drivers/ Drivers of trucks… 0% 15% Skilled manual labor / Builders/ Carpenters/ Painters/… 8% 16% Sales and services workers in personal Sevices/ Hair Stylists/… 8% 15% Workers in Personal Care/ Child Care/ In-Home personal care 3% 2% Customer Services/ Telephone operators/ Reception,… 1% 5% Clerical employees /Office work/ Secretary/ Data processing/… 10% 0% Social sciences/ Economists/ Sociology/ Philosophy/ History/… 1% 1% Judicial / Lawyers/ Judges 1% 1% Business/ Accountants/ Human resources 4% 6% Education Learning skills/ Inspectors/ Special education 5% 1% Education Basic and before/ 24% 2% Education / Secondary/ 6% 1% Education/ Higher education/ / 1% 0% Health/Nurses/ Midwives 5% 2% Health/ Physicians/ Dentists/Veterinarians/ Pharmacists 2% 1% Life sciences / Biology, Plant and Animal / Medicine and… 0% 0% Engineering/ Engineers / Architects 2% 2% Computers/ Programmers/ Analysists/Computer Engineers 1% 1% Natural sciences/ Physicists/ Chemists/ Statisticians 1% 0% Legislators/ Senior managers/ General managers/ CEOs 4% 4% 0% 5% 10% 15% 20% 25% Female Male For women, structural features of the job, such as availability of childcare appear to be important. While only 5% of men work with employers who provide childcare, 27% of working women do so. Child care options for working women with children below the age of five include close relatives (44%) and daycare (23%). Almost forty percent commute by car (half self-driven), 24% use public transportation or shared taxis, and 14% use private taxis. On average, the commute time for working women is around 28 minutes. Women who work primarily do so due to financial reasons, while those who do not, prefer taking care of children and household over working. Working women in our sample primarily need to work due to financial reasons (65%), or because they want to be financially independent (14%) or “make something of her lifeâ€? (13%). Men’s responses, when asked the same question about their female relatives, are roughly comparable. When non-working women were asked why they do not work, 37% indicated that they prefer to take care of their children and household, 22% cannot find good opportunities, and 19% were discouraged from working by their husbands. While men’s responses were somewhat similar, they tend to overestimate the extent to which women prefer to stay at home to take care of children and the household (42%), and underestimate the role of discouragement from men (14%), i.e. themselves . Figure 13. Reason for Working 2% Other 1% 6% To use free time 6% 14% To be financially independent 11% 13% To make something of her life 19% 65% Need to work for financial reasons 63% 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 70.00% Female Male Figure 14. Reason for not working Other 4% 6% Due to health reasons 7% 8% Cannot find jobs/no good opportunities 22% 20% Prefer to take care of child(ren) and household 37% 42% Salary is too low 2% 1% Do not want to work at mixed gender places 1% 1% Worried about exploitation at work 1% 0% Discouraged from working by parents 2% 2% Discouraged from working by husband 19% 14% Still studying 5% 6% 0% 5% 10% 15% 20% 25% 30% 35% 40% 45% Female Male If women are not currently working, it does not necessarily mean that they have never worked in the past. Indeed, 47% of all female respondents, including those who are currently working, were employed at some point in their lives, indicating that 17% dropped out of the labor force at some point (or are currently unemployed). In contrast, 73% of male respondents indicate that they have worked, which includes 70% who are currently working. This means that while women have a tendency to drop out of the labor force as they get older, men, once they have entered the labor force, continue to stay on. For women, the biggest reasons for dropping out of the labor force are marriage (26%) and having children (19%). Nineteen percent also quit due to the job being too stressful/not enjoyable or because they were laid off. Once again, we see a slight difference in men and women’s responses to this question, with men underestimating the potential role of marriage and overestimating the role of children relative to women in explaining women’s decision to drop out of the labor force. Figure 15 Reason to stop working Other 15% 19% Heath conditions deteriorated 7% 4% Financial situation improved 1% 1% Parents said no to working 2% 1% Husband said no to working 2% She got married 26% 21% She had children 19% 26% She was not able to succeed at her job 3% 1% Job became too stressful or tiring or not enjoyable She was 19% laid off 19% She was too old/retirement 2% 3% She felt exploited 4% 5% 0% 5% 10% 15% 20% 25% 30% Female Male Similarly, non-working women are not altogether inactive in their effort to find work. Twenty-four percent indicate that they have asked someone about job prospects in the past six months, with more than 80% inquiring more than once. Across all female respondents (including working and non-working women), 18% have looked for jobs in the past six months, among which 19% contacted job centers close to 5 times on average, and 39% sent CVs to approximately 8 potential employers on average. Based on experience of from working women who have had to look for jobs in the past, finding a job in Jordan takes approximately 2.5 months on average. 4.3.2. Job Preferences Low labor force participation and high unemployment among women in Jordan are frequently attributed to a range of structural barriers, such as low wages, limited job growth, lack of safe transportation, limited childcare options, etc. It is useful to understand what women’s preferences are, and in what sectors. The survey asked both female and male respondents about job preferences for women. While women were asked directly about their own preferences as well as what they thought their male relative would prefer for them, men were asked about their preferences when it comes to jobs for their female relatives. When it comes to structural aspects of a job, men and women have broadly similar preferences. As shown in Table 14 below, reservation salary for women, according to both men and women, is around 300 JD (USD 422.83) per month, which is close to the average monthly salary of women in our sample. What respondents consider a good salary for themselves is considerably higher, at around 462 JD (USD 651.18) per month. The majority (82%) are willing to take public transportation, and thought the longest acceptable commute time is relatively low- 32 minutes on average. Confirming our priors from Section 4.3.1., public sector jobs are strongly preferred over private sector jobs, even though both men and women agree that the latter pays higher wages. Men have a slightly higher preference for public sector jobs for their female relatives, despite a larger proportion (54%) agreeing that private sector jobs pay more. Table 14. Job Preferences Variables Women preference Men’s preference for Women’s expectation for themselves female relative of men’s preference Minimum monthly salary 293 307 311 (JD/month) (128) (142) (135) Good monthly salary 459 466 463 (JD) (257) (229) (236) Longest acceptable 32 32 commute (mins) (18) (47) Willing to take public 82% 84% 81% transportation Work sector Public 65% 69% Not asked Private 27% 20% Either 8% 10% Where are wages higher? Public sector 46% 36% Not asked Private sector 47% 54% About the same 7% 10% Note: Standard deviation in parenthesis Women’s occupational preferences are very different from what they actually do and what their male relatives would prefer they do. A third of all working female respondents are currently employed in the education sector, even though only 19% indicate jobs in this sector to be “ideal jobsâ€?. In contrast, 30% of male respondents find jobs in the education sector to be ideal for their female relatives. Similarly, while only 8% of working women are in the sale and service industry, 14% consider jobs in this sector to be ideal, compared to 8% of men. This shows that despite having different preferences, women’s occupational choices tend to be more closely aligned to occupational preferences that male relatives hold for them. Figure 16. Ideal Job 160 146 146 140 120 109 108 107 102 96 100 80 73 63 60 62 57 60 52 50 49 45 42 43 43 45 37 40 33 18 27 26 21 23 21 12 17 18 18 17 20 14 12 12 85 10 10 8 10 5 6 4 11 2 2 0 4.3.3. Decision making According to both men and women, male relatives are extremely influential when it comes to decisions about women’s employment. Around 90% suggest that opinions of male relatives are either important or extremely important, with 9% of men reporting that they actually get to make the final decision for their female relative (5% of women hold that view, however). Support from male relatives also appears to be critical as 88% of working women and 84% of men with working female relatives indicate that they are supportive of the woman working. In contrast, 46% of non-working women in our sample have male relatives who would oppose their decision to work. This is consistent with men’s answer to the same question regarding their non-working female relatives, where 42% indicate that they would oppose. However, it is worth noting that over 41% of both men and women also indicate that men would be supportive if women who are currently not working decided to work. Figure 17. Husband’s Support for Work Non-Working Women: Partner's Support if Working Men with Working Partner: Support for Work Strongly support it 8% Strongly support it 17% Support it 33% Support it 68% Neither oppose nor support it 13% Neither oppose nor support it 10% Oppose it 18% Oppose it 4% Strongly oppose it 28% Strongly oppose it 1% 0.00% 20.00% 40.00% 60.00% 80.00%100.00% 0.00% 20.00% 40.00% 60.00% 80.00%100.00% Men with Non-working Partner: Support if Working Working Women: Partner's Support for Work Strongly support it 26% Strongly support it 7% Support it 62% Support it 36% Neither oppose… 7% Neither oppose nor support it 16% Oppose it 4% Oppose it 21% Strongly oppose it 0% Strongly oppose it 21% 0.00% 20.00% 40.00% 60.00% 80.00% 100.00% 0.00% 20.00% 40.00% 60.00% 80.00%100.00% Non-working women in our sample are interested to work, but not all men may be aware of this. Sixty percent of non-working women in our sample indicate that they would like to work; however, when men are asked if their non-working female relatives want to work, 46% think they do. This points to a potential discrepancy in men’s perceptions of what women want and what women actually want, which is important to consider given the influential role men ultimately play in women’s decision-making process. 4.3.4. Education Women’s job preferences are somewhat reflected in their educational background. While over a third of the female respondents have not studied beyond high school, of those that did, Economics and Management (11%), Literature and Humanities (10%), and Education (10%) are the most frequently studied subjects. Figure 18. Main Goal of Education Women Men 2% 4% Can find work more easily Can help with household finances 16% 20% Helps with better marriage prospects 35% 32% Improves the status my family 10% Can bring up children better Can be financially independent 17% Can develop more confidence 8% 7% 19% Others: 11% 6% 12% 1% 2% While women see education as a means to improve employment opportunities, financial independence, and confidence for themselves, men also value education (for women) for its positive impact on child rearing and improving the family status. This suggests that to men, having educated female relatives has benefits beyond her own development as nearly 31% associate it primarily with improved family status and better parenting (compared to 17% of women). This, taking into account the potentila influence of male relatives on women’s employment-related decision-making, may explain why despite having college degrees, many women still refrain from entering the labor force. 4.3.5. Financial Independence and Behavior Irrespective of work status, women’s financial independence is relatively low in our sample. Seventy- six percent of women do not have bank accounts: this includes 44% of working women and 90% of non- working women. Over 70% of women, irrespective of work status, do receive monthly allowances from their spouses for personal or household/children’s expenses, though not all receive allowances for both. Table 15. Monthly Allowance from Husband Variables Working Women Non-Working Women Allowance for women 59 / 83 52 / 73 (JD/USD month) (51) / (72) (44) / (62) Allowance for Children 276 / 389 250 / 352 (JD/USD month) (209) / (295) (207) / (292) Note: Standard deviation in parenthesis When women work however, their spending is more evenly split between household/children expenses and personal expenses. Working women spend on average 31% of their income on personal expenses, and 37% on household and children’s expenses. One-tenth on their income on average is saved, and around 7% is shared with their husbands. When men are asked to estimate spending patterns of working female relatives, they tend to slightly overestimate the proportion of income women spend on themselves (37% on average). 4.3.6. Share of Household Chores Men’s contribution to household chores is negligible, regardless of women’s work status, except when it comes to feeding children. According to both men and women, most men (around one-fifth) do not contribute to chores such as cooking, washing dishes, cleaning, or taking children to school. Figure 19. Share of HH Chores Performed by Men 16% Taking children to school 16% 42% Feeding children 41% 18% Cleaning/laundry 12% 23% Cooking/dishes 21% Male Female Lab-in-the-Field: The challenge of light touch interventions in affecting behavioral change The lab in the field experiment attempted to assess whether a positive portrayal of working women (such as through a 90 second docufilm) could in turn motivate female respondents to seek information about work. However, the job advisor received only 64 calls (from men and women) from April 4 – May 16, 2018 and of that only 17% saw the video. Many of the callers were under the assumption that the jobs advisor would offer them a job or help them fund-raise for home-based businesses, despite the disclaimer on the flyer stating the advisor will only provide advice and guidance to callers Nevertheless, majority of callers accepted the advice offered by the advisor. Less than 20% of calls were from men most of whom called on behalf of their spouses or female family members to see if jobs were available. Women who called did so looking for work either for themselves, their children and in some instances for husbands, or for support in starting or building home-based businesses. Many callers (women and men) expressed desire to increase household income citing increased cost of living, rent, loans or debts, university tuition for their children and medical expenses as some of the reasons for their search. Some women asked for jobs that have flexible or short working hours while others reported to prefer to work from home or telecommute so as to only work a few hours a day, demonstrating limited understanding that home- based work could be as time-consuming as full-time work unless made explicit by the employer. For female callers asking about home-based businesses, the majority did not know what kind of business they could do from home and were interested in learning about options. Overall, most callers (female and male) did not associate the prospect of a double-income household with the ability to purchase (quality) services such as child-care and transportation that could facilitate day to day household needs. Overall, results from the light touch intervention were weak with only few calls (6.4%) and a reverse causality taking place: majority of callers were from outside of the treatment group. Many had not seen the video, demonstrating a weak link, if any, between the light-touch intervention and seeking employment. Explanation for this could vary: (i) that the video was too quick of an intervention to have an effect; (ii) that family’s circumstances in the video did not resonate with the respondents who came from a variety of socio-economic backgrounds, and (iii) that in fact results reinforce overall analysis that men and women’s attitudes towards women working is not significantly impacted by what societal expectations of whether women are actually working or perceived to be behaving when it comes to work. RECOMMENDATIONS 5. Recommendations What can the World Bank do to be more effective in supporting female labor force participation? What has this study taught us and how can we support the GoJ and sectoral teams to design interventions that have meaningful and sustainable impact? If the GoJ’s is fully committed to promote female economic empowerment, then it is necessary that proposed structural policy changes be accompanied by efforts that contribute to strengthening women’s bargaining position with their male counterparts, and invest in ‘changing mindsets’ activities. Furthermore, packaging the study’s key findings and recommendations will be equally important. Understanding the conceptual underpinnings of measuring social norms work is complex therefore it is important that the study be translated into an easily digestible format. Based on lessons from similar studies, it is recommended that the study be converted into a short brief, an infographic, and a 15-20 pages publication to be translated into Arabic for wider dissemination and reach. This section provides three-tiered recommendations specific to Jordan structured around: (i) policy level (for both the World Bank teams and the GoJ); (ii) sector-specific recommendations, and (iii) media campaign interventions coupled with broader recommendations to be considered for other countries in the MNA region. 5.1. Policy-level Recommendations The policy level recommendations are meant to address structural barriers impeding women’s active labor force participation and to do so within the context of social norms. While many of these were acknowledged prior to the study, the study has confirmed that in fact, these factors do contribute more significantly than expected in preventing women’s participation. The study has however provided a greater depth to some of these issues (ie., childcare: it is insufficient to provide affordable and reliable childcares if not accompanied by efforts to change stigma around mothers who leave their children in childcare before the age of 4.5). • Childcare: Having younger children can restrict labor force participation for married women as well. According to respondents, a child should be at least 4.5 years old before the mother can leave him or her at daycare (or elsewhere) and go to work. Working women think the appropriate age in around 3.4 years on average, while for men with non-working spouses, the average is even higher at 5.2 years. A study by Sadaqa demonstrated that vouchers for childcare were ineffective and that what was important for working women was proximity to their kids. Recommendation: increase access to affordable child care by complying with article 72 in the labor law to establish day care centers at the workplace and include men in the number of employees counted for the establishment of a daycare center at the workplace. Support the work of local organizations such as Sadaqa on achieving this goal. Currently, if a company has 19 women or more with children under the age of four, the company has to establish daycare center. Subsequent tranches of the First Equitable Growth and Job Creation Programmatic Development Policy Financing (DPF) could require an amendment to this law that is gender neutral, (ie. Require that 19 employees who have children, rather than women). • Moreover, additional information and role models regarding how women currently handle work and children care duties could also help. Child care is important for working women, but people do not appear to understand that Jordanian working women are able to overcome this hurdle nor how they manage to do so. Most women leave their children with close relatives, but most people think that working women are most likely to leave their children in daycare. Most respondents, men and women, believe that working women should return to work, and do return to work, after their children are 4.5 years old. These restrictive personal and social normative beliefs can potentially delay a woman’s re-entry into the labor force (if she was already working), and can do so to a point where she does not return to work at all. Recommendation: Provide community based peer to peer mentorship of working or prospective working mothers within a network of family and neighbors/community members to learn from each other how to best manage the double burden, share resources and information about quality day care services, best commutes, provide support and create newrole model especially for young women. • Family Friendly Policies: The most strongly reinforced social norm in the study was not opposition to women working outside the home, but that women should return home by 5 pm. Personal beliefs, social normative expectations, social empirical expectations, and intrahousehold expectations all mutually support the idea that working women should be home by 5 pm. The DPF introduces a Flexible Work System which is likely to support women’s economic empowerment. The Borrower’s Council of Ministers has approved a Bylaw on Flexible Work that benchmarks the minimum wage rate by unit of hours. The DPF is expecting an increase in growth of formal, private, part-time female workers from 31.6 percent (2014-2016) to 35 percent during (2018-2019). Recommendation: Subsequent tranches could further advance flexibility and part- time arrangements. It could introduce measures that facilitate and encourage employers - particularly the private sector – to create jobs that can be done from home, such as technology- based activities which could be developed and ideal for working women. Communication about the Flexible Work System would be important to ensure that workers and prospective workers are aware of these options and minimum wage requirements, and that employers include clear and gender-sensitive guidelines about flexible arrangements in their employment policies. • Information About Jobs Available: Information about women’s work appears to be inaccurate, leaving people with uncertainty about where women should look for work and how to go about the process. For instance, respondents vastly over-estimated the share of women and married women who are working. We interpret this to be an awareness on the part of respondents that “things are changingâ€? in Jordan. But this finding also shows Jordanian households do not have specific information regarding women’s labor force participation, and are instead relying on vague generalities when thinking and talking about women’s work options. Moreover, although 72% of women in our sample worked in the private sector, and less than 1/3 worked in traditional fields for women (teaching or nursing), respondents thought that only about 30% of women work in mixed gender environments. Recommendation: Develop plans to support both public and private sector entities to strengthen information availability and dissemination about share of women in the labor force and employed in the private sector, types of jobs available and composition of employees, etc. 5.2. Sector-specific Recommendations • Education Sector: 31% of men associate women education with improved family status and better parenting. Changing the narrative to associate education with better careers that would result in improving family status and better parenting could shift perceptions (see social media campaigns section). Education and sustainable job creation linkages have to be better articulated through World Bank operations. o Female teachers face wage discrimination, discouraging them to work. We also know that employers have the practice of paying teachers in cash, as many do not have bank accounts. The Ministry of Education, through the Education Reform Support Program for Results could provide incentives for female teachers to open bank accounts. o Efforts to change mindsets about gender roles and stereotypes that are currently being propagated through the curricula and teaching methods employed by the Ministry of Education could be supported. Invest in changing the teaching materials and in building capacity on gender sensitive approaches to teaching. Whenever possible, work through parent teacher associations (PTA)s, targeting fathers in particular, to support girls’ education and its benefits that go beyond being more “eligible for marriageâ€?. Recommendation: change the narrative to associate education with better careers that would ultimately result in improved family status and better parenting • Social Protection & Jobs: In terms of job opportunities, propose and integrate solutions and global lessons from the IFC into job creation that fosters an environment that can sustain women’s productive engagement in the long-term, (not just short-term jobs, such as public works). In terms of social protection, international best practices show that giving benefits to women will result in better outcomes – particularly for children and may contribute to women’s economic empowerment. Through the DPF, the GoJ has approved a Plan to improve and expand the coverage of the National Aid Fund (NAF) cash transfer program to at least 85,000 additional households in 2019 and 2020 (42,000, and 43,000 households in 2019, and 2020 respectively). The program gives preferences to female beneficiaries e.g. prioritizing women-headed households, families with children, and working poor (including working women) and can extend its coverage in subsequent tranches. Recommendation: For future tranches, learning from other DPF’s such as Brazil’s Development Policy targeting vulnerable groups Loan Bahia Strengthening Fiscal Management and Promoting Better and More Inclusive Service Delivery (DPL) may be considered. http://www.worldbank.org/en/results/2017/12/04/improving-access-public- services-vulnerable-groups-bahia-brazil. Lastly, the pipeline operation Safety Nets and Youth Employment Project may also be a venue to promote these approaches. • Transport Sector: Affordable and reliable transportation would contribute to women’s labor force participation however it is not enough. The majority (82%) of women are willing to take public transportation; however, women believe that the longest acceptable commute time is 32 minutes on average. Results from the survey also show that concerns around harassment can be prohibitive: a large share of respondents (44%) think that working women expose themselves to harassment either commuting to and from work or at work. Recommendation: Design a public transportation system that facilitates safe and reliable transportations needs for women and can provide shorter commutes. Concerns about harassment could be accompanied by efforts, such as that supported in Mexico through a World Bank financed pilot project, “Hazme el Paroâ€? (Have My Back), which empowered 46 male bus drivers to question ‘machista’ attitudes around harassment that had become normalized. The project gave drivers the tools to detect different types of harassment and respond unwaveringly fast according to an agreed protocol. • Financial Inclusion of Women: Provide global best practices to the GoJ on regulatory frameworks and procedures to facilitate access of women to bank accounts. Over 90% of women in our sample, and 46% of working women, did not possess their own bank account. The working women who do not possess a bank account are likely being paid in cash. Not having a bank account may limit the kinds of jobs they pursue and are able to obtain. At the same time, status concerns are an issue for Jordanian households. Almost 70% believe that women work because of financial need (not to make something out of their lives or some other reason), more than half estimate some or most of working women have husbands who cannot provide, and about a third estimate that households with working women have husbands who are not in charge. Several projects in the financial sector could support some of these initiatives. Possible entry points may include: (i) MSME Development Project for Inclusive Growth, (ii) Innovative Start Up Fund Project, (iii) Strengthening the Regulatory and Institutional Framework for MSME Development, and (iv) Promoting Financial Inclusion Policies in Jordan. • Infrastructure (Energy and Extractives, Water) and Urban/Municipal Planning: Draw from the results of the survey to generate a deeper understanding of how intrahousehold dynamics affects decision-making processes around women’s participation in various economic activities related to infrastructure. This will ultimately promote cross-sectoral linkages with infrastructure, calling attention for the need to identify gender entry points that incorporate a family centered approach when engaging women in activities related to hard sectors and project planning. • Cross Cutting Themes (Gender, Social Development, Communications): Leverage social development and gender entry points in all operations to integrate awareness raising campaigns that can accompany proposed structural (policy) and design (project level) issues, and that specifically target constraints highlighted in report and policy recommendations section. Working with the International Finance Corporation (IFC) & Private Sector Building on IFC’s ongoing efforts to strengthen opportunities for young women and men in private employment while leveraging its regional experience on financial inclusion can open doors for stronger private sector engagement in World Bank operational projects. • On employment: Linking to and learning from IFC’s education for employment program in Jordan could be an important entry point to widen the benefits of market-relevant programs especially to women. This can be done by exploring the potential of incorporating behavioral aspects to skills building or employability programs as well as keeping in mind normative constraints revealed through the study such as on travel time, mixed workplaces (or training places, expected time to be home, etc. What can be implemented to avoid additional burdens on companies? How can we ensure that the companies don’t end up refusing to hire women because they have to be flexible? • A second entry point is through a focus on financial inclusion: According to results of the survey, 76 percent of female respondents did not have bank accounts with a large gap between working women (44 percent) and non-working women (90 percent). Lack of access to and control of assets can be a major barrier for women’s economic participation, especially in countries where women are not granted equal inheritance and land rights. Such inequities also reduce women’s bargaining power in the home, which ultimately impacts her voice and agency with regard to work. Working with IFC on exploring the potential of adapting its financial inclusion programs, such as the Banking on Women program in Lebanon and the Palestinian territories, can provide greater opportunities for women especially entrepreneurs to access financial services. It can also enable banks to rethink their client market. For example, the Bank of Palestine (BoP) in partnership with the IFC has made a concerted effort to changing its banking culture across the Palestinian territories by training its BoP staff on gender-responsive customer service to ensure financial information and resources are shared equally with both female and male clients. The BoP has also been working on bank-wide policy changes that would overcome gender-based legal restrictions to financial access, such as allowing women to open an account for their children without a male guardian consent.59 The potential of such a program could be explored in Jordan, within the context of strengthening financial access to women and strengthening their household bargaining power. Strengthening access to women and girls, and incorporating financial literacy in early education may also help increase women’s agency in financial inclusion. Tax incentives to companies that provide flexible working hours for women could also be explored. Media campaign interventions The GoJ could implement a multi-pronged effort that invests in shifting mindsets of men, women, and boys from an early age. For example, actual male beliefs are more liberal than men imagine them to be. This opens the door to information campaigns and narratives that change their understanding of what others believe. For instance, if mixed gender work environments were normalized through discussion, discourse, and role models (e.g., educational attainment could serve this role), men would learn that other men know about these environments, and become less concerned about them. For example, families of working women are not perceived as less traditional nor are they accused of the husband’s inability to provide for the familyâ€?. Only 19% of respondents associate married women working with their husband’s inability to provide for her (slightly higher for non-working women and their male counterparts); and less than 10% believe that the husband is not in charge in families with working women (slightly higher for women compared to men). However, the false narrative of the statements above make Jordanians believe that most people hold such views. There is evidence that media campaigns, entertainment education, and aspirational interventions can, under the right circumstances, change behavior (World Bank 2014). Following are some media campaigns that could be implemented to encourage higher female labor force participation in Jordan: • Aspirational videos: Aspirations are shaped by exposure to and observations of outcomes of individuals whom one can observe and identify with. Aspirational videos, featuring working women from relatable backgrounds who have managed to find a way around structural and intra- household barriers, can be shown to raise Jordanian women’s aspirations and broaden the scope of livelihood options they consider viable for themselves. Aspirational videos shown in rural Ethiopia, featuring Ethiopians from similar backgrounds who were successful in agriculture and 59 Hillis, Samira Ahmed; Constant, Samantha M. 2018. Second Country Gender Action Plan (C-GAP II) for Palestinian Territories (FY2018-2021) (English). Washington, D.C.: World Bank Group. business on their own, were more effective at raising villagers’ aspirations when people in the community were exposed to the content (Bernard and others 2014). • Media campaigns targeting men: Given the critical role men play in women’s labor force participation decisions, media interventions need to focus on and target men as well. More than 90% of women respondents said that, in the decision to work, the views of male household members (mostly husbands) are important or extremely important. Only 25% agreed that a woman has the right to work if it makes the men in the household uncomfortable. For example, a Promundo type campaign can be developed to engage men as active partners in women’s empowerment and transform harmful gender attitudes and norms and impede women’s labor force participation. Advocacy campaigns, community outreach, and workshops targeting men and boys in India, Brazil and Chile lead to statistically significant changes in attitudes towards gender- based violence, and decrease in intimate partner violence (Instituto Promundo, 2012). • Soap operas: Story-telling and media exposure can have a powerful influence on real life behavior. Soap operas can be used to deliver entertainment education across a society to change prevailing beliefs and mental models around the role of women in society and within the household. For example, a soap opera could be produced in Jordan with a storyline featuring a working woman, her family, and their struggles and experiences, to normalize the concept of families with working women and potentially even provide a role model for women. Soap operas could address the disconnect between what Jordanian men observe and what they assume other men think and believe regarding women working in mixed gender environments (40% believe it is sometimes okay, men estimate that half of all other husbands would not be okay if their wives worked in mixed gender settings). In Brazil, access to the TV Globo network—which was dominated by soap operas with independent female characters with few, or even no children—has been linked to the country’s rapid drop in fertility. Viewing the soap operas had an effect equal to 1.6 years of additional education (La Ferrara, Chong, and Duryea 2012). Approaches like these could be used to stimulate women’s labor market participation in Jordan. Soap operas, aspirational videos, social media campaigns, or a combination, could be used to broadcast information about how and where to find work, correct misperceptions regarding aspects of Jordanian society that estimated to be more restrictive than they really are, empower women by lifting aspirations, and provide new mental models and role models for family life. 5.3. Recommendations for other MNA Countries The research study in Jordan both confirmed and demystified some of the constraints believed were preventing women’s labor force participation. It demonstrated that, in the case of Jordan, personal beliefs and intra-household bargaining are more influential than societal norms and expectations, though expectations of what others do can also be influential. While we cannot ascertain that is the case in other MNA countries, some lessons from this study that could be taken into account for others are: • It is important that Project Teams that aim to promote favorable labor market policies are cognizant that proposing changes to structural issues alone may not be sufficient to achieve the expected outcomes. It is paramount that they be coupled with “softâ€? interventions such as those that can influence intra-household decision-making. • Linked to this, employability and skills building (i.e. financial inclusion, vocational, etc) programs that target women and youth, especially young women, should also consider a behavioral component to help women navigate intrahousehold dynamics that may disadvantage them from fully benefiting from such programs. This component could also be an opportunity to engage men and boys in order to ensure male members of a household are not alienated during the process of women’s economic empowerment. • Invest in interventions that support the changing of attitudes towards shared household responsibilities, (chores, etc.) emphasizing messages that it is ‘manly’ to help around the house. • Support activities – wherever possible – to enable women to open their own bank accounts and gain financial literacy. Women’s economic empowerment is core to the WBG’s and MNA specific gender strategies, and this should be promoted whenever possible (via MFM, PSD but not limited to these sectors). • Invest in interventions that encourage women to aspire for a career path beyond obtaining a college degree. Focus on messages highlighting women with career paths in various sectors and the impact of having careers and two incomes on the living standards of their homes and the aspirations of their children. • When promoting women’s labor force participation, consider integrating global lessons to promote longer-term employment and job growth for women. What processes, procedures, laws, and/or incentives can be included to encourage women to stay in the job, so that they can envision professional growth opportunities? • When promoting female job creation is one of the project’s objectives, ensure that flexible jobs (flexibility in hours and location) is part of the design. For example, a positive – unintended – externality of the Egypt Inclusive Housing Program for Results, was that the Program produced home-based jobs for women through online loan application processes and responding to grievances and queries. Bibliography Bernard, Tanguy, Stefan Dercon, Kate Orkin, and Alemayehu Seyoum Taffesse. 2014. “The Future in Mind: Aspirations and Forward-Looking Behaviour in Rural Ethiopia.â€? Centre for the Study of African Economies, University of Oxford, Oxford. Bicchieri, Christina. (2006). The grammar of society: The nature and dynamics of social norms. Cambridge University Press. Bicchieri, Christina, & Mercier, Hugo. (2014). Norms and beliefs: how change occurs. In The Complexity of Social Norms (pp. 37-54). Springer International Publishing. Bicchieri, Christina, Lindemans, Jan W., & Jiang, Ting. (2014). A structured approach to a diagnostic of collective practices. Frontiers in psychology, 5. Care Jordan. (2013). Syrian Refugees in Urban Jordan: Baseline Assessment of Community-Identified Vulnerabilities Among Syrian Refugees living in Irbid, Madaba, Mufraq, and Zarqa. Rapid Participatory Community Assessment. Amman, Jordan. Crossley, Michelle. (2002). Introducing Narrative Psychology. In C. Horrocks; N.Kelly; B. Roberts; D. Robinson (Eds.) (2002): Narrative, Memory and Life Transitions. University of Huddersfield, pp. 1-13. Dababneh, Abeer and Salah Taher. (2016). Women’s Economic Participation in Jordan: Realities and Challenges in the Private Sector. Amman, Jordan. Eriksson, Lina. (2015). Social Norms Theory and Development Economics. Policy Research Working Paper No. 7450. World Bank, Washington, DC. https://openknowledge.worldbank.org/handle/10986/22863 License: CC BY 3.0 IGO.â€? Hashemite Kingdom of Jordan Official Notice of Legal Amendment. (2014). Law No. 34 (Year 2014): Income Tax Law. Hillis, Samira Ahmed; Constant, Samantha M. 2018. Second Country Gender Action Plan (C-GAP II) for Palestinian Territories (FY2018-2021) (English). Washington, D.C.: World Bank Group. Instituto Promundo. (2012). Engaging men to prevent gender-based violence: A multi-country intervention and impact evaluation study. Report for the UN Trust Fund. Washington, DC: Promundo. International Finance Corporation (IFC). (2015). Gender Diversity in Jordan: Research on Impact of Gender Diversity on the Economic Performance of Companies in Jordan. In collaboration with the Canadian Department of Foreign Affairs, Trade and Development and Switzerland State Secretariat for Economic Affairs and UKAid. Washington DC. International Labour Organization. (2012). Response to Classic Fashion Apparel Industries Allegations of Sexual Assault and Better Work - Jordan Follow-up on the Issue of Sexual Harassment. Working Paper. Geneva. International Labour Organization. (May 22, 2013). Gender Pay Discrimination in Jordan: A Call for Change, Feature Story. http://www.ilo.org/global/about-the- ilo/newsroom/features/WCMS_213754/lang--en/index.htm La Ferrara, Eliana, Alberto Chong, and Suzanne Duryea. 2012. “Soap Operas and Fertility: Evidence from Brazil.â€? American Economic Journal: Applied Economics 4 (4): 1–31. Lundvall, Jonna Maria; De Paz Nieves, Carmen; Constant, Samantha M.; Atamanov, Aziz. 2017. Progress towards gender equality in the Middle East and North Africa region. World Bank Washington, D.C. Ministry of Labor. Jordanian Labor Law no. (8) of 1996. Hashemite Kingdom of Jordan. Ministry of Planning and International Cooperation. Jordan’s National Employment Strategy (2011- 2020). Hashemite Kingdom of Jordan. Namri, Nadine. (March 7, 2016). How to open doors for women in Jordan. Washington DC: World Bank. Oxford Analytica. (2018). More Jobs for Middle East women could boost growth. United Kingdom. Posner, Richard A. (1997). Social norms and the law: An economic approach. The American economic review, 365-369. Promundo-UN Women. (2017). Understanding Masculinities: Results from the International Men and Gender Equality Survey (IMAGES), Egypt, Lebanon, Morocco, and Palestine. Razzaz, Susan. (2017). A Challenging Market Becomes More Challenging: Jordanian Workers, Migrant Workers, and Refugees in the Jordanian Labour Market. International Labour Organization (ILO). Geneva. Sarbin, Theordore R. (1986). The narrative as root metaphor for psychology. In T. R. Sarbin (Ed.), Narrative psychology: The storied nature of human conduct (pp. 3-21). New York: Praeger. (19 sider). UNDP. (2012). Gender Equality and Women’s Empowerment in Public Administration. Jordan Case Study. New York. World Bank. (2013). Economic Participation, Agency and Access to Justice. Hashemite Kingdom of Jordan: Jordan Country Gender Assessment. Washington DC. ----. (2014). Social Insurance Reform in Jordan: Awareness and Perceptions of Employment Opportunities for Women in Jordan. Social Protection & Labor Discussion Paper. No. 1402. ---- (2016). Paving the way for women’s economic inclusion in the Gulf Cooperation Council (GCC). World Bank, Washington D.C. ---- (2018). Women, Business and the Law. World Bank, Washington D.C. --- (2014). Mind, Society, and Behavior: The World Development Report 2015. World Economic Forum. 2017. The Global Gender Gap Report: Insight Report. Geneva. Annex 1: The Global Gender Gap Report 2017 Table 1: Structure of the Global Gender Gap Index 1. Sub Index: 2. Variable 3. Source 4. Economic 5. Ratio: female labour force 6. 7. International Labour Organization, ILOSTAT Participation and participation over male value database, 2016 or latest available data Opportunity 8. Wage equality between women and 9. World Economic Forum, Executive Opinion Survey men for similar work (survey data, (EOS), 2016-17 normalized on a 0-to-1 scale) 10. Ratio: female estimated earned 11. World Economic Forum calculations based on the income over male value United Nations Development Programme methodology (refer to Human Development Report 2007/2008) 12. Ratio: female legislators, senior 13. International Labour Organization, ILOSTAT officials and managers over male value database, 2016 or latest available data 14. Ratio: female professional and 15. International Labour Organization, ILOSTAT technical workers over male value database, 2016 or latest available data 16. Educational 17. Ratio: female literacy rate over male18. United Nations Educational, Scientific and Cultural Attainment value Organization (UNESCO) Institute for Statistics, Education indicators, database, 2016 or latest available data 19. Ratio: female net primary enrolment 20. UNESCO Institute for Statistics, Education rate over male value indicators database, 2016 or latest available data 21. Ratio: female net 22. secondary UNESCO Institute for Statistics, Education enrolment rate over male value indicators database, 2016 or latest available data 23. Ratio: female gross tertiary enrolment24. UNESCO Institute for Statistics, Education ratio over male value indicators database, 2016 or latest available data 25. Health 26. Sex ratio at birth (converted to female- and 27. United Nations Population Division, World Survival over-male ratio) Population Prospects, 2016 or latest available data 28. Ratio: female healthy life expectancy 29. World Health Organization, Global Health over male value Observatory database, 2015 or latest available data 30. Political 32. 31. Ratio: females with seats in parliament Inter-Parliamentary Union, Women in Politics: Empowerment over male value 2017, reflecting elections/appointments up to 1 June 2017 34. 33. Ratio: females at ministerial level over Inter-Parliamentary Union, Women in Politics: male value 2017, reflecting appointments up to 1 January 2017 35. Ratio: number of years with a female 36. World Economic Forum calculations, reflecting head of state (last 50 years) over male situation as of 30 June 2017 value