Female BREAKING Entrepreneurs Who Cross Over to Male-Dominated Sectors BARRIERS © 2021 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. 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. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Cover design: Vito Raimondi Female BREAKING Entrepreneurs Who Cross Over to Male-Dominated Sectors BARRIERS Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors 2 Table of Contents Foreword 4 What Do Our Data Tell Us about the 40 Drivers of Sectoral Segregation in Acknowledgments 5 these Countries? Executive Summary 6 Section 3: Unpacking the Profitarchy 44 What’s New in This Report? 8 The Correlates of Crossing Over to 48 MDS Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors The Main Findings 9 Spouses, Domestic Work, and 48 Partnerships Introduction: Sectoral Segregation 14 and the Gender Profit Gap Exposure to Male-Dominated 51 Sectors The Differences in Profits between 15 Male and Female Entrepreneurs and Capital, Wealth, and Number of 53 the Role of Gender-Based Sectoral Workers Segregation Skills and Abilities 55 Why Do Female Entrepreneurs 18 Concentrate in Less Profitable Sectors? Context-Specific Factors 60 Understanding the Common Barriers 21 Which Came First, the Chicken or the 60 to Crossing Over to Male-Dominated Egg? Sectors across Ten Countries in Three Regions Section 4: Conclusions and Policy 62 Recommendations Section 1: Gender-Based Sectoral 22 Segregation of Firms Support, Networks, and Exposure to 65 the Field How Do We Determine if a Sector of 23 a Firm is Male Dominated or Female Enhancing Skills and Training 69 Concentrated? Capital/Assets and Access to Loans 71 What Are Male-Dominated Sectors and 24 Female-Concentrated Sectors? Strengths and Limitations of This 73 Report Building on Previous Research 27 Agenda for Future Research and Policy 73 Section 2: The Profitarchy 32 Appendixes 76 What Is the Profitarchy? 33 Appendix A 77 Horizontal Segregation: The Role of 37 Appendix B 81 Sectoral Segregation in Creating the Profitarchy Appendix C 83 Vertical Segregation: Female-Owned 38 Appendix D 87 Firms within the Same Sector Continue to Lag behind Male Firms References 105 3 Foreword Female entrepreneurship is on the rise globally concentrated sectors. Even though the sectoral and in some countries, women are as or more composition might change from one country likely than men to own a business, yet they to another, helping women cross over to more operate smaller businesses and concentrate profitable male-dominated sectors could in less profitable sectors than men. Social contribute to their business performance norms, unconscious biases, lack of exposure to more generally, and may also make them as male-dominated sectors, and time and capital profitable as male entrepreneurs. And this will constraints are just some of the factors holding contribute to economic growth as skills are Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors women back from entering more profitable more efficiently distributed. male-dominated sectors. Policy makers could explore a number This new report, Breaking Barriers: Female of interventions that appear promising Entrepreneurs Who Cross Over to Male- to support women to cross over. These Dominated Sectors, produced by the Africa include encouraging spousal support; safely Gender Innovation Lab, the East Asia and connecting women to mentors and role Pacific Gender Innovation Lab, and the Latin models; providing early exposure to and America and the Caribbean Gender Innovation training in male-dominated sectors; enhancing Lab under the guidance of the World Bank’s women’s education; and increasing access to Gender Group offers insight into who are capital and loans. Complementary measures, the women who break into male-dominated such as tackling discrimination and harassment, sectors. Based on this analysis, the report which work against female crossovers, highlights policy and program options for could help women establish and grow their policy makers and other key stakeholders businesses once they have crossed over. including development partners, civil society, and corporations. The broader economy loses when individuals consider only a limited set of occupations The report spans three regions and ten based on their gender. As businesses start countries, while also drawing upon a to recover from the COVID-19 pandemic, global survey of entrepreneurs through a we hope that the report’s findings will guide social media platform. This multi country key decision-makers to implement programs exploration unearths common themes across and policies that enable women and girls to the globe while also demonstrating country- enter more profitable sectors, diversify their specific findings on gendered occupational economic opportunities, and contribute to a segregation. While programs and policies to faster, more equitable recovery. support female entrepreneurs who would like to enter male-dominated sectors will need to consider the country context to appropriately target specific sectors, several key messages emerge across countries in the report’s analysis. Mamta Murthi In almost all countries in the sample, women World Bank’s Vice President for Human who operate businesses in male-dominated Development sectors outperform women in female- 4 Acknowledgments Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors The Breaking Barriers: Female Entrepreneurs The authors would like to gratefully Who Cross Over to Male-Dominated Sectors acknowledge Elena Ianchovichina and Maurizio report was prepared by a team led by Naira Bussolo, who peer reviewed the report. Kalra and Markus Goldstein, and included Maria Emilia Cucagna, Fannie Delavelle, We are grateful to the United States Agency Leonardo Iacovone, Hillary C. Johnson, Paula for International Development (USAID) and Lorena Gonzalez Martinez, Eliana Carolina the World Bank Umbrella Facility for Gender Rubiano Matulevich, Elizaveta Perova, Rachael Equality (UFGE) for funding the research. The Pierotti, Gareth Roberts, and José Daniel UFGE is a multidonor trust fund administered Trujillo. Some sections are based on work by by the World Bank to advance gender Salman Alibhai, Ana Maria Munoz Boudet, equality and women’s empowerment through Niklas Buehren, Francisco Campos, Ludovica experimentation and knowledge creation to Cherchi, Daniel Kirkwood, Laura McGorman, help governments and the private sector focus Sreelakshmi Papineni, and Obert Pimhidzai. policy and programs on scalable solutions with sustainable outcomes. The UFGE is supported This report is a collaboration of the World with generous contributions from Australia, Bank’s Africa Gender Innovation Lab, the East Canada, Denmark, Germany, Iceland, Latvia, Asia and Pacific Gender Innovation Lab, and the Netherlands, Norway, Spain, Sweden, the Latin America and the Caribbean Gender Switzerland, the United Kingdom, the United Innovation Lab under the guidance of the States, and the Bill and Melinda Gates World Bank’s Gender Group. Foundation. The team thanks Cansu Birce Gokalp for leading the editing and production process, Tigist Assefa Ketema for providing research assistance and Amy Elizabeth Copley for her support with editing. The report was designed by Vito Raimondi and copyediting was carried out by Zuzana Johansen. 5 BREAKING BARRIERS Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Executive Summary 6 Executive Summary P icture the owner of a construction firm, a computer programming company, or an automotive repair store.1 Is it a man or a woman? If you pictured a man, you are not alone, and you are not far from reality. A recent study that uses global survey data from businesses with a Facebook Business Page showed that more profitable sectors, such as the three mentioned above, tend to be dominated by male-owned firms while female entrepreneurs tend to concentrate in sectors which are relatively less profitable (Goldstein, Gonzalez, and Papineni 2019). In addition to the glass Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors ceiling women face in rising to leadership positions, female entrepreneurs are also potentially surrounded by glass walls, making it challenging for them to enter more profitable, traditionally male-dominated sectors (MDS). Unconscious biases, societal norms, lack of exposure to the sector, and time and capital constraints are just some of the many push and pull factors holding women back from entering male-dominated sectors. Yet very little research has been undertaken to identify the factors that support women to enter such sectors. This report aims to fill this and other gaps in the literature on gender- based sectoral segregation. In the first section, the report provides an overview of the sectors that are typically dominated by male entrepreneurs, and the sectors that have a concentration of female microentrepreneurs across the ten countries studied. Based on this understanding of gender-based segregation of sectors, the second section then presents an overview of the hierarchy of profits, referred to in this report as “the profitarchy”2 (Goldstein, Gonzalez, and Papineni 2019), for male and female entrepreneurs who are operating microenterprises across male-dominated and female-concentrated sectors. Within this profitarchy, the specific focus is on exploring the difference in profits among female entrepreneurs who cross over into male-dominated sectors compared to those who remain in traditionally female-concentrated sectors (FCS). The third section of the report offers a snapshot of the factors associated with being a female entrepreneur who crosses over to MDS and identifies the most salient cross-country factors that are associated with breaking into and surviving in these more profitable sectors. In its final section, the report outlines priority action areas and aims to provide policy makers and other key stakeholders including development partners, corporations, and civil society with concrete solutions to galvanize action around this agenda while highlighting the remaining knowledge gaps that require more experimentation and research. 1 This exercise was inspired by a study by Tina Kiefer (2018), a professor of organizational behavior at the University of Warwick, where she asks the question, “Picture a leader. Is it a man or a woman?” 2 The profitarchy term was coined by the authors to describe the hierarchy of profits for male versus female entrepreneurs across sectors found in the Future of Business Survey. 7 Executive Summary What’s New in This Report? → A focus on microentrepreneurs: Data → A multicountry exploration: This report in this report, with the exception of aims to conduct an in-depth analysis of the Future of Business (FoB) study, the characteristics of “crossovers,“ women focus on microentrepreneurs. Globally, who cross over into male-dominated microenterprises account for an average industries, in multiple countries to identify of 39 percent of all firms (Aterido, which policies and interventions could Hallward-Driemeier, and Pages 2009), and support more women to enter higher-paid, in low-income countries, the “informal” male-dominated sectors. The studies Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors microenterprise sector accounts for were conducted in Sub-Saharan Africa: an average of 47.2 percent of gross Botswana (Cherchi and Kirkwood 2019), domestic product (GDP), with women Ethiopia (Alibhai et al. 2017), Guinea, and disproportionately overrepresented in this Uganda (Campos et al. 2015, 2017); in category of entrepreneurs (ICRW 2019; Latin America: Mexico and Peru; and in Mayoux 1995). Southeast Asia: Cambodia, Indonesia, Lao People’s Democratic Republic (Lao PDR), → Synthesis of rich data: The report builds and Vietnam. The report also draws from the on previous research using data from 10 previously mentioned global multicountry low- and middle-income countries across FoB survey of entrepreneurs carried out three regions as well as a global survey through Facebook (Goldstein, Gonzalez, and of entrepreneurs in 97 countries through Papineni 2019). the Future of Business (FoB) data set to provide (1) a multicountry overview of male- → Policy focus: The report presents an dominated sectors and female-concentrated overview of the types of policies and sectors and (2) patterns in the hierarchy programs that can support crossing over of profits, with a focus on comparing to the typically more profitable male- average profits of enterprises run by female dominated sectors. entrepreneurs in FCS vs. those in MDS. Photo by Christina Morillo from Pexels 8 Executive Summary The Main Findings Female-owned firms tend to concentrate in trade and retail industries, especially in textile and footwear, and pharmaceutical and perfume products, whereas male- Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors owned enterprises dominate most manufacturing sectors as well as agriculture, forestry, and fishing. The services sectors are neither male- dominated nor female-concentrated sectors. There appears to be a FEMALE-CONCENTRATED SECTORS concentration of female entrepreneurs MALE-DOMINATED SECTORS in some services such as hairdressing, personal services, hospitality, and food in all countries where these were assessed, while other services like automobile maintenance and sales, and small transport services are dominated by men. However, despite some common trends across countries we discuss in the report, there is no universal definition of which sectors are male dominated and which ones are not. As such, policies to support female entrepreneurs who would like to enter male-dominated sectors will need to consider the specific country context to appropriately target specific sectors. PRODUCTION SERVICES 9 Executive Summary Female entrepreneurs who cross over to male-dominated sectors perform better than female entrepreneurs in female-concentrated sectors in all countries studied except for Cambodia, where operating in FCS appears to be more profitable. In Botswana, Guinea, Lao PDR, and Uganda, crossing over to male-dominated sectors also ensures that female entrepreneurs are on average as profitable as male entrepreneurs operating in MDS. FEMALES IN MDS FEMALES IN FCS Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors % GAP IN AVERAGE PROFIT GUINEA 90 BOTSWANA 119 UGANDA 140 ETHIOPIA 80 MEXICO 51 PERU 71 LAO PDR* 80 VIETNAM 32 INDONESIA 67 -19 CAMBODIA NOTE: Orange bars compare the average profits of female-owned firms in male-dominated sectors (MDS) to those in female- concentrated sectors (FCS). Further information can be found in table 3 on page 35. * Sales were used in Lao PDR instead of profits due to data constraints. In some countries, firms owned by men continue to do better than those owned by female entrepreneurs irrespective of the sectors they operate in, suggesting crossing over as only one promising tool in a broader set of policies to support female entrepreneurs. This is observed in Indonesia, Mexico, and Vietnam. This might be a result of women clustering in less profitable roles and activities within the same sector, discrimination faced by women in male- dominated sectors, or other gendered barriers that limit female profits. 10 Executive Summary What factors are Characteristics such as women’s education, past associated with exposure to male-dominated sectors through work crossing over? experience or training, exposure to MDS through male relatives, mentors, or role models, and spousal support appear to be positively associated with crossing over in almost all countries where these were assessed. Other factors that are explored in this report include sociodemographic and family- related characteristics such as age, marital status, Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors women’s education household size, number of children, household assets, parental education, and parent’s occupation; skills and training-related factors including education levels, self-efficacy and locus of control, cognitive abilities, and exposure to a male-dominated sector through previous work experience; factors associated with social capital, networks, and family support such as receiving support from a spouse or partner, having work experience a role model or mentor, and inheriting the business. or training Last but not least, we explore the characteristics of the business itself, such as whether it is run with a partner and jointly owned or owned solely by the entrepreneur, age of the business, number of workers it has, and whether the entrepreneur also works as an employee of the business. Even though only a handful of characteristics are consistently assessed spousal support across all studies, we find that each one of the abovementioned factors is associated with crossing over in at least one country if not more. Furthermore, spousal support appears to be key in successfully crossing over and operating in male-dominated sectors among married women. socio emotional skills male relatives, mentors, or role models 11 Executive Summary Evidence-based programs and policies could support women to cross over and contribute to their business performance more generally. At its heart, occupational segregation is a constraint on growth. Limiting workers to certain sectors based simply on their sex prevents the economy from making the best use of the skills available. Public policy has a role to play in breaking down these barriers. The goal of these recommendations is not to make women more like men, but rather to level the playing field, opening up all sectors of the economy to women. Overall, we recommend policies and programs that encourage spousal support, safely connect women to mentors and role models, and provide early exposure to and training in MDS. Increasing spousal support, task sharing, and joint decision-making through couples’ training and Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors mass media programs may help women cross over and continue to operate successfully in MDS. Initiatives that safely link female entrepreneurs to mentors in male-dominated sectors and expose them to MDS in general, such as internship and mentorship programs, can also be a promising approach to open these sectors to women. Furthermore, providing trainings that improve women’s socioemotional and cognitive skills, as well as their technical skills and access to information on MDS at early stages are other ways of encouraging women to cross over. In addition to addressing skills, it is also important to lift constraints around access to loans for businesses and access to networks within sectors, as well as to improve the overall conditions of working in MDS for women. Safety in the workplace should be guaranteed through measures including safe transport, reporting and redressal mechanisms, and gender sensitization for men in these sectors. Since the impact of the COVID-19 pandemic has been disproportionately faced by female entrepreneurs, particularly those employed in the services sector, we hope that this report will shape policies and programs that will enable female entrepreneurs to diversify to other sectors of employment, strengthening their income potential and ensuring the stability of their income in crisis situations. 12 © Simone D. McCourtie / World Bank 13 Executive Summary Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors INTRODUCTION Sectoral Segregation and the Gender Profit Gap 14 Introduction The Differences in Profits between Male and Female Entrepreneurs and the Role of Gender-Based Sectoral Segregation W omen make up a greater share of entrepreneurs worldwide than ever before. Globally the gaps in the ratio of female to male participation in entrepreneurial activity are shrinking, and in some countries, women are as or Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors more likely than men to own a business (Buvinic, Knowles, and Witoelar 2018; Kelley et al. 2017; Kelley, Singer, and Herrington 2016; Meunier, Krylova, and Ramalho 2017).3 Women’s greater participation, however, does not mean that they are benefiting equally. Profits and sales of female entrepreneurs continue to be lower on average than those of male entrepreneurs and their businesses have higher closure rates and less potential for growth (Bardasi, Sabarwal, and Terrell 2011; Buvinic, Knowles, and Witoelar 2018; Carranza, Dhakal, and Love 2018; McKenzie and Paffhausen 2017; Rijkers and Costa 2012; World Bank 2019). The gender gap in profits may also be partly explained by the fact that women tend to run smaller firms than men (Hallward-Driemeier 2013; Carranza, Dhakal, and Love 2018; Buvinic, Knowles, and Witoelar 2018). In addition to firm characteristics, the concentration of male and female entrepreneurs in different sectors of the economy may also explain this gap. Globally, women who enter male-dominated sectors earn 67 percent higher profits on average than women who remain in traditionally female-concentrated sectors (Goldstein, Gonzalez, and Papineni 2019). When comparing profits between male and female entrepreneurs across 26 countries in Eastern Europe and Central Asia (ECA), Sabarwal and Terrell (2008) found that women run smaller firms in terms of the number of employees and there is a gap in revenue between the average female and male entrepreneur in a given country. However, this gap significantly shrinks from 63.1% to 37.2% when the industry they operate in is accounted for. Even when we look among microentrepreneurs,4 where the firm’s size does not vary substantially, this gender difference in profits persists in many countries. Figure 15 shows the gender differences in profits among microentrepreneurs, where the size of the firm ranges from 0 to 5 or 6 employees. Microenterprises owned by men earn 12% more than those owned by women in Benin, 34% more in Botswana, and 80% more in Mexico. 3 This is true for Sub-Saharan Africa as a whole, which is the only region where women make up the majority of those who are entrepreneurs (World Bank 2019). 4 The studies varied in their definition of microentrepreneurs. Some defined microentrepreneurs as those who own small or ‘micro’ enterprises/businesses with up to 5 employees while others defined them as having up to 6 employees. 5 While these are not all the results of representative surveys, they do provide a useful picture of gender gaps in profit. 15 Introduction Figure 1 Differences in past month or past year profit between female-owned and male-owned microenterprises GENDER GAP IN MONTHLY PROFITS BETWEEN MALE AND FEMALE MICROENTREPRENEURS FEMALE MALE Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors 12% Benina Congo, Dem. 49% Rep.a (census) Ethiopiaa 45% (manufacturing) 36% Ghanaa 31% Malawia -7% Togoa 30.5% Ugandaa 41% Guineaa 34% Botswanab 80% Mexicob 53.5%* Indonesiab 1.6%* Cambodiab 48.5%* Vietnamb Sources: a Profiting from Parity report (World Bank 2019). b Data from the individual Worls Bank studies included in this report. * =Yearly profits. 16 Introduction Overall, economic growth is hampered when individuals consider only a limited set of occupations to pursue work opportunities or open businesses. When this results from restrictive gender norms or gender differences in access to resources, growth potential is stymied for both female-owned enterprises themselves and the broader economy. Macroeconomic research on the consequences of occupational segregation of the workforce indicates that over Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors the past 50 years, the increased participation of women and other marginalized groups in highly skilled professions in the United States has led to a 20–40 percent growth in gross domestic product (GDP) per capita (Hsieh and Klenow 2016; Hsieh et al. 2019). Complete removal of gendered occupational segregation could result in approximately another 10 percent increase in GDP today (Hsieh et al. 2019). Studies that focus specifically on gendered sectoral segregation among just entrepreneurs, instead of the occupational segregation of the entire workforce, continue to highlight that segregation of male and female entrepreneurs into different sectors results in a misallocation of talent that can contribute to the gender gaps in profit. Firm profits are significantly lower in the sectors where women are overrepresented compared to sectors where they are underrepresented (Hallward-Driemeier 2013; Rijkers and Costa 2012; Bardasi, Sabarwal, and Terrell 2011; Hardy and Kagy 2020; Singh, Reynolds, and Muhammad 2001; World Bank 2019). A study from Michigan, USA, estimates that choosing to operate in the personal services sector, a typically female-concentrated sector, over construction and professional services, which are typically male dominated, explains as much as 9 percent to 14 percent of the earnings gap between male and female entrepreneurs (Hundley 2001). Globally, women who enter male-dominated sectors earn 67 percent higher profits on average than women who remain in traditionally female-concentrated sectors (Goldstein, Gonzalez, and Papineni 2019). Despite this, women tend to start businesses in fewer and less profitable sectors than men (Kalleberg and Leicht 1991; Fairlie and Robb 2009; Bardasi, Sabarwal, and Terrell 2011). 17 Introduction Why Do Female Entrepreneurs Concentrate in Less Profitable Sectors? Female entrepreneurs tend to cluster in low-profit-yielding sectors, with lower returns, more informality, and lower potential for growth, compared to men who dominate a wider selection of industries (Carranza, Dhakal, and Love 2018; Hallward-Driemeier 2013; Bardasi, Sabarwal, and Terrell 2011). This is potentially driven by factors such as social norms about women’s interactions with men, the social roles attributed to them, their smaller, more family-related Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors networks, financial discrimination, and lack of gender-equal property and labor laws, as well as the risk of facing gender-based violence, to name a few (World Bank 2019; Carranza, Dhakal, and Love 2018; Jayachandran 2020). These characteristics can impact both sector choice and profits. For example, social norms around what is acceptable for a woman to do in a particular culture may lead to women concentrating in a handful of sectors that then tend to be overcrowded, thereby making them less profitable (Hardy and Kagy 2020). We highlight some of the many ways in which sector choice and resulting profit gaps are a cause and consequence of the systematic gender differences in societal expectations from female entrepreneurs. segregation literature also suggests that mothers tend to concentrate in stereotypically female occupations compared to nonmothers (Okamoto and England 1999). Flexibility However, the choice of location can impact One characteristic that may drive women’s the visibility and legitimacy of the businesses concentration in some sectors is the flexibility in the customer’s eyes and thereby reduce of timing and location offered by the sector. sales (Carranza, Dhakal, and Love 2018). Women may choose to operate in sectors that Furthermore, women who run businesses out of require less mobility and allow for home-based their homes are likely to make working capital– work due to safety concerns, care work, and related investments beneficial to both the household responsibilities (de Mel, McKenzie, business and the household. This contributes and Woodruff 2009). Women typically shoulder to lower returns on investments by women a greater burden of care for small children compared to men, who are less likely to work in and elderly relatives and are likely to locate sectors that allow flexibility of location and are their businesses in their homes to manage also less likely to factor in household needs in their household responsibilities (Carranza, their business investments (de Mel, McKenzie, Dhakal, and Love 2018). The occupational and Woodruff 2009). 18 Introduction Barriers to Entry Access to Capital Linked closely to the flexibility offered by a Male-dominated sectors are usually more sector are the barriers to entry and business capital intensive (World Bank 2019), which formalization. Female-concentrated sectors may be a barrier for women’s entry given the typically have a lower share of formalization constraints women face in accessing capital. In and fewer barriers to entry, such as a lower Sub-Saharan Africa, for example, male-owned Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors need for previous business experience, human firms have six times the capital investment of capital, or social networks (Hallward-Driemeier female-owned firms (World Bank 2019). Due 2013; Carranza, Dhakal, and Love 2018). The to these constraints, female entrepreneurs concentration of female entrepreneurs in may select sectors that typically have smaller sectors with lower barriers to entry is also not and less efficient businesses and require fewer surprising given that worldwide, men and capital inputs (Bardasi, Sabarwal, and Terrell women have access to different opportunities, 2011). Despite the fact that limited access to and there are essential gender gaps in capital is one of the primary drivers of sectoral educational attainment. This human capital segregation, we sometimes see that when the constraint inherently makes these sectors less sector is accounted for, the difference in profits profitable. or firm size due to access to capital is not always as stark. For example, when comparing The degree of formality within a sector, lower the number of employees of male- and barriers to entry, and higher labor intensity can female-owned firms within Ethiopia’s services predict female entrepreneurs’ participation and manufacturing sectors, Bardasi and in the sector (Hallward-Driemeier 2013). Data Getahun (2008) find that female-owned firms from Sub-Saharan Africa (SSA), for example, despite being younger on average, are larger demonstrate that the industry with the lowest and have greater revenues compared to male- share of formal entrepreneurs is the textile owned firms within the same sector. And our and garment industry, where 35 percent of the study in Uganda also finds that once they are workforce is women. In contrast, the industry established in male-dominated sectors, there with the highest share of formal entrepreneurs is no difference between male- and female- is basic metal and metal products, where owned firms in the levels of capital (Campos et the female share is only three percent. The al. 2017). resistance to work in formal sectors can be due to many factors such as a lack of In addition to constraints related to access to access to formalization processes because capital, investment decisions may also be a of literacy barriers, complex administrative function of men and women facing societal processes with high costs in terms of time pressure to perform different social roles, and money, or cultural barriers in engaging impacting their sector choice and profit with business associations (Babbitt, Brown, motive (Carranza, Dhakal, and Love 2018). and Mazaheri 2015; Benhassine et al. 2018 Demenet, Razafindrakoto, and Roubaud 2016; Jayachandran 2020). 19 Introduction For example, a series of experiments that in East Java, Indonesia, show that the gender provide grants to female entrepreneurs across gap in earnings would not shrink significantly if Ghana, India, and Sri Lanka, find that these women operated in the same sectors of activity grants, when given to female entrepreneurs in as men (Buvinic, Knowles, and Witoelar 2018). multientrepreneur households, result in lower In the study by Sabarwal and Terell (2008) in returns compared with when they are given Eastern Europe and Central Asia, accounting to those who are the sole entrepreneurs in for sector significantly reduced the gender gap their households (Bernhardt et al. 2019). This in profits, from 63.1% to 37.2%, but a sizable profit gap is largely because women invest in portion of the gap remained unexplained. The their husbands’ firms rather than their own in role of sector segregation in explaining profit multientrepreneur households and opt into gaps may also vary between countries in the Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors different sectors than if they were the sole same region. For example, the significant gap entrepreneur in their household. Another in return on capital investment between female- experiment in Sri Lanka finds that female and male-owned microenterprises in Brazil is entrepreneurs invest their existing income explained almost entirely by accounting for in capital that may benefit the household, female entrepreneurs’ concentration in select while male entrepreneurs do not invest industries. However, in the same study, the in the same way (de Mel, McKenzie, and gender gap in returns to investment in Mexico Woodruff 2009). The experiment in Sri Lanka is not explained by sectoral segregation (de also finds that unconditional cash grants to Mel, McKenzie, and Woodruff 2009). microentrepreneurs led to profit increases for men but not for women. The authors Regional differences in sectoral segregation are attribute this to sector choice where they find expected as gender norms continue to evolve. that returns to investments are lower when For example, in the past decade, qualitative the proportion of female entrepreneurs in research in Kenya and Zambia finds that men a sector increases (de Mel, McKenzie, and and women both report an increasing comfort Woodruff 2009). Therefore, the association with crossing gendered lines of work (Pike between reduced capital and sector choice 2018; Evans 2014). Drawing on this notion may run both ways: lower levels of capital that context matters, this report systematically may sort entrepreneurs into some sectors, but explores data across ten countries in three also sectors which require fewer capital inputs different regions and analyzes whether sectoral may not produce the same return to capital, segregation contributes to lower profits ultimately undermining women’s earnings. for female entrepreneurs who operate in traditionally female-concentrated sectors by comparing their profits to women who cross over to traditionally male-dominated sectors. In contexts where women in MDS make more profit than those in FCS, we then explore Evolving Role of Context Specific context-specific factors that may contribute to Factors crossing over and operating in sectors that are traditionally dominated by men. Although most existing evidence suggests that gendered sectoral segregation contributes to the gender gap in profits in many countries, other constraints may be more salient in some contexts. For example, data on entrepreneurs 20 Introduction Understanding the Common Barriers to Crossing Over to Male-Dominated Sectors across Ten Countries in Three Regions We draw the data6 from a mix of existing studies conducted in Botswana (Cherchi and Kirkwood 2019), Ethiopia (Alibhai et al. 2017), Uganda (Campos et al. 2015, 2017), newly conducted surveys in Guinea, Mexico, Peru, and existing data in Cambodia, Indonesia, Lao PDR, and Vietnam of female entrepreneurs in male- dominated and female-concentrated sectors. We also draw from Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors the Future of Business survey, which entailed a multi country survey of female entrepreneurs using a social media platform (Goldstein, Gonzalez, and Papineni 2019). Across all countries where female entrepreneurs who cross over to male-dominated sectors stand to make more profits, we aim to understand the characteristics that differentiate those in the more profitable male-dominated sectors from those in the less profitable female-concentrated sectors. Given that our data are cross-sectional, we cannot separate whether the characteristics we identify are constraints that hold women back from entering more profitable male-dominated sectors or whether they result from sectoral choice. As we see from the complex literature on women’s sectoral choices, many individual and cultural constraints that drive sectoral segregation also drive gendered profit gaps. Identifying these characteristics and potential points of intervention can help maximize profits for female entrepreneurs in general, even if not through helping them cross over. However, before answering these questions, we must first understand what male-dominated and female-concentrated sectors are. In the following section, we build on previous studies and research conducted for this report to offer an overview of male-dominated and female-concentrated sectors in individual countries, within and across regions. Finally, we compare and contrast existing evidence to establish trends and highlight differences in sectoral segregation across countries in the three regions that are the focus of this report. 6 More details of the surveys can be found in tables A.1 and A.2 in appendix A of this report. In some settings we also obtained data from male entrepreneurs in male-dominated sectors and female- concentrated sectors. 21 SECTION 1 Gender-Based Sectoral Segregation of Firms 22 Section 1: Gender-Based Sectoral Segregation of Firms How Do We Determine if a Sector of a Firm is Male Dominated or Female Concentrated? T he research we synthesize uses two approaches to classify sectors in each study as male dominated or female concentrated7 (see appendix B for details on sector classification in each study). The first approach used in most of the studies in this report Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors (8 studies) classifies sectors as male dominated if men own at least 70 percent of enterprises in that sector, otherwise the sector is classified as female concentrated (except for Uganda, where a 75 percent threshold was used).8 Data from representative surveys9 were used for sector classification in Cambodia, Indonesia, Lao PDR, Mexico, and Vietnam, and nonrepresentative surveys were used in Guinea10 and Uganda11 (see appendix A for details on surveys used). This approach to sector classification builds on the percentage-based approach for sector classification that has been used in previous research. Notably, in their study of entrepreneurs in Brazil, Mexico, and Sri Lanka, de Mel, McKenzie, and Woodruff (2009) use a 75 percent threshold to determine if a sector is dominated by one gender. They consider industries in which both the male and female share of firms exceed 25 percent to be gender-mixed industries. 7 Use of the term “female concentrated” does not imply Rijkers and Costa (2012) also used a similar approach in their study of female-dominated sectors. Indeed, in about a third of nonfarm enterprises in Bangladesh, Ethiopia, Indonesia, and Sri Lanka, in the cases documented for this report, male ownership which they divide the sectors reported in their surveys into three categories: exceeded 50 percent. 8 For the studies using a 75% manufacturing, trade, and services, and provide percentages of female threshold, a lower threshold risked including sectors that participation across these sectors. Like de Mel, McKenzie, and Woodruff are more mixed and not truly male dominated; an analysis of (2009), our study goes a step further by attempting to categorize sectors as mixed-gender sectors would not capture the constraints (such male dominated or female concentrated (MDS or FCS) using a threshold. as social norms and cultural barriers) that crossovers typically face and help us distinguish In the second approach, women’s perceptions of sectoral segregation were between women who cross over and those who do not. For used instead. Sectors were classified as male dominated if at least 70 or 75 the one study that used a 70% threshold, a higher threshold percent of respondents answered that men own most enterprises in their would lead to a dramatic reduction in the sample size of business sector, otherwise the sector was classified as female concentrated. crossover women, which reduces statistical power to detect The Future of Business (FoB) survey (with a 70 percent threshold) and differences between crossover women and those who do not surveys carried out in Botswana and Ethiopia (with a 75 percent threshold) cross over. 9 Survey weights were not used this approach. This use of perception for sector classification was used used when classifying sectors in Southeast Asia but were used in and tested in a previous study of businesses by Anna et al. (1999) in Illinois Mexico. 10 The data used for Guinea and Utah, who asked survey respondents to provide their perceptions of came from a World Bank survey of small and medium-size the gender composition of their respective industries. They found that enterprises as well as qualitative this approach to sector classification results in a very close estimate of data to identify male-dominated sectors. the data they collect from the US Census. For example, women business 11 The data used for Uganda came from a survey of 735 firms owners operating in industries identified as nontraditional perceived that who were part of the Katwe Small Scale Industry Association the industry consisted of 12 percent female owners, and US Census data (KASSIDA). See the Uganda country page in appendix D for showed an average of 17 percent women ownership in those sectors. Their more details. 23 Section 1: Gender-Based Sectoral Segregation of Firms study reinforces that using women’s perceptions of whether their sector is male dominated or female concentrated is also a valid approach to classifying sectors. While several sectors can be classified as male dominated based on these two approaches, only one sector (hairdressing and personal services) is consistently dominated by female entrepreneurs, i.e., where more than 70 or 75 percent of entrepreneurs are women across all the countries in which data on that sector are available.12 Previous studies that used such thresholds to categorize sectors as female dominated were carried out in industrialized countries. For example, Sappleton (2009) defines an industry Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors as female dominated if it consists of at least 65 percent women, based on the European Social Survey. However, this approach cannot be applied in developing countries, as potentially there is relatively lower overall participation of women in entrepreneurial activity compared to men (Kelley et al. 2017; Kelley, Singer, and Herrington 2016), resulting in too few sectors that would meet this criterion. Given the paucity of truly female- dominated sectors in our studies, for the purpose of understanding gendered sectoral segregation, we consider all sectors that have less than 70 or 75 percent male entrepreneurs as female-concentrated (see box 1 and appendix table B.1 for details on cutoffs used in each country). What Are Male-Dominated Sectors and Female-Concentrated Sectors? We find consistent patterns in the sectoral concentration of female entrepreneurs across the three regions (see figure 2 and table 1 below for a summary of MDS and FCS by country).13 Female-owned firms tend to 12 No data were available concentrate in trade and retail industries, whereas male-owned enterprises for that sector for Cambodia, Indonesia, Lao PDR, or Vietnam. dominate manufacturing sectors and agriculture. In manufacturing, male- 13 It is important to note that not all sectors are captured dominated sectors include metal works and engineering, construction, in all countries and that the sampling strategy differs from other manufacturing and repair, and leather manufacturing/shoemaking country to country. The trends reported in this section are and repair (except Indonesia and Vietnam, where some of these are broad and based on the data available as well as a function female concentrated).14 We find that only two manufacturing sectors are of who was surveyed in each country. A visual representation female concentrated across almost all studies: food processing15 (except that specifies where we are drawing these trends from in Mexico) and textile manufacturing and repair. Additionally, agriculture is provided in table 1. More details of the individual (including forestry and fishing) are dominated by men across all countries countries’ survey samples are provided in appendix D. (except in Cambodia, where it is female concentrated).16 14 Metal works and engineering, and construction in Vietnam, and mother In contrast, automobile maintenance and sales are the only sectors manufacturing and repair, and leather manufacturing/ dominated by male entrepreneurs on the trade and retail side, whereas shoemaking and repair in Indonesia. many trade and retail sectors are female concentrated (i.e., are composed 15 No data were available for that sector in Guinea. of more than 30 or 35 percent female entrepreneurs). These include the 16 No data were available for that sector in Botswana, trade of textiles and footwear (except in Mexico), other retail trade (except Ethiopia, or Uganda. 24 Section 1: Gender-Based Sectoral Segregation of Firms in Mexico), trade of pharmaceutical products, perfume, and small electronic products, and wholesale trade (except in Botswana and the FoB survey). These results confirm findings from previous studies that female-owned firms tend to concentrate in trade and retail, whereas male-owned firms dominate manufacturing17 (see appendix table B.2 for a comparison of our findings on sectoral segregation compared to previous research). Services as a whole are not clearly male dominated or female concentrated in our study. A number of services are male dominated across almost all countries, such as transportation Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors and storage, electricity and gas supply, water supply and waste management, and small transport services. Still, many services are female concentrated, notably waitering and food services/ accommodation, hairdressing, personal services (no data available for Southeast Asia), and tourism services (except in Ethiopia). © Jessica Belmont / World Bank 17 In their literature review of female entrepreneurs globally, Klapper and Parker (2010) find that female-owned firms tend to concentrate in labor-intensive sectors such as trade and services, and similar trends are shown by Bardasi, Sabarwal, and Terrell (2011) in Europe and Central Asia, Sub-Saharan Africa, and Latin America and the Caribbean; Hallward-Driemeier (2013) in Sub-Saharan Africa; and Anna et al. (1999) in the United States. 25 Section 1: Gender-Based Sectoral Segregation of Firms Figure 2 Likelihood of being categorized as male dominated or female concentrated in sectors that were identified across all three regions Male-dominated sectors Female-concentrated sectors d an an a ic be ica fr sia A rib er A n Ca Am t ra as ha he th tin Sa ut La b- So e Su Sector Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors a Agriculture, forestry, and fishing b Construction c Water supply and waste management Wood manufacturing and repair Leather manufacturing/shoemaking d e and repair Textile manufacturing and repair Automobile maintenance and sales f Metal works and engineering Food processing Trade of food, beverages, and tobacco retail g Trade of textile and footware Trade of pharmaceutical products and perfumes, small domestic electronic products Other retail trade Waiter/food services and accommodation Hairdressing and personal services Small transport services Real estate activities NOTES a. Female concentrated in Cambodia, male dominated in Vietnam and Indonesia b. Female concentrated in Vietnam, male dominated in Indonesia c. Female concentrated in Vietnam, male dominated in Cambodia d. Female concentrated in Botswana, male dominated in Uganda and Ethiopia e. Female concentrated in Indonesia, male dominated in Vietnam f. Female concentrated in Indonesia, male dominated in Vietnam g. Female concentrated in Indonesia and Cambodia, male dominated in Vietnam and Lao PDR 26 Section 1: Gender-Based Sectoral Segregation of Firms Building on Previous Research This report goes one step further than previous and beverage and textile manufacturing studies and provides a more granular analysis are female concentrated. However, without of sectoral segregation. In most sectors our more information on the proportion of findings bring to light the need for more subsectors represented in Rijkers and Costa’s nuanced analyses by country and subsector. manufacturing category, it is difficult to assess whether this difference in findings is Broad Sector-Level Data May Hide due to their results being skewed by specific Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Important Differences in Subsectors subsectors within manufacturing or are a result of differences in the respective sample Our findings call for more granular analyses characteristics. by subsector. Where available, sector- disaggregated data in our studies show that Highlighting the Need for Finer classifying broad sectors as male dominated Country-Level Analyses or female concentrated may hide gender differences in subsectors. For example, Our data show that few sectors are consistently Bardasi, Sabarwal, and Terrell (2011) find that categorized as MDS or FCS across all countries the garments and leather goods are female studied in the report or even within regions. concentrated in Europe and Central Asia and We only detect homogeneity in sectoral Sub-Saharan Africa. While we also find that segregation within regions in one sector, wood textile manufacturing and repair is female manufacturing and repair, which is dominated concentrated in all countries in our study, by male entrepreneurs in Africa (Botswana, except for Botswana, leather manufacturing/ Ethiopia, Guinea, and Uganda) and Latin shoemaking and repair is male dominated America and the Caribbean (Mexico). However, in Ethiopia, Mexico, and Uganda. We also this sector classification is not consistent across see this nuance within our studies, where regions, and we find that this sector is female typically the general “textile manufacturing concentrated in Southeast Asia (Cambodia, and repair” sector is female concentrated. Lao PDR, and Vietnam). Apart from this sector, In the case of Guinea, where more granular all others have outliers within each region. For sectoral disaggregation is available, sewing is example, while a regional analysis of most of a female-concentrated subsector, but carpets our Asian studies would point to manufacturing and weaving, as well as textile dyeing, are male as a male-dominated sector, several dominated. manufacturing sectors are female concentrated in Indonesia.19 Similarly, in Ethiopia, Rijkers and Costa (2012) find that women are overrepresented relative to men in manufacturing, which includes food and beverages, brewing, grain milling, and manufacturing18 in their classification. This differs from the findings of our study, which shows that when we look at subsectors, 18 Excluding grain milling, food and beverages, distilling, and wearing leather, building, and other manufacturing apparel. 19 That women are concentrated in certain manufacturing sectors in are male dominated in Ethiopia, while food Indonesia is aligned with other studies, including Asia Foundation (2013). 27 Section 1: Gender-Based Sectoral Segregation of Firms Similarly, comparisons with other studies also highlight regional and classification-based differences. For example, previous studies find that male-owned firms concentrate in the information technology (IT) sector much more than female-owned firms in ECA, SSA, and the US (Bardasi, Sabarwal, and Terrell 2011 and Anna et al. 1999). However, country-by-country analysis in this report brings to light a more nuanced gender distribution of entrepreneurs in the IT sector. This sector is male dominated in Botswana Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors (information and communication), Indonesia (internet service providers, web search information services), and Mexico, as well as in the Future of Business study. Still, it is female concentrated in Ethiopia (information and communication) and Vietnam (internet service providers, web search information services). Other studies also find that while sector sorting patterns by gender are very pronounced, they can vary from country to country. For instance, Rijkers and Costa (2012) find that in Bangladesh and Ethiopia, female firms are engaged in manufacturing, and almost none are trading firms. In contrast, in Indonesia they find that only 3 percent of all female firms are in manufacturing, compared with 14 percent of all male firms. Therefore, we highlight that policies to encourage female entrepreneurs to cross over into a specific male-dominated industry should not be based on broad region-level sector classifications. Analyses based on gender-disaggregated sectoral data from one or two regions are not generalizable across countries, and context- appropriate analyses are needed to provide policy recommendations. Now that we have established that female and male entrepreneurs cluster in certain sectors within each country, and the need for a granular © Stephan Gladieu / World Bank look at male and female concentration in sectors and subsectors, we explore whether and how gender-based sectoral segregation/clustering is associated with business outcomes in the next section. 28 29 Section 1: Gender-Based Sectoral Segregation of Firms Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Section 1: Gender-Based Sectoral Segregation of Firms Table 1 Which sectors are male-dominated or female-concentrated for firms? Female-concentrated sectors Male-dominated sectors Few observations No data LATIN AMERICA SUB-SAHARAN AFRICA AND THE ASIA CARIBBEAN Cambodia Facebook Botswana Indonesia Lao PDR Vietnam Ethiopia Uganda Mexico Guinea Peru Agriculture, forestry, and fishing 1 2 1, 3, 4 Poultry Construction 5 6 7 Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Electricity and gas supply Mining and quarrying Water supply and waste management 8 9 9 10 MANUFACTURING Wood manufacturing and repair 13 11 12 14 15 15 16 Building materials manufacturing 17 Manufacturing furniture and related products 18 (mattresses, curtains) Leather manufacturing/shoemaking and repair 19 20 22 21 21 Textile manufacturing and repair 23 24 26 25 Other manufacturing and repair 28 27 29 21 30 Fitting and machinery Automobile maintenance and sale 31 32 33 33 33 34 Metal works and engineering 35 36 22 21 21 37 Foundry 38 Plastic and rubber industry Food processing 39 41 40 42 44 43 41 TRADE AND RETAIL Trade food, beverages, and tobacco retail 45 47 46 46 41 Trade of textiles and footwear 49 48 50 46 46 Trade of pharmaceutical products, perfume, small 51 52 domestic electronic products Wholesale trade (no automobile) 53 53 53 53 Other retail trade 46 54 46 55 46 46 46 Waiter/food service and accommodation 56 56 57 57 57 58 SERVICES Administration and support services Information and communication 59 60 Tourism services 57 57 57 Human health and social work 61 Computer programming activities 62 Engineering services Business services Leasing services Internet service providers, web search 63 information services Hairdressing and personal services 64 Transportation and storage 65 66 67 66 66 Small transport services 68 69 70 Real estate activities 71 Arts, entertainment, and recreation 72 73 Accounting and auditing Storage and warehousing 74 Printing 75 76 77 Chemistry 30 Section 1: Gender-Based Sectoral Segregation of Firms Note: FCS = female-concentrated sectors; MDS = male-dominated sectors. 1. Fishing 65. Storage and warehousing 2. Agriculture 66. Includes small transport services 3. Agriculture/farming 67. Transportation and communication 4. Forestry 68. Land transportation (except railway) 5. Electrical 69. Service: transportation 6. Electricity, gas, and water 70. Land transportation 7. Construction services (plumbing, electrical, etc.) 71. Finance, insurance, real estate 8. Waste management 72. Video club and CD rental 9. Water, sewage, and waste 73. Creative, arts and entertainment 10. Water transportation 74. Warehousing 11. Artisan: carpentry/woodwork 75. Printing/internet cafe Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors 12. Carpentry 76. Printing and reproduction of recorded media 13. Woodwork 77. Printing or reproduction 14. Wood industry 15. Manufacturing: wood 16. Wood products 17. Brick making 18. Artisan: furniture making 19. Shoemaking and repair 20. Manufacture of leather and fur products 21. Other manufacturing 22. Other 23. Textile: sewing is FCS, but Textile: carpets/weaving and Textile: dye are MDS 24. Tailoring /knitting 25. Industry: clothing 26. Manufacturing: wearing 27. Artisan: handicraft and Artisan: windows 28. Maintenance: non-auto 29. Industry: other 30. Repair of computers or household goods 31. Mechanic 32. Automotive repair and maintenance 33. Sales of motor vehicles 34. Automotive repair 35. Metal fabrication 36. Manufacture of products based on minerals and basic metal industries; manufacture of metal products 37. Fabricated metal products 38. Foundry and forgery 39. Food and beverage 40. Catering 41. Food/beverage production 42. Custom-made food preparation services 43. Industry: food 44. Manufacturing: food 45. Butcher, Baker 46. Retail trade 47. Restaurants, food sales 48. Shop/store 49. Retail trade textiles, clothing 50. Finance, insurance, real state 51. Pharmacy 52. Trade of perfumery and jewelry 53. Wholesale 54. Retail trade in ironmongery articles, chain and glass; motor vehicles, fuels and lubricants 55. Not food sales 56. Bars, cafés, and restaurants 57. Accommodation and food services 58. Food and beverage services (e.g., restaurants) 59. Communication services 60. Information services activities 61. Orientation services and social work 62. Software development 63. Telecommunications 64. Beauty salon 31 SECTION 2 The Profitarchy 32 Section 2: The Profitarchy What is the Profitarchy? I n the global Future of Business (FoB) study, a hierarchy of profits by gender—the profitarchy—emerges when comparing firm profits. Male entrepreneurs in male-dominated sectors are the top earners, female entrepreneurs in male-dominated sectors and male entrepreneurs in female-concentrated sectors are in the middle tier, and finally, female entrepreneurs in female-concentrated sectors are located at the bottom (Goldstein, Gonzalez, and Papineni 2019). Simply put, the profitarchy is the gendered hierarchy of profits for entrepreneurs driven by the concentration of Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors women in less profitable sectors compared to men, and the concentration of women in less profitable economic activities within sectors. Globally, the average male-owned firm in a male-dominated sector has slightly more than double (+116 percent) the profits of a female-owned firm in a female-concentrated sector (Goldstein, Gonzalez, and Papineni 2019). From the multicountry FoB study, which surveys businesses on a social media platform, we find that this profitarchy differs slightly in developing and 20 The sample for the FoB developed countries and further differs when we consider only the subset of study includes only business owners with Facebook microentrepreneurs in the data (see table 2). However, irrespective of these business pages. Given the small deviations in rankings, the data from the FoB study generally indicate large gap in digital access faced by women compared that women in male-dominated sectors (MDS) on average make more profits to men in developing countries, the study may be than female entrepreneurs in female-concentrated sectors (FCS), who are at the underpowered to detect a difference between female bottom of this profitarchy.20 We also find that male entrepreneurs, irrespective microentrepreneurs in FCS vs. those in MDS in the of their sectors, typically perform better than female entrepreneurs in FCS. developing country context. Table 2 The “profitarchy” rankings using the Future of Business survey with entrepreneurs across 97 countries Multicountry analysis by groups* % gap in average In MDS In FCS profit between females in FCS and females in MDS Males Females Males Females The profit rankings → 1st 2 nd 3rd 4 th 1 st 2 nd 3 rd 4 th 1 st 2 nd 3 rd 4th 1st 2nd 3rd 4th All countries in the survey: SME + microentrepreneurs 61% Developing countries: SME + microentrepreneurs 63% Developed countries: SME + microentrepreneurs 54% All countries in the survey: microentrepreneurs 31% Developing countries: microentrepreneurs 19%* Developed countries: microentrepreneurs ** 58% NOTE: FCS = female-concentrated sectors; Male-owned firms MDS = male-dominated sectors; SME = small and medium enterprises; Female-owned firms * = not statistically significant; ** = not statistically different from males in FCS but significantly lower than males in MDS. 33 Section 2: The Profitarchy Looking at the rankings of average firm profits21 by a combination of gender and sector (table 3), we find that female entrepreneurs who operate in male-dominated sectors outperform female entrepreneurs in female- concentrated sectors in all countries studied except Cambodia, where the opposite is true. While we lack the data to look at how women perform compared to men within MDS and FCS in all countries, where we do have these data, we see two trends emerge (figures 3 and 4, based on appendix table C.1). First, in Sub-Saharan Africa and Lao PDR, female entrepreneurs who cross over to male-dominated sectors are on average as profitable as Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors the men who operate in these sectors. However, these are not the trends everywhere: in Indonesia and Mexico, men outperform women irrespective of sector. In Vietnam, men in MDS continue to perform better than women who cross over. Second, among firms in FCS, male entrepreneurs continue to outperform female entrepreneurs. These two findings are in line with previous research, showing in some settings the gender gap in profits may diminish when sector is accounted for while in others, gender gaps remain even within female-concentrated or male-dominated sectors. In a study that uses data from Latin America and the Caribbean, Sub-Saharan Africa, and Europe and Central Asia but unlike our studies does not focus on microentrepreneurs, Bardasi, Sabarwal, and Terrell (2011) find the extent to which sector choice and gender of the entrepreneur explain gender gaps in profits depends on the region. On the one hand, in Sub-Saharan Africa, female-owned firms operating in MDS are as large as the male-owned firms in these sectors. While no gender gap in sales remains after the sector is accounted for, male-owned firms compared to female-owned firms perform slightly better in terms of sales revenue within the FCS. In Europe and Central Asia, sales revenue for male entrepreneurs in MDS was 70 percent greater than that of female entrepreneurs in FCS. However, the gender gap in sales remains after accounting for sector, indicating that other factors beyond sectoral choice contribute to this gap, an issue we explore further below. On the other hand, in Latin America and the Caribbean, female-concentrated sectors have firms that are larger and on average have a higher value added compared to firms in male-dominated sectors. However, male-owned firms in FCS still outperform female entrepreneurs. While this trend is not what we see among microentrepreneurs in our studies in Mexico or Peru, it is similar to what is observed in Cambodia (see table 3). 21 The studies differed in their sampling and approach to calculating profits and hence While gender-based sectoral segregation is one main reason women in we caution against making cross-regional comparisons in FCS perform relatively poorly compared to women in MDS across almost all absolute profits. More details on the diversity in survey countries in our data, the extent to which sectoral segregation explains the approach can be found in appendix A, table A.1, and gender gap in profits of male and female microenterprises also varies by more details of the firms that were surveyed are available in country and region. Analysis from Indonesia, Lao PDR, and Vietnam shows table A.2. 34 Section 2: The Profitarchy that adding controls for the sector of activity does not significantly reduce the gender gap in the performance of microenterprises. Moreover, male-owned firms in any sector, including FCS, continue to perform better than women- owned firms in Indonesia, Mexico, and Vietnam. Nevertheless, in Sub-Saharan Africa and Lao PDR, once women cross over to male-dominated sectors, they perform as well as men in MDS. Here we discuss ‘horizontal’ and ‘vertical’ segregation—the two key contributors to the gendered hierarchy of profits. Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Table 3 The “profitarchy” ranking for the individual countries in our analysis Country* Males in Females in Males in Females in % gap in average MDS MDS FCS FCS profit between females in FCS and females in MDS Guinea 1 1 — 2 90.24% Botswana 1 1** — 3 119.11% Uganda 1 1 — 3 140% Ethiopia ­— 1 — 2 80% Mexico 1 3 2 4 50.55% Peru — 1 — 2 70.5% Lao PDR 1 1 1 4 79.5% Vietnam 1 2 2 4 31.9% Indonesia 1 3 1 4 66.9% Cambodia 3 3 1 2 -18.5% NOTE: FCS = female-concentrated sectors; MDS = male-dominated sectors; * = calculated using the difference in average monthly profits (appendix C) between female entrepreneurs in MDS-FCS/FCS*100; ** = females in MDS rank lower than males in MDS if windsorized results are used and when jointly owned firms are removed. Appendix A provides details of the data used in each country. 35 Section 2: The Profitarchy Figure 3 Average profit for female entrepreneurs in FCS vs. MDS for all countries (USD) Profits comparison (monthly, USD) Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Figure 4 Average profit for female entrepreneurs in FCS vs. MDS for all countries except Botswana (USD) Profits comparison (monthly, USD), without Botswana Note: FCS = female-concentrated sectors; MDS = male-dominated sectors. * Sales were used in Lao PDR instead of profits due to data constraints. 36 Section 2: The Profitarchy Horizontal Segregation: The Role of Sectoral Segregation in Creating the Profitarchy The profitarchy emerges in part due to a or occupations women should choose. Findings concentration of female entrepreneurs in from agribusiness entrepreneurs in Nigeria (Das different economic activities compared to male et al., forthcoming), show that when given a entrepreneurs. This type of segregation by choice of 11 different value chains, 54 percent sectors, known as “horizontal segregation,” of entrepreneurs chose to enter into poultry, is not surprising given that previous literature one of the least profitable value chains among strongly suggests that women concentrate the choices, and that women were significantly Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors in less profitable, fewer, and more crowded more likely to choose poultry than men. Their sectors and industries. As discussed earlier, analysis points to restrictive gender norms as multiple studies find that female entrepreneurs preventing many women from crossing over largely seem to operate in sectors into the more lucrative value chains, along characterized by lower initial investments and with differences in land ownership and work growth, lower barriers to entry, and higher experience in the chosen value chain, as well as labor intensity compared to traditionally male- differential access to tertiary-level education. dominated sectors such as manufacturing, construction, and mining (Nissan, Castano, Overcoming this horizontal segregation by and Carrasco 2012; Hallward-Driemeier 2013; operating an enterprise in a male-dominated Rosa and Sylla 2016; Bernhardt et al. 2019). sector can help bridge the gender gap in Furthermore, female entrepreneurs selecting firm performance in MDS in many, but not into a small number of sectors contributes to all, countries where this was explored. On sector crowding, making these sectors less the one hand, in Sub-Saharan Africa and Lao profitable to operate in. A study in Ghana finds PDR, we see that once women cross over to that female microentrepreneurs continue to male-dominated sectors, they perform as well crowd into fewer sectors, despite the absolute as men in MDS. On the other hand, in some number of women in nonagricultural self- countries like Indonesia and Mexico, while employment being far greater than the number female entrepreneurs in MDS perform better of self-employed men (Hardy and Kagy 2020). than female entrepreneurs in FCS, male-owned This demand-side constraint is also reflected in firms in both MDS and FCS continue to perform the qualitative data from our study in Guinea, better than female-owned firms. In Vietnam, where women operating in FCS report that women in MDS perform at par with males they are less likely to overcome the constraint in FCS, but men in MDS continue to make of finding customers compared to women who more profits. These findings are consistent cross over to MDS. with the results of the multicountry FoB study by Goldstein, Gonzalez, and Papineni (2019), Other studies find that socially constructed where we see a similar hierarchy of profits push and pull factors such as safety, care emerge in middle-income countries indicating responsibility, mobility, time use, training, that in addition to horizontal segregation, there and economic constraints may drive women’s are patterns of vertical segregation as well, selection into less profitable activities and i.e., men do better than female entrepreneurs sectors. For example, one reason for selecting irrespective of the sectors they operate in. a less profitable income-generating activity may be social norms around which businesses 37 Section 2: The Profitarchy Vertical Segregation: Female-Owned Firms within the Same Sector Continue to Lag behind Male Firms While this report is focused on horizontal key factors exacerbate vertical segregation segregation, it is worth examining some specifically: clustering in less profitable roles reasons why in certain countries—e.g., or activities within the same sector, lack of Indonesia, Mexico, and Vietnam—firms owned access to capital, and the discrimination by men continue to outperform those owned faced by female entrepreneurs. The following by women even when they operate in the same paragraphs explore how these key factors sector. This pattern of hierarchical or vertical contribute to vertical segregation. Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors gendered segregation is commonly known as the ‘glass ceiling.’ Hallward-Driemeier (2013) in her analysis of firms in Sub-Saharan Africa emphasizes that Before turning to vertical segregation explicitly, women tend to crowd into less profitable we need to see if there is possible sorting subsector activities, even in more profitable within the set of sectors that make up the industries like manufacturing. She finds MDS. Goldstein, Gonzalez, and Papineni evidence of gender segregation within sectors, (2019) identify two types of male-dominated with most female-headed businesses clustered sectors: those that are capital intensive and in activities with low profit margins such as those reliant on skills that may be more labor food preparation, textiles, and garments in intensive. It is possible that women, who often the manufacturing sector. This is supported by lack access to capital, sort into the latter, which qualitative research with women who work in could have lower profits. To examine this in construction in South Africa, who report that Indonesia, Mexico, and Vietnam, we carried out certain aspects of the business such as bidding an analysis of profit difference between male- are only open to men while women carry out and female-owned firms within MDS while work in the background (Aneke, Derera, and accounting for specific sectors. We found that Bomani 2017). A study in Ghana finds that men a significant gender gap in profits persisted, outperform women entrepreneurs within the indicating that the gap is not entirely because traditionally female-concentrated garment- women are choosing to concentrate in less making industry (Hardy and Kagy 2018). When profitable male-dominated sectors. unpacking the cause of this gap in profits, the authors found that male-owned firms were Instead, the gender gap potentially remains older, had more capital, assets, and sales, and due to gender-related barriers that influence their owners worked longer hours. However, productivity in ways other than through sectoral even after accounting for these factors, the choice. Similar to horizontal segregation, gender gap in profits remained large. Through a wide array of gender-specific constraints an experiment which involved randomly contribute to vertical segregation. These increasing demand for gender-neutral clothing, include differences in skill, education, and the authors establish that female-owned hours worked, self-imposed limitations to firms have lower demand and access to business growth and a desire for home-based customers compared to male-owned firms. flexible work, and a lack of access to networks This low demand may be due to crowding (Rose and Hartmann 2004; Goldin 2014a; of female entrepreneurs in specific activities. Carranza, Dhakal, and Love 2018; Hallward- For example, female-owned firms produce Driemeier 2013). Additionally, several other women’s clothing, while men’s shirts are 38 Section 2: The Profitarchy typically produced by male-owned firms. In this loss of prestige, men would be incentivized way, segregation even within sectors continues to create social norms and barriers that keep to impact women’s business outcomes. women and the associated ‘pollution’ out of these occupations. Another key factor driving vertical segregation is access to capital and gendered differences In the case of female entrepreneurs, as in investment decisions. Several experiments asymmetry of information arises, discrimination have highlighted reduced returns to capital for can come from both within the industry, i.e., female entrepreneurs based on their choice men operating in the same sector, and from of investments (Bernhardt et al. 2019; de customers. Qualitative research with female Mel, McKenzie, and Woodruff 2009). Another entrepreneurs in MDS documents that while Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors study which seeks to explore the male-female operating in MDS, women report many gap in performance within a mixed-gender barriers to success, including lack of training sector (vegetable selling) in India finds that by the industry, discrimination, overcoming sellers’ and buyers’ behavior do not drive the men-only networks, and stereotyping by both gender difference in profits when business the industry and customers who perceive that characteristics are held constant (Delecourt men would perform better in traditionally and Ng 2019). In terms of characteristics, they male-dominated fields (Haupt and Ndimande find that capital constraints faced by female 2019; Aneke, Derera, and Bomani 2017). This entrepreneurs that translate to lower inventory discrimination from customers toward female are potentially the primary cause of the gender entrepreneurs in MDS was also seen in our gap in profits. qualitative data in Uganda, where clients placing orders in the carpentry and metal Discrimination may be an another explanation fabrication sector preferred placing orders with for lower profits of enterprises in female- male-owned firms (Campos et al. 2015). This concentrated sectors. We draw upon Goldin’s preference may arise from the asymmetry of (2014b) ‘pollution’ theory of discrimination information that Goldin (2014b) refers to in her which aims to explain why vertical segregation ‘pollution’ theory. occurs for women employed in male- dominated occupations. She posits that “male employees discriminate against prospective female employees as a way of protecting their prestige in an asymmetric information context” (p. 316). According to this theory, prestige for men is derived from how society views their occupation. An asymmetry in information arises as women, a group that is traditionally perceived as less skilled, enters male-dominated occupations. Society is unable to distinguish whether women’s entry in the male-dominated occupation reflects parity in skills between men and women in the occupation or if the occupation itself is no longer as highly skilled or prestigious. In this absence of information, society may lean toward presuming the latter. To prevent this 39 Section 2: The Profitarchy What Do Our Data Tell Us about the Drivers of Sectoral Segregation in These Countries? While vertical segregation will continue to be Mexico, Peru,24 Vietnam, and across the FoB an issue in explaining the male-female gap study’s multicountry data, controlling for in profits, horizontal segregation through other characteristics somewhat shrinks the the sector of operation plays a significant performance gap between women operating role in profits for female entrepreneurs. In all in MDS and FCS; however, the gap remains countries except Cambodia, women in MDS sizeable and statistically significant. This Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors have higher profits (or sales) than women in suggests that in these countries, both observed FCS (figure 3 and figure 4).22 While gender- entrepreneur characteristics and sector of based sectoral segregation is one main reason activity help explain why women in MDS that women in FCS perform relatively poorly outperform women in FCS. In Indonesia, we compared to women in MDS across almost all find that women in MDS have higher profits countries, we know little about the extent to than women in FCS even after accounting for which the individual, household, and business some key observed business characteristics characteristics are driving the profit gap and for entrepreneurs’ personal and household between female entrepreneurs in MDS and characteristics and accounting for these those in FCS. does not shrink the gap in profits. However, in Botswana, the gap in profits shrinks To unpack what drives the profit difference significantly and loses statistical significance between female entrepreneurs in MDS once business characteristics are taken into and those in FCS, we adjust for business account. This suggests that, in this case, the characteristics (other than sector) and for difference attributable to sector may come entrepreneurs’ personal and household largely from differences in the characteristics characteristics. We find that the difference in of the businesses between MDS and FCS. performance between female entrepreneurs At the same time in Ethiopia and Uganda, in male-dominated sectors and those in where we account only for a few business female-concentrated sectors remains, but characteristics, the gap in profits between the magnitude of the gap in profits/sales female entrepreneurs in MDS and those in diminishes in most countries. While we account FCS remains. Furthermore, in Guinea and Lao for only a limited number of characteristics, PDR, accounting for individual characteristics these analyses suggest that perhaps the removes the profit gap between women in choice of sector itself continues to play the MDS compared to those in FCS. This may main role in the underperformance of women indicate that, in these cases, individual factors in FCS23 but that individual, household, that largely drive sectoral segregation or are a and business characteristics also contribute result of sectoral segregation drive the gaps in to this gap. For example, in countries like profits between women in MDS compared to those in FCS. 22 As discussed above, in Cambodia, the women in FCS outperform women in MDS, and this relationship exists even when controlling for observable individual, household, and business characteristics. 23 We carry out regression estimates of the gap in profits among female entrepreneurs due to sector of operation. We then check if the coefficients for the gap in profits attributable to being a female in MDS remain constant in magnitude and significance when we account for business, individual, and household characteristics. While we are limited in the variables we can account for and their temporality, we hope to provide suggestive evidence as to where the differences between women in MDS and women in FCS are coming from. More details of these analyses and adjustments can be seen in table C.2 in appendix C. 24 In Peru, accounting for household characteristics increases the profit gap between women in FCS and those in MDS. 40 Section 2: The Profitarchy characteristics of women that are successfully In the next section, we aim to unpack what operating in male-dominated sectors and are enables women to enter and operate in on average earning higher profits can help profitable (male-dominated) sectors. We design policies and programs that enhance identify and explore individual, household, these qualities. Policies encouraging diversity and business characteristics that potentially in sector choice may not only lead to greater drive the selection into MDS or FCS. Due to profits for female entrepreneurs, as seen in the the cross-sectional nature of the data we draw profitarchy patterns, but may also spur growth, from, it is also possible that these differences as talent is allocated more efficiently across not only drive, but also result from sectoral sectors. Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors segregation. In either case, identifying the © Shynar Jetpissova / World Bank 41 Section 2: The Profitarchy Box 1 The curious case of Cambodia In contrast to this finding from nine studies synthesized in this review, an interesting trend is seen in Cambodia, where MDS are less profitable and crossing over into male-dominated sectors is associated with lower profits Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors for women. The types of sectors characterized as male dominated and female concentrated in Cambodia are similar25 to those of other countries in the region, and the results are not driven by one or two sectors. Moreover, the pattern holds across multiple years, suggesting that the finding is not driven by either a data aberration or something specific to the survey year. What then may be driving this surprising finding? The unique results in Cambodia are most likely due to its economic, social, and political history. In 1970, a military coup overthrew Prince Norodom Sihanouk, who had governed Cambodia since its independence from France in 1953, and the country entered a period of civil war. In 1975, the Khmer Rouge took power and embarked on a massive campaign to radically transform Cambodian society, during which approximately 1.5 to 2 million people—one quarter of the Cambodian population in 1975— died as a result of politically motivated killings, starvation, exhaustion from overwork, lack of medicine, and increased exposure to malaria (Heuveline and Poch 2007). Adult men were most likely to die (de Walque 2005), leading to a skewed sex ratio which, while it has partially recovered from the rate of 89 men per 100 women in 1980, remains at 95 men per 100 women today (United Nations 2019). In addition, under the Democratic Kampuchea regime governed by the Khmer Rouge from 1975 to 1979, schools were decimated, and educated professionals and urban dwellers were forced to relocate or were killed (de Walque 2006). Economically, private ownership was banned, and entrepreneurship disappeared (Chhair and Ung 2016). Although households could engage in small trade or handicraft activities during the following regime of the People’s Republic of Kampuchea from 1979 to 1989, Cambodia did not transition back to a market-based economy until after 1989 (Chhair and Ung 2016). 25 Agriculture, forestry, and fisheries are FCS in Cambodia and MDS in other countries. 42 Section 2: The Profitarchy Demographic trends resulting from Cambodia’s recent history may have influenced the composition and performance of traditionally male-dominated sectors. Research from other countries has shown that Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors when the sex ratio is skewed with more men than women, there are lasting effects on gender norms about women’s work (Grosjean and Khattar 2019), and occupational segregation along gender lines is more entrenched (Baranov, de Haas, and Grosjean 2018). While similar research has not been done exploring cases with sex ratios skewed toward more women than men, it is possible that the skewed sex ratio in Cambodia linked with the reign of the Khmer Rouge influenced norms about occupational segregation or shifted the gender composition within these MDS and FCS. Moreover, de Walque (2006) showed that the age and education gap between spouses lowered in the period after the Khmer Rouge, which may shift women’s bargaining power in the household. An increase in bargaining power could enable women to have more access to capital or resources to support their activities. Alternatively, changes in roles in the household and small businesses may be linked with higher rates of disability among men who experienced the Khmer Rouge regime as children, adolescents, or young adults. Businesses in Cambodia are relatively recent. As entrepreneurship was decimated under the Democratic Kampuchea regime, the oldest businesses in Cambodia were founded in 1979 (Chhair and Ung 2016). A rise in entrepreneurship did not occur until after 1993, when the Kingdom of Cambodia began to liberalize the economy more fully (Chhair and Ung 2016). Perhaps the relatively recent establishment of businesses may have changed the dynamics of traditionally male- or female-dominated sectors. Another possibility is that lingering effects from the decline in education during the Khmer Rouge period and the following years also affected the skill base of the population, affecting the productivity of different sectors. 43 Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors 44 Profitarchy Unpacking the SECTION 3 Section 3: Unpacking the Profitarchy I n an attempt to explore how to overcome horizontal segregation and support female entrepreneurs to cross over to more profitable male-dominated sectors, we analyze what is different about women who cross over from those who do not. In table 4, we present a summary of common characteristics associated with female Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors entrepreneurs in male-dominated sectors across the nine countries where women operating in male-dominated sectors (MDS) outperform those in female-concentrated sectors (FCS). We also corroborate these findings from previous research as well as the multicountry data collected via the Future of Business survey. 45 Section 3: Unpacking the Profitarchy Table 4 Characteristics correlated with female entrepreneurs crossing over to MDS and the direction of the correlation Positively correlated Negatively correlated No observable correlation No data // No controls developing developed Botswana Indonesia countries countries Characteristic Lao PDR Vietnam Ethiopia Uganda Mexico Guinea Peru FoB FoB Age Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Being married (vs. single) a b c d Household size // e Number of children // g f i h Father’s education Mother’s education Father ran a business/owned an enterprise j k Mother had a business/owned an enterprise l m n Household asset, quality of dwelling, or o p q s r t source of drinking water Education u Self-efficacy, locus of control, and decision- v w x making power Cognitive ability (intelligence/abstract y z aa thinking) Support from spouse in running the business // ab ad ac ae ae (or in the form of money) Had a (male) role model or mentor af ag ah Inherited the business Exposure to male-dominated sectors through aj ai work experience Joint ownership/has partners ak // // Age of the business // Number of workers al // am // // an Worked as an employee in the last 7 days 46 Section 3: Unpacking the Profitarchy NOTES FoB = Future of Business; MDS = male-dominated sectors; OLS = ordinary least squares. a. Association is no longer significant after controlling for u. Having less than primary education was negatively whether spouse owns a business himself and whether the associated with being in an MDS. Other categories of woman has a male role model. education such as secondary education and more than b. Association is no longer significant (in the OLS regression secondary education were not associated with crossing over. with the full sample) after controlling for exposure v. A negative association is seen with a household decision- origination, i.e., entry to sector was a result of someone’s making power index, but a positive association is seen with suggestion, was offered a job by family or friends, observed making business decisions. others in sector, and worked for a wage in the sector. w. No association with internal locus of control. A positive c. Relationship continues to be significant in developing association is seen with self-efficacy that is no longer countries after accounting for socioemotional skills. significant after accounting for exposure-enabling factors, d. Relationship no longer significant when controlling for i.e., entry to sector was a result of a suggestion, job offered Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors five noncognitive skills: generalized self-efficacy, error by a family member or friend, observing others, working for management score, entrepreneurial identity, career fit a stranger, or training. confidence: more committed to staying in sector, and career x. No association with self-esteem but a positive association fit confidence: good relationships in sector. with higher internal locus of control – which is an indicator e. Only significant when controlling for mentorship. No longer of how much agency and control an individual feels they significant when controlling for cognitive skills. have over their life. f. Association is no longer significant when factors such as y. Positive association with Digit Span Score, no association household size and socio economic status are accounted with Ravens Test Score. for. z. Negatively associated with Ravens Test score but no g. Not associated with number of children nor with at least association with Digit Span Score. one child in the household between the ages of 0 and 2 and aa. Positive association with Ravens Test score but no with at least one child in the household between the ages association with Digit Span score. of 3 and 5. ab. Husband helped in some way to begin the business. h. Not associated with at least one child in the household ac. Spouse is the main source of borrowing emergency funds. between the ages of 3 and 5 but associated with at least ad. Spouse/partner started the business; analysis with no one child in the household between the ages of 0 and 2. controls shows a positive effect of assisting with advice i. Not associated with at least one child in the household but negative association when support was in the form of between the ages of 0 and 2 but associated with at least 1 money and assets. child in the house between the ages of 3 and 5 years of age. ae. Association is negative when the spouse only supports in j. Defined as, father was owner/manager of a firm in a male- the form of money, but positive when spouse helps in the dominated sector when respondent was a child; no longer form of registering with authorities to get a license and is a significant when controlled or after controlling for ‘Any last co-owner. five jobs were in an MDS’ or ‘Exposed to the current sector af. Became insignificant in the OLS regression with the of operation by someone.’ matched sample after controlling for exposure origination k. The variable indicates father owned or managed family (idea originated from self vs. others) or exposure-enabling enterprise. Only significant in matched analyses. factors (exposure was self-initiated or came from training or l. Only significantly associated with reduced likelihood of others). crossing over when businesses that are not jointly owned ag. Significant positive association with male mentors but are considered. significantly less likely to cross over with female mentors; m. Crossing over is negatively associated with the mother or only positively associated with male role models. father working on a farm but positively associated with ah. Role model was a man was significant and positively father having a government wage job. associated, but no longer significant when controlling for n. Association is no longer significant after accounting for noncognitive skills. exposure origination, i.e., entry to sector was a result of ai. Having a first job in a non-male-dominated sector is someone’s suggestion, was offered a job by family or negatively associated with being a crossover. friends, observed others in sector, or worked for a wage in aj. Crossover women report prior experience in the sector the sector. before starting the MDS business. Having any previous o. Car and internet connection are considered as a measure of work experience is negatively correlated with crossing over, household asset and were positively associated, but owning although having any previous experience as an entrepreneur at least one plot of land and livestock were not associated is positively correlated with crossing over. with being a crossover. ak. Based on test of differences of means between crossover p. Positively associated with asset index for the household. firms and noncrossovers firms. Sensitivity analysis also finds q. Walls of brick or concrete block and ceiling of reinforced that ‘profitarchy’ results are partly driven by the profits concrete have positive correlation while source of drinking of high-performing outliers who tend to be foreign-born water has no significant association with operating a owners or those who jointly own their business with another business in an MDS. owner. r. Source of drinking water being inside or near the house in al. Based on t-tests, test of differences of means between dry season is positively associated with crossing over, the crossover firms and noncrossover firms. drinking water source being nearby or in the house in wet am. Positively associated with number of full-time family season is negatively associated with crossing over. employees. Negatively associated with number of full-time s. Cooking source is used as a measure of household asset. women employees. t. Water source and cooking source are used as a measure of an. It is for unpaid workers in the last four weeks. household asset. 47 Section 3: Unpacking the Profitarchy The Correlates of Crossing Over to MDS This section summarizes some of the findings across the ten studies on the key factors associated with crossing over to male-dominated sectors. These factors include spousal support, domestic responsibilities, business partnerships, exposure to the field through work experience, parents, role models, mentors, access to capital and wealth, number of workers, as well as cognitive and noncognitive skills and abilities. Spouses, Domestic Work, and Partnerships Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors → Being married versus being single: associated with an increased likelihood of In Nigeria, a study among female operating in male-dominated sectors. entrepreneurs who operate technology- An entrepreneur’s spouse’s entrepreneurial based firms finds that unmarried women experience may be one potential pathway tend to be similar to men in terms of through which marital status can influence the diversity of their networks, while the the decision to cross over in Ethiopia. Once networks of married women are focused the spouse’s entrepreneurial experience is on their family (Aderemi et al. 2008). Given accounted for, the association with marital the role of social and business networks status disappears. This is also reinforced by in operating in a sector, this can be one the finding that among married women who pathway through which marital status crossed over to MDS, 42 percent reported negatively impacts women’s ability to that the idea for the business came from cross over into MDS. We find support for their spouse. In Ethiopia and Peru, the this in Uganda, where female crossovers spouses of crossovers are themselves more were in fact more likely to have never been likely to be engaged in running or managing married, but at the same time less likely a business. In both countries, there is an to be divorced or widowed. However, association between crossing over to MDS the association between marital status and having a spouse who runs a business and crossing over is not consistent across of his own. It is possible that spouses or settings. For example, marital status is not partners who work in businesses themselves associated with sector choice in Botswana, are providing some exposure that could Indonesia, Lao PDR, or Peru. On the other facilitate crossing over in some settings. hand, in Ethiopia,26 Mexico, Vietnam, and for entrepreneurs in developing countries in → Spousal support among those who are the FoB study, we find that being married is married: Among married entrepreneurs, 26 In Ethiopia, the association between being married and being a crossover is no longer statistically significant after controlling for whether the spouse owns a business himself and if the woman has a male role model. 48 Section 3: Unpacking the Profitarchy spousal support appears to be consistently When moral support and money are associated with crossing over. In all the provided along with other types of support countries where the role of the spouse in running the business, they are only in crossing over was studied, a positive then positively associated with being in a association between some form of spousal male-dominated sector. When looking at support and operating in MDS was developed and developing countries, the observed. In Botswana and Ethiopia, women FoB study finds that in developed countries who cross over to male-dominated sectors the relationship is no longer significant when are more likely to report that their spouse accounting for generalized self-efficacy, provided support by providing the business error management abilities, entrepreneurial idea. Similarly, spouses are more likely identity, commitment to staying in the Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors to have created the business for female sector, and having good relationships in the entrepreneurs who cross over to MDS in sector. Similarly, in Ethiopia, we find that Guinea. the association between being married and being a crossover is no longer significant In Uganda, spouses are an important source after controlling for the influence of others of finance, with 10 percent of crossovers (spouse owns a business himself and the reporting that their spouse was a primary woman has a male role model). This absence source of borrowing money compared of a relationship between marital status to 3 percent of noncrossovers. This was and crossing over to MDS when accounting also seen in the survey in Ethiopia, where for noncognitive skills and the influence of crossovers reported that their spouses others indicates that it is possible that for provided financial support. Furthermore, the those who are married, spousal support and qualitative research from Ethiopia finds that the strengthening of noncognitive skills are partners helped crossovers by expanding pathways for crossing over into MDS. their networks and providing startup capital for the crossover business. In Botswana, → Having fewer domestic responsibilities: crossovers are also more likely to report that On average across 64 countries, women their spouse shared know-how from their contribute more than three-fourths of own business experience, and supported unpaid care work, averaging 265 minutes them with business registration, acquiring a per day compared to men’s 83 minutes license, and in skills-related aspects of the per day. This influences their labor market business. choices (International Labour Organization 2018). Entrepreneurship can offer more The Future of Business multicountry study flexible working hours and location, attempts to unpack the types of spousal facilitating the combination of labor market support associated with crossing over to work and domestic tasks (Bahramitash and male-dominated sectors in their sample of Kazemipour 2011; Chen 2001; Manning married female entrepreneurs on a social- 1998; Vanek 2013). This may also orient media platform. It finds that spousal support sectoral choice, if working hours and in the form of money or moral support alone location are more flexible in certain sectors was negatively associated with crossing of activity. over, while support in administrative tasks such as licensing and registration, and co- In several countries, crossover women have ownership alone is positively associated characteristics that would suggest they face with being in a male-dominated sector. fewer time constraints related to domestic 49 Section 3: Unpacking the Profitarchy tasks than women in FCS. For example, no longer perform as well as men in MDS in Guinea, Peru, and Vietnam, crossovers (see note in table 3). are less likely to have small children who would require more care, and in Lao PDR, Research from Botswana attempts to unpack crossovers are less likely to have elderly who co-owns these firms and finds that 18 household members who also may require percent of the female-owned businesses assistance. However, the trend is not who cross over into MDS co-own their universal. In Ethiopia, crossovers are more businesses with their spouses. Among likely to have a greater number of children. these female crossovers they find that Evidence is mixed for Indonesia, where those businesses co-owned by spouses far women with children aged 0–2 are more outperform the firms in male-dominated Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors likely to operate in MDS, but crossovers are sectors that are co-owned by other men or also more likely to operate their businesses run by women alone. In Ethiopia, qualitative outside of the home, where it would be data also find that husbands were often more challenging to simultaneously conduct partners in the business, providing financial entrepreneurial and domestic tasks. In support or support in the form of networks, addition to the association with childcare skills, or management. Due to these reasons, responsibility, there is some evidence from partnerships, particularly with a spouse, Lao PDR and Vietnam that reduction in other become a success strategy for women- household responsibilities is also associated owned businesses operating in MDS. with crossing over. For example, in Lao PDR, crossovers are more likely to have water in Interestingly, selecting a partnership, the home during the dry season, which can especially with a male business partner, may reduce the time needed for fetching water. not be what drives crossing over to MDS but Similarly, in Vietnam, crossovers are more instead can be a survival strategy for those likely to be in households that use cooking facing gender-based barriers to operating fuels that require less time to gather and in a sector dominated by men. A qualitative ignite. study of female entrepreneurs operating in the construction sector in South Africa finds → Having joint ownership or a business that women overcome the gendered barriers partner: In Uganda, there is no association of working in this predominantly male sector between joint ownership/partnership and by strategically partnering with men to avoid crossing over. However, the studies in harassment and the challenges of operating Botswana, Ethiopia, and the FoB survey in an environment that is not welcoming of find that crossover businesses are more them (Aneke, Derera, and Bomani 2017). likely to be jointly owned compared to The following quotes from participants in noncrossovers. In Botswana, female-owned the study explain the reasons why women in firms that operate in partnership with any MDS partner with men in the sector and how male report being able to access more a division of labor is formed: “In general, capital. Access to capital may help women men hold prominent positions and are at the working in capital-intensive male-dominated hub of affairs in the committees responsible sectors perform as well as men in MDS. for the awarding of tenders… this gives men So much so that in Botswana, when the the upper hand…” “Men can go for bidding subgroup of jointly owned enterprises is while women work from the background…” removed from the analysis, women in MDS (Aneke, Derera, and Bomani 2017:45). 50 Section 3: Unpacking the Profitarchy Exposure to Male-Dominated Sectors → Having a (male) role model or mentor: and generally boosting women’s confidence Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Having a male role model while growing about their fit and other noncognitive skills up appears to be positively associated with that help women operate in and select male- crossing over into male-dominated sectors dominated sectors. in Mexico and Uganda. While it was not associated with crossing over in Ethiopia The study in Uganda attempts to unpack in the full sample, the top 25 percent of the role of role models and finds that earners among crossovers were more likely crossovers are likely to be introduced to to report having a male role model while their sectors by men, including fathers, growing up. The FoB study finds that having male friends, male community members, a male role model is significantly associated and male family members. The influence of with crossing over in pooled data from all these role models is not limited to owning 97 countries. Additional analyses highlight an enterprise in that sector but exists above the various pathways through which role and beyond that. Qualitative research models may be associated with crossing in Uganda finds that male role models over. In the subset of developing countries, and mentors help by providing training the association between having a role to gain skills specific to male-dominated model and crossing over disappears when sectors or coming up with the idea for the noncognitive skills such as generalized business. Relatedly, a study of vocational self-efficacy, career-fit confidence, and training applicants in the Republic of entrepreneurial identity are accounted Congo (forthcoming study by the World for. In Uganda, the association between Bank) emphasizes both the importance role models and crossing over disappears of exposure to MDS and women’s when exposure to the sector is accounted disadvantaged access to early exposure. The for. These analyses highlight the various research finds that prior technical experience pathways through which mentors and role in MDS makes it more likely that women will models influence women to cross over to opt for training in MDS. However, women male-dominated sectors and are reminiscent are less likely to have network connections of the pathways through which marital with people operating businesses in MDS, status may be associated with crossing over and men are more likely than women in the developed country analysis in the to obtain technical experience through FoB study and in Ethiopia. For example, their network connections. As a result, role models could be providing exposure while technical experience is a correlate through suggesting or offering jobs in of selecting MDS, especially for women, these male-dominated sectors, providing women may have a hard time gaining that introductions and networks information, experience. 51 Section 3: Unpacking the Profitarchy Noncrossovers, on the other hand, are more because I was discouraged by school, so he likely to be introduced to their sector by advised me, he encouraged me.” mothers and teachers, highlighting the role that teachers and female role models may → Father or mother running an enterprise: play in perpetuating gendered segregation A study from Denmark finds strong of sectors. Data from Mexico, where having empirical support for the hypothesis that a female mentor is negatively associated entrepreneurial parents serve as role models with crossing over into male-dominated for their children (Hoffmann, Junge, and sectors, also reinforce this finding. This Malchow-Møller 2015). Most interestingly, is likely due to the fact that female role they find that while having self-employed models are more likely to operate in female- parents in general increases the probability Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors concentrated sectors. of becoming self-employed, the relative effect of a self-employed father is roughly While it is possible that having a role model twice as high for men as for women. or mentor can increase the likelihood of In contrast, the relative effect of a self- crossing over, it is just as likely that we employed mother is similarly about twice as see this association only because women high for women as for men. These findings looking to cross over seek a male role also hold when they control for parental model or mentor to help them develop wealth and work experience from the skills and networks in a field that is new parents’ firms and when they disregard cases to them (Martin and Barnard 2013). An in- where the offspring takes over the family depth qualitative study in Guinea uncovers business. It even holds to some extent when evidence of this complex relationship they exclude individuals that start in the between role models and crossovers. In same industry as their parents. this study of women who were working or receiving training to work in MDS, In Botswana, having a father who was some women articulated a long-standing an owner/manager of a firm in a male- interest in working in crossover sectors dominated sector when the respondent and described their efforts to identify male was a child is associated with operating in role models to help them succeed. Other a male-dominated sector. This potentially women indicated that they had not had highlights the important role of early these aspirations initially but had been exposure. When the study accounts for encouraged by friends and family members, ‘female entrepreneurs’ last job being in both men and women, to consider training the male-dominated sector’ and for ‘being or working in MDS. Some of these mentors exposed to the sector by someone,’ the helped with getting admission to technical father being an owner or manager in MDS schools, securing apprenticeships, and is no longer associated with crossing over, paying for transport costs. For example, one indicating that exposure and networks that woman said that she initially got the idea to result in the first job in a male-dominated go into plumbing when she was at a building sector are a potential pathway through site where her aunt worked as a mason. She which fathers in MDS can influence crossing decided to go to technical school after she over. This path dependence is corroborated failed her Brevet exam and her boyfriend, by research from Uganda, where both initial who was himself a plumber, suggested that employment in a female-concentrated she pursue plumbing. She said, “I told him sector and having a mother who managed that I couldn’t continue with my studies or owned an enterprise are independently 52 Section 3: Unpacking the Profitarchy associated with a reduced likelihood of in Botswana, we find support for the role being a female entrepreneur in a male- of previous exposure to and training in dominated sector.27 male-dominated sectors in helping women cross over to MDS. Women who cross over In addition to providing exposure to a to MDS in Guinea not only report previous sector, parents may directly enable crossing work experience, but we also find that over by passing on their business which having any previous work experience is operates in a male-dominated sector. In the negatively correlated with crossing over, FoB study, inheriting a business is associated although having any previous experience with crossing over. In Guinea, too, crossovers as an entrepreneur is positively correlated were more likely to start the business with crossing over. In Uganda, we see that Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors because they inherited it from a family previous exposure to FCS through the first member. job results in women being less likely to cross over. It is, however, not possible based → Exposure to male-dominated sectors on our data to determine if those with this through work experience: Having been exposure are encouraged to cross over a worker in MDS or receiving previous or those who plan to cross over seek this training in the sector was not associated exposure and training in MDS. with being a crossover in Ethiopia. However, Capital, Wealth, and Number of Workers → Household Assets: An assessment of with crossing over and overcrowding or household wealth using a household asset access to water or electricity. However, the index in Ethiopia finds a positive association quality of their dwelling in terms of having between assets and crossing over. The more permanent and sturdy structures (a association between having more assets common measure of wealth and means) is and crossing over is also seen in Lao PDR, associated with being a crossover. As the where those who have sources of drinking discussion above and de Mel, McKenzie, water inside the house in the dry season are and Woodruff (2009) show, the association more likely to cross over to male-dominated between household assets and crossing over sectors. No association is seen between may be driven by the fact that crossovers these assets and crossing over in Indonesia generally have higher profits and are able to or Vietnam. In Peru, no association is seen use those profits to increase their household 27 The Uganda paper does not specify whether the mothers who manage or own an enterprise are in male-dominated or female-concentrated sectors. It is likely that women whose mothers are entrepreneurs continue the mother’s profession and are less likely to cross over. 53 Section 3: Unpacking the Profitarchy assets. On the flip side, households with cash investment were lower for women in greater assets may have more capital female-concentrated sectors such as lace to invest in relatively capital-intensive work and coir compared to women in more businesses in MDS. mixed industries such as bamboo and retail (de Mel, McKenzie, and Woodruff 2009). → Financial capital, investments, and access This gap in return on investment, according to loans: In Botswana, Indonesia, Uganda, to them, may be partly explained by the and Vietnam, crossovers operate with a tendency of women in food processing higher level of capital and production than and garments sectors—typically female noncrossover women. This is in line with concentrated—to invest in equipment that what is found by Rijkers and Costa (2012), benefits both the home and the business. Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors who note that FCS operate at a smaller scale compared to MDS that require more While more capital may certainly be needed capital. Many crossover firms in Ethiopia are by women who are starting businesses in also concentrated in the capital-intensive capital-intensive MDS, the study in Uganda transport and construction sectors, where finds that ‘owning an enterprise in another the capital required to start a business is sector’ was also consistently associated three times that of a business in a female- with being a crossover. The Uganda study concentrated sector. In Guinea, too, also makes another important observation accessing credit was listed as a primary that the difference in capital and inputs constraint faced by crossovers in starting alone does not drive the differences in sales their businesses. Given that MDS are more between female entrepreneurs in MDS vs. likely to be capital intensive, it is likely that those in FCS, as these differences exist even access to financial capital may facilitate after accounting for capital invested. crossing over. Crossovers in Vietnam were more likely to be in households that have → Number of workers: Linked closely to the borrowed money. However, it is not possible need for capital are factors like the size to determine whether access to loans of the firm as assessed by the number of allowed them to cross over into capital- workers it employs. The FoB study finds intensive male-dominated areas or whether that crossover businesses run by women they were more likely to seek out capital in developing countries are more likely to invest once they crossed over. Similarly, to employ more workers compared to in Ethiopia, women in MDS report having noncrossovers. This pattern closely follows more access to finances in an emergency the ‘profitarchy’ that emerges in developing compared to noncrossovers, and in countries, while in developed countries Botswana noncrossovers are less likely than the female-owned firms operating in MDS crossovers to be able to access the bank as do not employ more workers compared a source of emergency funds. to those in FCS. In Ethiopia, Guinea, and Uganda, we also see this positive association The difference in profits may arise from between crossing over and the number how the capital is invested. An experiment of employees.28 However, in Lao PDR and in Sri Lanka that provided cash grants to Vietnam, no significant difference is seen microentrepreneurs found that return on in the number of workers between women- 28 In terms of composition of employees, as seen in table 4, not surprisingly, crossover firms in Indonesia are less likely to employ unpaid labor and in Guinea they are less likely to employ women. 54 Section 3: Unpacking the Profitarchy owned firms in FCS and MDS, indicating that a positive association between crossing over the businesses do not operate at a different and women starting a business because scale in these countries. As noted in the they saw an opportunity and chose to opt previous paragraph, in Uganda, the authors into the business. Similarly, in Indonesia, account for labor, capital, and materials and female crossovers are more likely to also find that crossovers continue to perform have a wage job, suggesting that they may better than noncrossovers. be running the business out of interest than because they do not have other income- → Necessity vs. opportunity entrepreneurs: earning opportunities. The necessity vs. An opportunity entrepreneur is someone opportunity findings align with previous who has started a business because they research by Anna et al. (1999), who find that Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors identified a good business opportunity women operating in traditionally female- or a good business idea. A necessity dominated sectors have higher expectations entrepreneur is someone who started a of work-life balance while those who cross business because no other opportunity over have higher expectations for money for employment presented itself. Data and wealth.29 These findings are also echoed from Mexico and Botswana support the in a study of women who started businesses hypothesis that opportunity tends to in the technology sector in Nigeria, who drive crossing over to MDS as opposed indicated that personal interest was a to necessity. The study from Ethiopia also primary motivator. In contrast, for those attempts to unpack what drives those who who operated in the nontechnology sector, choose to cross over into MDS and whether unemployment was the most common they are necessity- or opportunity-driven reason for starting a business (Aderemi et al. entrepreneurs. Like in Mexico, they too find 2008). Skills and Abilities → Education: The number of years of can be associated with being a crossover education was associated with being an (Brouwers, Van de Vijver, and Van Hemert entrepreneur in male-dominated sectors in 2009). However, no association between Botswana, Guinea, Indonesia, Lao PDR, and education and crossing over to MDS was Mexico. Having less than primary education observed in Peru, Uganda, Vietnam, or the was negatively associated with crossing multicountry FoB study. The association over in Ethiopia. This is in line with previous seen in some studies could result from research that finds years of education the relationship between education and 29 Also, in line with these is the finding from Uganda, where crossovers are more likely to be entrepreneurial compared to noncrossovers. However, interestingly, the research from Peru finds that those who cross over are more likely to be risk averse. 55 Section 3: Unpacking the Profitarchy wealth, or the prestige of male-dominated self-efficacy: ‘verbal persuasion.’ Exposure- businesses, which could attract more enabling factors such as persuasion or educated entrepreneurs in some settings. suggesting to someone that they possess It could also reflect the fact that in some the skills to cross over can positively countries, more educated women may have reinforce self-efficacy. Therefore, it is likely higher cognitive abilities, are able to access that the perception that others believe capital, have greater self-efficacy, or can you to succeed in male-dominated fields overcome societal norms and cross over to drives both self-efficacy and crossing over. male-dominated sectors. In Ethiopia, for For example, Barnir, Watson, and Hutchins example, the association between education (2011) find that role models have a strong and crossing over may be captured through influence on self‐efficacy, which, in turn, Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors indicators of cognitive ability, as this influences entrepreneurial career intention. association disappears when Digit Span Test scores are accounted for. Furthermore, given Needing greater self-efficacy to operate as the positive association between education a crossover is another factor that emerged and wealth, the greater average profits in the qualitative component of the study. observed in the MDS could also be driven Among women working in MDS in Conakry, by these more educated entrepreneurs Guinea, women talked about having to clustering in these sectors. prove themselves as capable. For example, a woman painter said, “You have to work → Self-efficacy: Self-efficacy as per Albert hard to show those people that even if you Bandura (1999) is the belief in one’s own have work at home, you have children, you capability to execute a task, perform a have your husband, you can also work like behavior, or achieve an outcome. A person’s men do.” This suggests that women may self-efficacy is considered to be a function need a high degree of self-efficacy to be of performance accomplishments, vicarious able to persevere in work environments experience, verbal persuasion, interpretation dominated by men where they can face of one’s emotional arousal or physiological regular questions about their competence state that collectively determine these and commitment to the work. beliefs (Bandura 1977). The belief in one’s ability to achieve an outcome in turn drives On the other hand, performance perseverance, which could be an important accomplishments can drive self-efficacy skill needed to cross over. The association and we find that some of the women who between self-efficacy and crossing over is managed to succeed in MDS reported not consistent across settings in this report. increased self-confidence and pride due In Ethiopia, Mexico, and the multicountry to their success. For example, when asked FoB study among developing countries, no what she thought about crossovers, a female association is observed between self-efficacy mechanic said, “I am proud of them and and crossing over. However, a positive myself, I feel proud to be amongst men.” association is seen in developed countries Of the four aspects that drive self-efficacy, in the multicountry FoB study and in some ‘personal accomplishment’ is a key driver regression specifications in Uganda. For of self-efficacious belief and it is likely that example, after accounting for exposure- being in a male-dominated sector that enabling factors in Uganda, self-efficacy is yields greater profits increases a sense of no longer associated with crossing over. This personal accomplishment and reinforces may be explained by a key factor that drives self-efficacy of women in male-dominated 56 Section 3: Unpacking the Profitarchy sectors. A small survey of women engineers → Cognitive ability: No clear or consistent in the United States finds that women who association is seen between crossing over persisted in engineering careers articulated and cognitive ability. Cognitive ability was higher levels of self-efficacy, were more likely measured across the studies in this report to describe themselves as engineers, and primarily in two ways. One was using the were more motivated by the challenges and Ravens Progressive Matrices, an assessment novelty of the profession (Buse, Bilimoria, of analytical or fluid intelligence and abstract and Perelli 2013). This association of success thinking, and another was the Digit Span and self-esteem was also seen by Rauch Test, an assessment of short-term working and Frese (2007), who find in a meta-analysis memory. In Ethiopia and Mexico, a positive that self-efficacy is positively correlated with association is found between higher scores Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors business success. on at least one of the two tests and crossing over to male-dominated sectors.31 However, While self-efficacy was not associated with in Uganda, a small but negative association crossing over in Peru, having a greater was seen between analytical intelligence internal locus of control instead of an and being a female entrepreneur in a male- external locus of control30 is associated with dominated sector. The link between crossing increased probability of crossing over to over and cognitive ability may be a function MDS. At the same time, self-efficacy but not of the nature of the sectors that comprise internal locus of control was associated with FCS and MDS in a country. Not all MDS being a crossover in Uganda. would attract those with higher memory ability or fluid or visual/spatial32 intelligence. → Decision-making: In Guinea, crossovers For example, in Uganda, where FCS are likely to have greater decision-making comprise sectors such as hair dressing and power within the home. However, in tailoring that may require abstract thinking, Botswana, a negative association is seen a negative association between MDS and between a higher score on a household Raven’s Progressive Matrices score is not decision-making index and crossing over. fully surprising. Overall, further research This is possibly due to what this index is needed to establish the link between captures. For example, it is possible that cognitive abilities and operating in MDS, women who cross over, especially at the keeping in mind the nature of the cognitive scale that MDS operate at in Botswana, do skills needed in the male-dominated sectors not partake in daily decision-making around in the country. households’ tasks compared to women who remain in FCS. Those who do participate in household decision-making are the women who are less likely to operate a business in MDS. Prior research has shown that FCS tend to be sectors that allow women the flexibility to manage daily housework and 30 Locus of control refers to self-determination and the extent to which one believes they can control their own outcomes and events in their lives. childcare along with their business (Anna et A greater internal locus of control indicates that they feel greater control al. 1999). On the other hand, when asked over these outcomes and events, while a greater external locus of control refers to the external factors exerting control over an individual’s life. specifically about decision-making in terms 31 In Ethiopia, there is a positive association with Digit Span score but no association with Raven’s Progressive Matrices scores. However, in of their business, the crossovers in Botswana Mexico, the positive association is with Ravens Test score, while there is no association with Digit Span score. report a greater ability to participate in such 32 Colom and Garcia-Lopez (2002) find males outperform women in Ravens Progressive Matrices in Spain and this improved performance of decisions compared to noncrossovers. males can be attributed to items related to spatial and visual processing. 57 Section 3: Unpacking the Profitarchy Box 2 What are some of the challenges that female entrepreneurs face after crossing over? Even after crossing over, operating in male-dominated sectors can Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors be fraught with discrimination and challenges (Haupt and Madikizela 2009; Haupt and Ndimande 2019). Qualitative research with female entrepreneurs that operate in the male-dominated construction sector in South Africa highlights the challenges with having to deal with “men in the industry” (p. 43), in obtaining loans due to a lack of collateral, as well as a low quality of employees (Aneke, Derera, and Bomani 2017). Qualitative data from Guinea and Uganda highlight challenges such as harassment, including sexual proposals and threats to shut down the business. In Guinea, women reported challenges including harassment by colleagues, paternalism, and consistent expectations of failure due to the incompatibility of work in MDS with women’s physical characteristics (e.g., strength and cleanliness) and social roles (e.g., housework and caregiving responsibilities). Women working in workshop environments reported more harassment. In the study sample, these women were usually apprentices or employees rather than business owners. Nonetheless, their stories are illuminating. A female mechanic explained that she has to resist advances from colleagues and has to confirm her commitment to the work repeatedly. She said that others tell her that “people will make you pregnant and you’ll go sit at home.” She counters by saying, “Me, I have decided that I am going to do this, I am going to do it.” In the Guinea study, women working as business owners in sectors requiring higher levels of education articulate different types of experiences of discrimination from clients and suppliers. A pharmacist explained, “Discrimination is always there. People don’t talk about it but it’s there…” Similarly, a graphic designer described potential clients questioning her capabilities. She said, “When you go to an office to describe the services that you offer, there will be people who will underestimate you and say that a girl like you will not be able to do this work.” Women in the study did not report direct social sanctions for working in MDS; instead, they often described receiving encouragement. Moreover, both women and men expressed admiration for women operating in MDS. Rather than stemming from explicit social disapproval, 58 Section 3: Unpacking the Profitarchy challenges to operating in MDS were related to more subtle demands to prove competence and suitability for the work in the face of low expectations. Crossing over is also associated with reduced social capital. Current research indicates that operating in male-dominated sectors results in lower social capital through reduced trust, social network, and community participation for women (and men) compared to those operating in Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors FCS (Sappleton 2009). Gender norms make it more difficult for women to network with men and ‘old boys’ clubs’ can be particularly difficult to penetrate for women who cross over, leaving them with the option of either finding a male business partner or networking with only women in their sectors (Aneke, Derera, and Bomani 2017; Sappleton 2009). Furthermore, even if they network with women, this is not always beneficial, as it is the men who hold powerful positions as the ‘gatekeepers’ of resources in most male-dominated sectors (Sappleton 2009; Aneke, Derera, and Bomani 2017). Reduced social capital can contribute to or amplify other harmful effects of operating in a sector that is dominated by the opposite sex. For example, women in male- dominated occupations are more likely to report depressive symptoms compared to women in female-concentrated economic activities (Tophoven et al. 2015). This persistent discrimination and expectation to fail may be internalized and lead women to second-guess their abilities. This could even explain why we see an inconsistent association between crossing over and self-efficacy. In South Africa, female entrepreneurs already operating in the construction sector reported that they lack existing knowledge and experience of the industry (Aneke, Derera, and Bomani 2017). Female entrepreneurs operating in MDS in Uganda (Campos et al. 2015) similarly felt that they lacked technical and managerial skills. Another study of women who work in primarily male-dominated occupations in South Africa (Martin and Barnard 2013) also highlighted negative work-identity perceptions among women which were characterized by low self- esteem and low self-efficacy. Women report a lack of confidence in their competence and a reluctance to take on roles that are perceived to be more competitive and masculine. While it is possible that there are gaps in skill sets and the technical knowledge of crossovers, it is just as likely that these are the internalized societal expectations. It is important to unpack the extent to which they are real or simply a gender-role-induced skepticism of professional ability that holds women back. 59 Section 3: Unpacking the Profitarchy Context-Specific Factors Given the variety of settings in the review and the diversity in data sets, very few factors are associated with crossing over to MDS across all countries. This may be driven in part by the fact that sampling and what comprises MDS and FCS differs somewhat in each country. The enablers of crossing over to those country-specific male-dominated sectors may also differ. In almost all countries though, women’s education and past exposure to male-dominated sectors through work experience or training, or through their father, spouse, male mentor, or role model appear to be associated with crossing over into Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors MDS. Similarly, among married women, spousal support appears to be key in successfully crossing over. The role of support from family members and mentors in crossing over is not surprising, as women’s choice of sector is potentially driven by beliefs of those important to them. When the traditional social norms of which sector a woman should operate in are questioned by influential others in women’s life, it could support women to also challenge these societal norms and dare to operate in sectors that are not traditionally female friendly. A second and more direct pathway through which these influential male relatives and role models are likely to be encouraging crossing over is by providing their networks, financial support, information, and guidance, and by presenting opportunities for exposure and work experience in the sector. Which Came First, the Chicken or the Egg? While it is likely that the factors identified above encourage women to cross over, it is possible that those who already wish to cross over seek out factors such as role models, work experience, and training in that sector. This is less likely to be the case for education or role models identified at a younger age. However, these studies are correlational and not causal, and could be capturing the approaches female entrepreneurs adopt to deal with the challenges faced after crossing over to MDS. For example, having a male business partner is often a strategy adopted to deal with harassment and difficulty in operating in a male-dominated space (Haupt and Ndimande 2019). Qualitative research identifies other strategies that are adopted by female entrepreneurs to survive in a male-dominated sector, such as partnering with men and established women in the sector, joining business networking organizations, and attending training workshops (Aneke, Derera, and Bomani 2017; Haupt and Ndimande 2019). These women also emphasize the need for strong support systems to help them succeed in male-dominated sectors (Aneke, Derera, and Bomani 2017). Martin and Barnard (2013) in their 60 Section 3: Unpacking the Profitarchy qualitative study find that women seek male mentors to help them deal with the challenges of crossing over and for emotional support. Further research is needed to unpack context-specific enablers of crossing over from survival and success strategies adopted by female entrepreneurs operating in MDS. Whether the factors we identify existed before setting up a business and truly distinguish who chooses to operate in a male-dominated sector, or whether they reflect best practices after crossing over does not necessarily take away from the fact that female entrepreneurs in MDS are more successful. The factors identified above not only make them different from those who operate in FCS but also potentially contribute to their success and should be identified Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors and strengthened. © Jessica Belmont 61 Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors 62 and Policy Conclusions Recommendations SECTION 4 Section 4: Conclusions and Policy Recommendations I n countries where women in male-dominated sectors (MDS) continue to outperform those in female-concentrated sectors (FCS), it is important to pursue policies that enable women to operate in a diverse set of sectors. Based on the studies analyzed in this report and the broader literature on drivers of sectoral Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors segregation, we offer recommendations of successful interventions that key decision-makers can adopt to help female entrepreneurs cross over into male-dominated sectors. As we note throughout this report, several factors ultimately drive sectoral choice and segregation, including (1) support, networks, and exposure to the field; (2) skills and training; and (3) capital/assets and access to loans. Based on these characteristics, we propose a number of interventions that appear promising to support women to cross over into more profitable, male- dominated sectors. These include (1) encouraging spousal support, (2) role models and early exposure to MDS, (3) enhancing women’s education and socioemotional skills, and (4) providing access to capital and loans. Whether the correlates of crossing over—observed in our studies and the literature—cause women to cross over to more profitable male- dominated sectors, or whether these are just factors that are found in successful businesses, is something we cannot unpack with the current evidence we have. Due to the lack of evidence on what works to help women cross over into MDS, the evidence we present below focuses on correlates and interventions that have transferability beyond helping women cross over, in that they are also generally associated with successful entrepreneurship. Additional research is necessary to grow the evidence base on the family of policies that work to support female entrepreneurs to cross over into more profitable, male-dominated sectors, particularly those in science, technology, engineering, and mathematics (STEM). 63 Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Photo by Christina Morillo from Pexels 64 Section 4: Conclusions and Policy Recommendations Section 4: Conclusions and Policy Recommendations 1 Support, Networks, and Exposure to the Field → Role models or mentors from MDS: There is some evidence that having male role models/mentors and fathers in MDS may be helpful to support women to Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors cross over. Our data in Uganda, for example, highlight the importance of male role models and ‘gate openers’ in engaging female entrepreneurs in MDS and breaking traditional norms. Similarly, the WeXchange initiative in Latin America, which connects female entrepreneurs with mentors and investors, found that two out of three STEM entrepreneurs have the support of a mentor, compared to one out of two non-STEM entrepreneurs. Matching female entrepreneurs with male mentors could be particularly helpful. The WeXchange initiative finds that among the STEM entrepreneurs who have mentors, 86 percent have at least one male mentor, compared to 62 percent of non-STEM entrepreneurs (WeXchange and IADB, 2020). While few studies have been developed on the impact of mentorship on crossing over into male-dominated sectors, mentors have been shown to increase business performance of the entrepreneurs they mentor. For example, 82 percent of women who participated in the Going for Growth peer mentoring sessions in Ireland in 2015 reported a 30 percent increase in business turnover, on average, during the 6-month program. Therefore, policy makers and other key stakeholders should design programs matching aspiring or current female entrepreneurs with mentors in male- dominated sectors. Developing role model interventions at the secondary or postsecondary level is particularly crucial as women’s educational choices and their first job set them on a path dependency which is subsequently more difficult to change at a later stage. This is perhaps why we see that early exposure to MDS also contributes to crossing over. For example, the Technovation Girls program enables girls in over 100 countries to work with female mentors to launch technology start-ups aimed at addressing a problem they have identified in their community. After participating in the program, 78 and 70 percent of the 32,000 girls who participated over the past nine years reported more interest in computer science and in entrepreneurship, respectively. Similarly, in a study of factors leading to innovation by women inventors in the US, Bell et al. (2019) find that early exposure to inventors, particularly female inventors, plays an important role in increasing women’s probability of holding a patent at a later age. 65 Section 4: Conclusions and Policy Recommendations While limited evidence and fewer interventions exist on the impact of female role models to support current female entrepreneurs in crossing over, role models in general have been shown to increase business success. For example, including visits by successful alumni in training programs for microentrepreneurs can lead to increased business participation and income, as shown through an experiment in Chile (Lafortune, Riutort, and Tessada 2018). Exposing students to female role models in male-dominated disciplines may also have the potential to increase their likelihood of obtaining degrees in such disciplines and subsequently entering male-dominated sectors.33 For example, in the United States, assigning a female STEM professor to high- ability female students has shown to substantially increase their probability Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors of receiving a STEM master’s degree and of working in a STEM occupation (Mansour et al. 2020). Governments could ensure that high schools and universities have a balanced gender representation among teachers of male- dominated courses. Universities can also encourage their female students to obtain degrees in male-dominated disciplines by organizing sessions with successful female graduates in such disciplines. For example, a study by Lancaster University (2020) of undergraduate women studying introductory economics classes in the US revealed that women were substantially more likely to continue studying the subject if they encountered successful female graduates in the same course. Similarly, an intervention in France showed that a one-hour session with a female role model increases the probability of high-achieving high school female students to enroll in STEM classes the next year (Breda et al. 2020). While implementing role model interventions, potential negative impacts should be minimized. These potential harms include increasing students’ awareness of female underrepresentation in science and reinforcing the belief that women face discrimination in STEM careers, which may further discourage young women from entering those sectors. → Apprenticeships: Another way in which women seem to obtain exposure to male-dominated sectors is through apprenticeships. Our studies indicate that female entrepreneurs working in male-dominated sectors tend to have benefitted from apprenticeships in those sectors. In a qualitative study with women registered in apprenticeship programs in the US, female respondents indicated that they pursued an apprenticeship to break into a new, higher- paying occupation (Reed et al. 2012). Similarly, in Tanzania, Structured Engineers Apprenticeship Program (SEAP) provided female participants with subsistence allowances, additional training, and mentorship opportunities with follow-up after they achieved professional registration. The records show 33 However, the sector of the female role model could matter. As discussed earlier, our study in Mexico showed having a female mentor is negatively associated with crossing over into male-dominated sectors due to the fact that female role models are more likely to operate in female-concentrated sectors. More research is needed to determine when and how female role models could support women to cross over. 66 Section 4: Conclusions and Policy Recommendations that female apprentices with funding and complementary support had a much higher completion rate (86 percent) than those who were self-supported (20 percent). However, there is still a need for more rigorous evidence in developing countries on whether apprenticeships could help women cross over and improve the subsequent labor market outcomes of women in male- dominated sectors. Women are underrepresented in apprenticeship programs in male-dominated sectors. Female recipients of the Women in Apprenticeship and Nontraditional Occupations (WANTO) technical assistance grant in the US indicated that they faced three main barriers in participating construction apprenticeships: lack of Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors information and skills, unrealistic expectations about working in skilled trades, and harassment and exclusion at male-dominated worksites (Reed et al. 2012). Similarly, a study with female participants in Swiss STEM apprenticeships programs cites gender stereotypic beliefs about young women and their professional competencies and several adjustment strategies women adopt to succeed in these programs (Makarova, Aeschlimann, and Herzog 2016). The literature offers outreach strategies that employers and public sector agencies could incorporate to increase women’s participation and retention in apprenticeship programs. These include awareness-raising sessions, adapting language in advertising and descriptions in apprenticeships, and profiling real- life examples of women in these jobs to ensure they avoid gender stereotypes and attract a diverse set of candidates (Schomer and Hammond 2020). Part- time apprenticeships and complimentary childcare interventions could be considered to make apprenticeships more flexible (Schomer and Hammond 2020). Tackling harassment and discriminatory social norms and behaviors could ensure that women successfully complete these programs. We explore some strategies to achieve this in box 3 below. → Support from partner in running the business: Support from partners seems to be a key factor in female entrepreneurs’ business performance. For example, in Sub-Saharan Africa, Wolf and Frese (2018) find that women perform worst when their entrepreneurial efforts are ignored by their husband. Since our findings seem to indicate that female entrepreneurs who succeeded in crossing over had support from their husband, interventions that successfully increase cooperation between spouses could be helpful for crossover entrepreneurs to increase their profits. One option is to engage men at the family and household level to provide direct support to their wives and other female relatives. This support can be in the form of economic empowerment or business support that leverages the skills, knowledge, and networks of male family members. For example, a gender transformation and joint training intervention in Côte d’Ivoire showed that male export crop farmers who filled 67 Section 4: Conclusions and Policy Recommendations out a two-year action plan with their wives shared more agricultural decisions, and enabled women to manage more cash-crop tasks (World Bank 2020). In addition, promoting spousal support by encouraging husbands to participate more in domestic work could also alleviate time constraints that can limit women’s sectoral choice. Gender transformation training for couples can result in increased task sharing and joint decision-making and reduce intimate partner violence perpetration by men (Doyle et al. 2014). Shifting individual and collective gender norms, practices, and beliefs of men and women can positively impact women to cross over to MDS. Another strategy that is emerging to be a potential source of gender transformation has been Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors mass media (Chang et al. 2020). Telenovelas and soap operas can be a source of aspirational transformation in gender roles and couples’ relations. Current studies have found that telenovelas can reduce fertility levels (La Ferrara, Chong, and Duryea 2008) and can impact attitudes toward intimate partner violence (Banerjee, La Ferrara, and Orozco 2019). There is also extensive literature on the impact of edutainment interventions. A randomized controlled trial of an edutainment program promoting entrepreneurship among young adult viewers in a popular channel in Egypt found that the program had an impact on respondents’ gender-related attitudes toward self-employment. In particular, male respondents were less likely to report gender-discriminatory beliefs when exposed to the intervention (Barsoum et al. 2017). Another intervention consists of inviting male family members and husbands to business-related trainings targeting women. This can help them understand what their household can gain from their wives’ businesses and how they can support them (Vu et al. 2015). Including men in these trainings can also help mitigate a sense of exclusion and encourage better communication and shared decision-making between household members. Focusing these mixed-gender trainings on business and economic empowerment rather than on gender discussions may be important, to increase the likelihood of men attending the sessions. The impact of this type of intervention on household cooperation should be evaluated in the future, as rigorous evidence on this topic remains scarce. 68 Section 4: Conclusions and Policy Recommendations 2 Enhancing Skills and Training → Self-efficacy and cognitive ability: Increasing self-efficacy may help facilitate female entrepreneurs’ transition to male-dominated sectors. Policy makers Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors and other key stakeholders could implement programs designed to enhance women’s socioemotional skills and their self-efficacy. Such programs could include personal initiative training: a psychology-based entrepreneurship training aimed at developing participants’ entrepreneurial mindsets. For example, in Togo, the training increased female entrepreneurs’ profits by 40 percent—compared to a 5 percent increase among those who underwent a traditional business-training program—in part as a result of increased self-efficacy (Campos et al. 2018). Findings from a similar training in Ethiopia showed that women participants have higher index levels of self-efficacy, personal initiative, and entrepreneurial locus of control in comparison to those who did not participate in the training, as well as higher profits after the training (Alibhai et al. 2019). A training in Kenya that aimed to enhance personal agency and develop an entrepreneurial mindset among female entrepreneurs in the clean cooking sector also found that those receiving the intervention went on to have greater sales compared to those who did not receive this training (Shankar, Onyura, and Alderman 2015). These findings show that trainings addressing socioemotional skills can lead to increased self- efficacy and high levels of impact on business performance. Increased personal initiative may also increase women’s likelihood to cross over into male- dominated sectors, although this type of intervention has yet to be tested. → Technical training, information, and education: Policy makers and other key stakeholders, including universities, may also provide technical training in male-dominated occupations to encourage university graduates to enter male- dominated sectors. Croke, Goldstein, and Holla (2018) find that an information and communications technology (ICT) training, which gave access to 85 hours of classroom-based training spread across 10 weeks, resulted in university graduates being 26 percent more likely to work in the ICT sector. This suggests the potential for trainings to support women’s employment in male-dominated sectors despite initial lack of sector-relevant skills. Interestingly, the switching was more pronounced among women who initially held deep-seated biases against women’s professionalism. Such job training programs offer a potential opportunity to reduce occupational segregation by shifting norms about the appropriate sectors for men and women to work in, even without explicitly encouraging participants to defy social norms. 69 Section 4: Conclusions and Policy Recommendations Organizing information interventions to highlight the earning potential of male-dominated sectors can also be a powerful incentive for women to attend training in these fields. Female entrepreneurs in traditionally female- concentrated sectors often incorrectly believe they make the same or more than their counterparts in male-dominated sectors. Accurate information can address misperceptions of earnings in traditionally female sectors and help women, especially young women, make more informed decisions when choosing their sector of activity. For example, providing information on real returns to vocational training and presenting persuasive messaging to encourage female participation in traditionally male-dominated fields led to increased take-up of male-dominated trades by women in Kenya (Hicks et al. Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors 2011). Similarly, in the Republic of Congo, a World Bank Gender Innovation Lab study (publication forthcoming) found that providing information on returns per sector through a video at the time of application to the program increased the likelihood that women pick a male-dominated trade by approximately 30 percent when applying to a vocational training, and that women who cross over as a result of the intervention are not more likely to drop out from the training. The study also finds that women who cross over after seeing the video had already overcome some barriers: women were at least three times more likely to cross over when seeing the video if they had higher technical experience, higher technical knowledge, or a male role model. This shows that interventions aimed at helping young women to cross over should target several constraints at once. This could be done through programs that provide information on earning through career guidance in schools or informational sessions accompanying technical skills training programs. Similarly, adolescent girls’ programs should enhance exposure to male- dominated sectors, to avoid path dependence in female-concentrated sectors at a young age. Integrating such programs within trainings that increase adolescent girls’ nonfarm employment, such as a vocational training in Rwanda, which increased the share of girls reporting businesses, wage employment, or internships from 50 percent to 75 percent, could help adolescent girls to obtain their first job in more lucrative sectors (World Bank 2015). 70 Section 4: Conclusions and Policy Recommendations 3 Capital/Assets and Access to Loans → Providing capital: Programs that enable women to have access to greater capital may support entry into male-dominated sectors with higher start-up Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors capital requirements. Facilitating access to credit may also help women stay and thrive in these sectors. This could be achieved through business plan competitions such as those tested by McKenzie (2017), that specifically target women who have proposals for business opportunities in male-dominated sectors. These competitions can take the form of grants or low-interest loans. Other innovative interventions, such as psychometric testing, can also be used to facilitate access to credit among women who want to establish or grow their businesses in male-dominated sectors. In an experiment in Ethiopia, Alibhai et al. (2018), for example, showed that customers who scored higher on psychometric tests provided by banks were seven times more likely to repay their loans compared to lower-performing customers. They conclude that psychometric testing could be a promising avenue to increase access to finance for female entrepreneurs, who otherwise tend to lack the collateral needed to obtain bank loans. Furthermore, insurance products that are specifically tailored to any disruptions likely to affect female crossovers are already being designed (IFC 2019). Addressing potential gender discrimination in access to formal finance could improve access to credit. For example, an experiment by Alibhai et al. (2019) showed that in Turkey, 35 percent of loan officers studied in their analysis are biased against female applicants, with women receiving loan amounts $14,000 lower on average compared with men. They find that experience in the banking sector can attenuate this bias, with each year of experience reducing gender-biased loan allocations by 6 percent. This suggests that newly recruited and less experienced loan officers may use gender bias as a heuristic device given limited information, and training could improve their ability to discern loan application quality. It is, however, important to mention that other studies, such as Bardasi, Sabarwal, and Terrell (2011), show no evidence of discrimination in access to formal finance in the three regions they analyze— Europe and Central Asia (ECA), Latin America and the Caribbean, and Sub- Saharan Africa (SSA)—although in ECA women are less likely than men to seek formal finance. 71 Section 4: Conclusions and Policy Recommendations Box 3 Ending discrimination and harassment against female crossovers Harassment and discrimination are cited as key constraints for female entrepreneurs in male-dominated sectors in our studies. While there is still little evidence on the policies that work to end discrimination and Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors harassment, certain steps could be taken to shift the norms around women’s participation in MDS, including launching communication campaigns (through the radio, TV, or newspaper) and showcasing women in male-dominated professions. Anti-harassment awareness campaigns may also be another approach that can be adopted. For example, Guizzo and Cadinu (2020) found that a web campaign against female objectification led to lower gender-harassing behavior, lower hostile sexism, and lower sexual coercion intention by the men who were shown the sensitizing video, in comparison to those in the control group. In addition, establishing public helplines to offer support and resources for targets of harassment and discrimination may be a helpful avenue. Finally, expanding networks of support by connecting female entrepreneurs in male-dominated sectors to one another may help them to build strategies to operate in these sectors. For example, qualitative research shows that women’s social networks among firms in Sri Lanka helped to gain and give emotional and informational support through dialogue and information sharing related to sexual harassment (Adikaram 2017). Although more research is needed on the topic, this points to social support as a potential means of dealing with discrimination and sexual harassment. 72 Section 4: Conclusions and Policy Recommendations Strengths and Limitations of This Report The primary limitation of the research At the same time, the strength of the studies synthesized in this report is that it is in this report lies in overcoming this lack correlational, cross-sectional, and primarily of existing data on gender-based sectoral descriptive in nature. While it speaks to segregation and profits by using creative characteristics associated with crossing sources of data, including data from existing over to MDS, it also potentially captures the impact evaluations and a multicountry characteristics of successful business owners survey through a social media platform. who are surviving in MDS. It is also possible This report and the studies that it comprises Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors that the surviving female entrepreneurs are among the first to provide a global captured in MDS are profitable because of overview of understanding the importance of characteristics such as higher education, self- female microentrepreneurs crossing over to esteem, ability to find mentors, inheriting a male-dominated sectors. This research has business, ability to find a business partner, or established that women operating in male- having a supportive spouse, and not because dominated sectors have greater profits than of the sector choice alone. Another limitation women operating in female-concentrated is that in many countries, national data sets sectors in all countries studied, with the were not available, and the sample is not exception of Cambodia. In addition, the report representative of all microentrepreneurs. As a highlights key correlates of crossing over, result of data-related limitations, this report and such as the role of a mentor or role models, the studies we draw from only assess individual supportive spouses, and prior experience and household-level factors associated with in a field, which can form the basis of future crossing over and do not explore broader research and policy action that can benefit sociocultural factors. female entrepreneurs. Agenda for Future Research and Policy This report deepens our understanding sectors. Moreover, because sectoral choice is of sectoral segregation by systematically not the only factor influencing the productivity examining what constitutes FCS and MDS in of women’s businesses in most contexts, various countries and regions, the performance identifying which constraints are most binding differences of women operating in MDS and in specific countries is needed. To enable FCS, and factors associated with crossing over. this type of detailed analysis, more nationally The differences in sectoral classifications across representative data across entrepreneurs and countries and regions highlight the need for firms are needed. further research to develop rigorous analyses and policy recommendations on female In addition, further research is needed to crossovers at the country level. Even within enhance our understanding of the correlates sectors, a more uniform and granular approach of crossing over in different contexts, notably in sector classifications will be helpful to unpack by including a mixed methods approach to exactly what constitutes MDS and FCS and better identify the barriers and facilitators to to understand continued segregation within crossing over. While these should continue to 73 Section 4: Conclusions and Policy Recommendations be explored at an individual level, as has been done in existing studies, there is also a need to explore the facilitators of crossing over at the household and community level. Longitudinal research may help with distinguishing what helps women to start operating businesses in MDS from what helps women survive and succeed in male-dominated sectors once they enter. In countries where crossing over leads to higher profits for female entrepreneurs, some of the promising policies outlined above Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors could be tested, as large evidence gaps remain. For example, while there is promising evidence on the importance of role models, more research could be undertaken to identify at what stage and in what ways role models are most helpful. Similarly, while there is some evidence that many of the policies suggested in this report can successfully increase female entrepreneurs’ profits, there is still little evidence on the impact of apprenticeship programs, spousal support training programs, or socioemotional skills training in encouraging crossing over. Conducting pilot projects based on the interventions outlined in this report and assessing their effectiveness would widen the evidence base on what works to support female entrepreneurs’ entry and success in more profitable, male-dominated industries. Finally, a major gap exists in the literature on ensuring women’s safety from sexual harassment and discrimination after crossing over into male-dominated sectors. Women’s safety should be a core focus in designing programs intended to support women to cross over. It is important to ensure that programs involving male role models and mentors do not contribute to women’s increased experience of discrimination or harassment. © Lakshman Nadaraja / World Bank 74 75 Section 4: Conclusions and Policy Recommendations Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors 76 APPENDIXES Appendix A: Description of the Sample Table A.1 Summary of data sources used to determine the sector classification, the profitarchy, and correlates of crossing over to MDS in each study Country/study Sample size Nationally Selected % of micro- Profit measurement representative sampling of entrepreneurs (survey question) household/firm enterprises/ in the sample survey entrepreneurs Cambodia Full survey: 4,274 Cambodia N/A N/A Net profit = Revenue – enterprises Socioeconomic Cost Survey (CSES) Sector classification, 2014 Where: profitarchy, and crossover correlates: 1. Revenue: How much Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors 4,118 did your household receive under the different items listed for activity 1, during the past 12 months, that is, since [month] last year? 2. Cost: How much did you spend on the different items listed for activity 1, during the past 12 months, that is, since [month] last year? Indonesia Full survey: 14,019 Indonesian Family N/A 95.2% What is the approximate enterprises Life Survey (IFLS) amount in rupiah of net 2000, 2007, 2014 profit generated by the Sector classification, business during the past 12 profitarchy, and months? crossover correlates: 10,571 Lao PDR Full survey: 2,015 Lao PDR N/A 88.7% In an average sales month, enterprises Expenditure and what is your level of sales Consumption per month? Sector classification, Survey (LECS V) profitarchy, and 2012 crossover correlates: 1,989 Vietnam Full survey: 2,406 N/A Vietnam Access 97.4% Net profit = Revenue – enterprises to Resources Cost Household Profitarchy and Survey (VARHS) Where: crossover correlates: 2008, 2010, 2012, 2,364 and 2014 was 1. Revenue: For the carried out in 12 months the business was provinces in rural under operation for the areas past 12 months, what was the total revenue of this activity? 2. Cost: During the past 12 months, how much was spent on raw materials and small nondurable tools? - During the past 12 months, how much was the labor cost, utilities, rent of land, workshops, transportation, loan interest, taxes and fees, water and solid sewage disposal, and other expenditures? 77 Appendix A: Description of the Sample Table A.1 (continued) Country/study Sample size Nationally Selected % of micro- Profit measurement representative sampling of entrepreneurs (survey question) household/firm enterprises/ in the sample survey entrepreneurs Mexico Full survey: ENAMIN (Mexican IE data was used 98.63% in the N/A 27,582 National Survey of to determine IE survey Microenterprises) correlates of Sector classification 2014 was used crossing over to 100% in the and profitarchy: to determine the MDS ENAMIN 27,540 classification of survey Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors the sectors and Crossover correlates: the profitarchy 3,907 Peru Sector classification: Mexico’s ENAMIN Survey of 100%* N/A 27,582 2014 was used women clients to determine the of the Program Profitarchy and classification of Palabra de Mujer crossover correlates: the sectors – PDM (Word 1,148 of a Woman) of Financiera Confianza were used for the profitarchy and to determine the correlates of crossing over Guinea 540 entrepreneurs N/A World Bank 94.62% What was the total profits Enterprise survey the business earned during and qualitative the past month (30 days) interviews 2014; after paying all expenses, Guinea crossover including salaries, rents, needs assessment materials, etc.? Expenses survey 2014 include payments to business owners if these were paid as a salary. That is, what were the profits of your business during the past month (30 days)? Botswana Full survey: N/A In-person survey Approximately What was the total income of 797 firms 80% the business earned during 797 firms in Gaborone, the past month after paying randomly all expenses, including Sample size for sampled from salaries, rents, materials, analysis: 637 the Botswana etc.? Expenses include Business Registry payments to business owners if these were paid as a salary. That is, what were the profits of your business during the past month? Ethiopia Sector classification: N/A Baseline survey 95% What was the total profit 2,369 of Women the business earned in the Entrepreneurship last 30 working days? Profitarchy and Development crossover correlates: Project 790 78 Appendix A: Description of the Sample Table A.1 (continued) Country/study Sample size Nationally Selected % of micro- Profit measurement representative sampling of entrepreneurs (survey question) household/firm enterprises/ in the sample survey entrepreneurs Uganda Sector classification: N/A Quantitative 100% What is the business NET 187 data from 2011 PROFIT per day? That is, sampling of 735 what was the total income Crossover correlates: entrepreneurs, the business earned 735 most of whom each DAY after paying belonged to all expenses, including Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors the Katwe Small salaries, rents, materials, Scale Industry etc.? Expenses include Association the payments to business (KASSIDA). owners if these were paid In addition, a as a regular salary. quantitative and qualitative survey was administered in 2012 to 63 crossovers and to 120 women working in FCS. Global: Future of Full survey: 55,932 N/A Survey offered 65.9%** To the best of your Business by Facebook to knowledge, in a typical Profitarchy: 17,351 administrators month, what are the profits of Facebook- to the owner(s) of this Crossover correlates: designated SME business? (How much?) 9,827 pages. NOTE: FCS = female-concentrated sectors; IE = impact evaluation; MDS = male-dominated sectors; N/A = not applicable; PDM = Palabra de Mujer; SME = subject-matter expert. * 1,905 female PDM clients were surveyed, of which 1,148 were microentrepreneurs. ** The total number of observations for which number of employees is available is 31,756. Out of these, 20,915 observations have 6 or fewer employees. 79 Appendix A: Description of the Sample Table A.2 Characteristics of the firms in the survey Study/Country Firm characteristics Global: Future of Business Cambodia Botswana Indonesia Lao PDR Vietnam Ethiopia Uganda Mexico Guinea (mean) Peru Women 9.54 8.59 6.3 14 - - 9.196 9.4 - 6.827 - MDS Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Age of Women 8.75 9.48 5.89 9 - - 8.78 9.912 - 6.112 - business FCS FCS – MDS – MDS MDS – MDS – MDS – Diff. MDS = - FCS = – FCS - - FCS = FCS = - FCS = - -0.794 0.41 = 5*** 0.415 -0.512 0.715 Women Number OLS 4.94 1.07 4.34 3.9 1.28 0.549 0.467 1.371 MDS of regression: employees OLS 1. without Women = 0.276*** regression region FE Number of 1.07 0.93 1.95 2.4 0.73 0.202 0.205 0.28 FCS and paid = 26.979* employees Number workers = and 2. MDS of paid 0.082*** with FCS – MDS – MDS – MDS – MDS – – FCS P-value employees region FE Diff. MDS = - FCS = FCS = FCS = FCS = = = 0.000 = 0.290*** = 24.798 -3.87*** 2.38*** 0.347*** 0.261* 1.091*** 1.5*** Women 1.41 - - - - - 1 - - - - MDS Number of Women business 1.15 - - - - - 1 - - - - FCS owners FCS – MDS – Diff. MDS = - - - - - - - - - FCS = 0 -0.262 Women - - - 44% - - - - - - - MDS What percent of Women firms have - - - 46% - - - - - - - FCS a business partner? MDS Diff. - - - – FCS - - - - - - - = -2% Women - - - - - - MDS What All firms OLS OLS OLS OLS percent of Women are in regression regression regression regression firms are in - - - - - - FCS urban rural = rural = rural = urban = urban/rural locations 0.003 -0.153 -0.021** 0.024* areas? Diff. - - - - - - NOTE: FCS = female-concentrated sectors; FE = Fixed Effects; MDS = male-dominated sectors; OLS = ordinary least squares; - = no data. *= p < .10; **= p < .05; ***= p < .01. 80 Appendix B: Sector Classification Table B.1 Methods used to determine whether a sector is MDS or FCS Country/study MDS if at least 70% of MDS if based on the MDS if at least 75% of MDS if based on the respondents answered survey data at least 70% respondents answered survey data, at least that most enterprises of enterprises in that that most enterprises 75% of enterprises in in their business sector sector were owned by in their business sector that sector are owned by are owned by men, men, otherwise FCS are owned by men, men, otherwise FCS otherwise FCS otherwise FCS Cambodia X Indonesia X Lao PDR X Vietnam X Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Mexico X Peru X Guinea X Botswana X Ethiopia X Uganda X Future of Business* X NOTE: FCS = female-concentrated sectors; MDS = male-dominated sectors. * Female respondents only. Table B.2 Comparing our findings on sectoral segregation to previous research Study Male dominated Female concentrated Gender Innovation Lab (Botswana, Construction, wood manufacturing Agriculture, forestry, and fishing Ethiopia, Uganda) and repair, electricity and gas (only in Guinea), ICT (only in supply, automobile maintenance Ethiopia), leather manufacturing and sale, transportation and (only in Botswana), tourism services storage, small transport services, (only in Botswana), computer trade food beverages and tobacco programming (only in Ethiopia), retail, metal works and engineering, trade of textiles and footwear (only water supply and waste in Botswana) management, tourism services (only in Ethiopia), trade of textiles and footwear (only in Uganda), real estate activities Bardasi, Sabarwal, and Terrell Other services, electronics, Wholesale and retail trade, Sub-Saharan Africa (2011): 22 countries IT, nonmetals, metals, other garments and leather goods, manufacturing, construction and hotels and restaurants, food, other transportation, chemicals services, textiles Hallward-Driemeier (2013): 41 Metals, construction, chemicals, Food processing, garments countries machinery Rijkers and Costa (2012): Ethiopia No data Manufacturing (including food and beverages, brewing/distilling), manufacturing (excluding grain milling, food and beverages, distilling, wearing apparel), grain milling 81 Appendix B: Sector Classification Table B.2 (continued) Study Male dominated Female concentrated Gender Innovation Lab study Automobile maintenance and sale, Textile manufacturing (Mexico), (Mexico and Peru) agriculture, forestry and fishing, clothes retail (Mexico), hairdressing, trade food, beverages and tobacco etc. (Mexico), waitering (Mexico) retail, transportation and storage (Mexico only), construction (Mexico only), wood manufacturing and repair (Mexico only), metal works Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors and engineering (Mexico only), Latin America and small transport services (Mexico the Caribbean only), water supply and waste management (Mexico only) Bardasi, Sabarwal, and Terrell IT, construction and transportation, Garments and leather goods, retail (2011): 13 countries other services, chemicals, other trade, machinery and equipment manufacturing, nonmetals production, food Gender Innovation Lab (Cambodia, Agriculture, forestry, and Agriculture, forestry, and fishing Indonesia, Lao PDR, Vietnam) fishing (except in Cambodia), (only in Cambodia), construction construction (except in Vietnam), (only in Vietnam), wood electricity and gas supply, other manufacturing and repair, trade manufacturing and repair, food, beverages and tobacco automobile maintenance and sale, retail (only in Indonesia), tourism transportation and storage, human services, trade of textiles and health and social work footwear, trade of pharmaceutical products, accounting and auditing, manufacturing furniture Rijkers and Costa (2012): Indonesia Manufacturing (including mining Wholesale and retail trade and excavation; manufacturing, Asia including processing of agricultural products; electricity; gas and water; construction) Rijkers and Costa (2012): Trade (including wholesale and Manufacturing (including mining Bangladesh retail trade, excluding repair of and quarrying, manufacturing motor vehicles, motorcycles, electricity, gas and water supply, personal and household goods) construction) de Mel, McKenzie, and Woodruff Repair services Making lace (2009): Sri Lanka Even numbers of men and women in retail trade, bamboo, and food sales Anna et al. (1999): United States High technology, construction, and Retail and service industries United States of manufacturing America Bardasi, Sabarwal, and Terrell IT, chemicals, others, electronics, Wholesale and retail, other services, (2011): 26 countries nonmetals, metals, other garments and leather, textiles Europe and manufacturing, construction and Central Asia transport, food NOTE: ICT = information and communications technology; IT = information technology. 82 Appendix C: Summary of the Data Used to Determine the Profitarchy Table C.1 Average monthly/past month profits in each study Women in MDS Men in MDS Women in FCS Men in FCS Percent change for women from FCS to MDS* Country LCU USD LCU USD LCU USD LCU USD Botswana 55,189 5,411 85,731 8,405 25,195 2,470 - - 119% Guinea 3,851,639 549 3,538,036 504 2,024,608 289 - - 90% Ethiopia 10,678 545 - - 5,932 303 - - 80% Uganda 494,000 197 515,000 206 205,000 82 - - 141% Mexico 3,826 288 6,222 468 2,542 191 5,474 412 51% Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Peru - 330 - - - 194 - - 70% Indonesia 443,333 37 483,667 41 265,667 22 522,750 44 67% Vietnam 1,253,167 59 2,180,500 103 950,417 45 1,739,667 82 32% Cambodia 994,500 246 891,500 221 1,220,667 302 1,620,333 401 -19% Lao PDR** 8,557,000 1,069 5,438,000 679 4,768,000 596 5,773,000 721 79% NOTE: USD average exchange rates for the year of the survey were used, using https://data.worldbank.org/indicator/PA.NUS.FCRF. FCS = female-concentrated sectors; LCU = local currency unit; MDS = male-dominated sectors; - = no data. * = calculated using the formula (women in MDS – women in FCS/women in FCS) * 100. ** = uses monthly sales instead of profit. Table C.2 Robustness of the profit gap between women in MDS compared to those in FCS with and without accounting for individual, household, and business characteristics Study/country Profit variable Controls Results No controls Crossovers have higher profit (statistically significant at 1%) Controlling for individual characteristics: married, completed Results hold at 1% more than secondary education; any of last 5 jobs was as business significant level for the log owner; any of last 5 jobs was in a male-dominated sector; role transformation while they model was male; mother completed more than primary education; are no longer significant mother was owner/manager of a firm when respondent was a child; for the IHS transformation father completed more than primary education; father was owner/ manager of a firm in a male-dominated sector when respondent was a child; foreign born; proportion of male siblings; firstborn in the family; number of siblings The IHS and log transformation of Controlling for business characteristics: received previous training No significant gap in Botswana profit during the in sector of current business; knows more than 30 owners in any profits for either the log past month sector; business formally registered; business has written business or IHS transformation of plan; business has a written annual budget; business records profits revenues and expenses; business is part of any business association; share of male employees Controlling for household characteristics: Household Decision- Results hold for both the Making Index (High: 5, Low: 1); spouse completed more than log (at 1% significance primary education; spouse is owner/manager of other firm; HH level) and IHS (at 5% asset index 0–14; number of children significance level) transformation of profit Controlling for all the above characteristics No statistically significant gap in profits in either the log or IHS transformations 83 Appendix C: Summary of the Data Used to Determine the Profitarchy Table C.2 (continued) Study/country Profit variable Controls Results No controls Crossovers have higher profit and the difference is significant at 1% Controlling for individual (household) characteristics: age and Crossovers are still more age squared; education; age and age squared started working; profitable, but the gap marital status; number of children; household size; region of birth; is no longer statistically mother was female guardian when young; female guardian had significant once individual some schooling; female guardian was an entrepreneur; female characteristics are guardian had wage employment; father was male guardian when accounted for Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors young; male guardian had some school; male guardian was an Past month’s entrepreneur; male guardian had wage employment Guinea profit winsorized at the top 1% Controlling for business characteristics: age and age squared of Crossovers have higher business; only owns business but does not manage; starting capital profit and the difference is obtained from loan by family/friends; starting capital obtained significant at 1% from gift by family/friends; starting capital obtained from savings; number of full-time employees that are members of household; most business activities take place at home; ever involved in day-to- day production or service delivery; pays self a regular salary Controlling for individual (household) + business characteristics: Crossovers are still more all the above variables profitable, but the gap is no longer statistically significant No controls Crossovers have higher profit and the difference is Last 30 days’ significant at 1% Ethiopia profit Without age of the business Crossovers perform better and the difference is still significant at 1% No controls Crossovers have higher Log sales and this is statistically transformation of significant at 1% Uganda previous month’s Controlling for capital, labor, and material inputs Crossovers have higher sales (revenue) sales (statistically significant at 10%) No controls Crossovers have higher profit and the difference is significant at 1% Controlling for business characteristics: retail; services; hours Still higher profit for worked per week; age of business; firm registry with local authority; crossovers (all are total workers significant at 1% except for the IHS transformation, which is significant at 5%) Controlling for household characteristics: married; has children; Crossovers have high level of education; household size; Poverty Dummy Index; no significantly (at 1%) higher Profit per month access to toilet; indirect water access profit (in all three data winsorized at the transformations) top 1% as well as Mexico the log and IHS Controlling for individual characteristics: age (in years); age Crossovers have transformation of squared; father's education level; mother’s education level; access significantly (at 1%) higher profit to finance; FE education level; cognitive index span test; cognitive profit (in all the three data index Raven transformations) Controlling for all the individual, household, and business Including all controls characteristics together does not change the significance level for the level (still significant at 1%). However, it is significant only at 10% for the log transformation, while it is no longer significant for the IHS transformation of profit. 84 Appendix C: Summary of the Data Used to Determine the Profitarchy Table C.2 (continued) Study/country Profit variable Controls Results No controls Crossovers have higher profits and the difference is significant at 1% Controlling for business characteristics: number of employees; Adjusting for business number of paid employees characteristics reduces the magnitude of the profit gap and it remains significant at 10% Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Controlling for household characteristics: water from public Adjusting for household network inside the dwelling; drainage from public network, inside characteristics increases the dwelling; electricity; walls of brick or concrete block; ceiling of the magnitude of the reinforced concrete; overcrowding in the household; household size profit gap and it remains Monthly profit significant at 1% Peru winsorized at the top 1% Controlling for individual characteristics: age (in years); language; Adjusting for individual marital status; years of schooling; number of children; woman has a characteristics reduces dependent job; women's independent labor weekly hours; partner the magnitude of the manages a business; partner's income; instrumental risk taking; profit gap and it remains brief resilient coping; self-control; brief resilience and survival; significant at 5% abbreviated self-leadership; locus of control; self-esteem; big 5-11 Controlling for individual + household + business characteristics Adding all controls together reduces the magnitude of the profit gap and it remains significant at 10% No controls Crossovers have higher profits and the difference is significant at 1% Controlling for individual characteristics: married; schooling; age; work and self-employed; father highly educated; mother highly educated; credit other sources; father works Controlling for individual + household characteristics: rural; kids 0 to 2; kids 3 to 5; elderly; cooking source; need help Log transformation of profit during the Controlling for individual + household + business characteristics: Indonesia past 12 months business outside; business age; unpaid workers Both adding each group of after winsorizing control variables at a time it at the top and and adding all together bottom 1% increases the magnitude of the profit gap and it Controlling for individual + household + business + community remains significant at 1% characteristics: nonmotor; market Controlling for individual + household + business + community + network characteristics: artisans; voluntary labor 85 Appendix C: Summary of the Data Used to Determine the Profitarchy Table C.2 (continued) Study/country Profit variable Controls Results No controls The magnitude of Log Controlling for individual characteristics: age; age squared; the profit gap in (log transformation married; schooling; work for wage; member of a political party transformed) profits is no of calculated longer significant once profit during the Vietnam region is accounted for. past 12 months Controlling for individual + household characteristics: rural; elderly; However, the gap remains after winsorizing kids 0 to 2; kids 3 to 5; water source; cooking source; housework; even when any or all Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors it at the top and borrowed money; HH savings the control variables are bottom 1% added to the model Controlling for individual + household + business characteristics: business in home; business age; unpaid workers No controls Crossovers have lower profit and the difference is significant at 5% Log Controlling for individual characteristics: married; schooling; age; The above result is still transformation age squared; work for wage; household head; spouse of household significant at 5% of calculated net head; violence victim profit during the Cambodia past 12 months Controlling for individual + household characteristics: rural; kids The above result is still after winsorizing 0 to 2; kids 3 to 5; elderly; cooking source; drinking water inside; significant at 5% it at the top and household members; men in HH bottom 1% Controlling for individual + household + business characteristics: The above result is still female helpers; male helpers; paid workers significant at 1% No controls Crossovers have higher profits and the difference is significant at 10% Log Controlling for individual characteristics: married; schooling; age; transformation age squared; work for wage; household head; spouse of HH of average sales The above result is no Lao PDR per month after Controlling for individual + household characteristics: urban; kids longer significant once winsorizing it 0 to 2; kids 3 to 5; elderly; own house; good walls; kitchen inside; control variables related to at the top and public electricity; household size individual characteristics bottom 1% are added to the regression model Controlling for individual + household + business characteristics: business age; fixed location business; HH business members; more than 1 business Controlling only for regional FE Crossovers have higher profits and the difference is significant only at 1% Controlling for individual characteristics: age of the respondent; The results still remain university or college; more than secondary, vocational training, or significant at 1% The IHS apprenticeship; secondary; primary; do you have a spouse or long- Global: Future transformation term partner of Business profit in a typical month Controlling for business characteristics: more than one owner; The results still remain single owner; number of employees significant at 1% Controlling for both individual and business characteristics The results still remain significant at 1% Note: FCS = female-concentrated sectors; FE = fixed effects; HH = Households; IHS = inverse hyperbolic sine; MDS = male-dominated sectors. 86 Appendix D: Details on Each Individual Study Botswana Our sample Number of enterprises: 637 informal and formal microenterprises in the urban context of Gaborone How and where were the data collected, and what are they representative of? Botswana Women Entrepreneurship Study is based on an in-person survey of 797 firms in Gaborone, randomly sampled from the Botswana Business Registry across all business sectors in our sample. Categories of business owners: Three categories of business owners: (1) women in male- dominated sectors (151); (2) women in female-concentrated sectors34 (262); and (3) men in male- dominated sectors (224). Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Identifying What methodology was used to define sectors as male dominated or female concentrated? As male-dominated the Botswana Business Registry does not include the gender of the business owner, we classify and female- sector gender dominance using a survey of a sample of business owners: the sector was classified concentrated as male dominated if at least 75 percent of the respondents answered that most enterprises sectors in their business sector are owned by men, otherwise we classified the sector as female concentrated. How many sectors are female concentrated? How many are male dominated? Out of 28 sectors, only 9 are classified as female concentrated. Our final gender classification of sectors is consistent with the gender-ownership data in the World Bank’s Enterprise Survey for Botswana. The profitarchy Women operating in male-dominated sectors have monthly profits (116,522 pula) that are statistically similar to those of men in male-dominated sectors (111,178 pula) and significantly higher than those of women operating in female-concentrated sectors (25,195 pula). However, the profits of women operating in male-dominated sectors appear to be partly driven by high- performing outliers, including businesses that are jointly owned with a husband and those that have a foreign-born owner. After winsorizing the profits to account for these high-performing outliers, men in male- dominated sectors have the highest profits (85,731 pula), followed by women in male-dominated sectors (55,189 pula), and then women in female-concentrated sectors (25,195 pula). Factors associated • Exposure to the sector through previous work with being a • Being a supplier to the sector, through friends and family or by receiving information about its female crossover potential • Completing more than secondary education • Having a mother who completed more than primary education When we exclude businesses with foreign-born owners, the level of education of the owner and of the owner’s mother appear to be the most important correlates of crossing over: it might be that Batswana women who received an education and come from a family where the mother has a high level of education are more informed and exposed to promising business opportunities. Constraints • Access to credit: Women in male-dominated sectors, women in female-concentrated sectors, to business and men in male-dominated sectors all identified access to credit as a constraint. performance faced • Location: Women in male-dominated sectors and women in female-concentrated sectors were by women who more likely than men to list finding a location as a constraint. cross over • Building networks and discrimination from clients and employees: Despite their success, women operating in male-dominated sectors face specific challenges that may have their roots in social biases against women operating in a traditionally male world, such as problems building networks and discrimination from clients and employees. Given these • Encourage women to cross over: Exposing women to more profitable male-dominated sectors constraints, what through training, apprenticeship, and mentoring programs could help more women cross over. policies could • Social norms and spousal support: Policies that address discriminatory social norms and that help women to sensitize husbands on the valuable role they can play in supporting their wives’ success in these cross over into sectors should be encouraged. While we do not find evidence that husbands help women to more profitable, cross over, spouses do appear to provide important support that could help women to succeed male-dominated in a given sector, such as by providing skills (either through their own labor or by imparting sectors? these skills on the wife), access to larger amounts of finance/capital, or by helping with business registration or the acquisition of a license. 34 In our study, female-concentrated sectors include any sector that is not male dominated. 87 Appendix D: Details on Each Individual Study Cambodia Our sample Number of enterprises: 4,274 enterprises How and where were the data collected, and what are they representative of? The data come from the Cambodia Socioeconomic Survey (CSES) for the year 2014. The survey is a nationally representative household survey, and information on the businesses comes from the household business module. Categories of business owners: Four categories of business owners: (1) women in male- dominated sectors (59); (2) women in female-concentrated sectors (2,238); (3) men in male- dominated sectors (935); and men in female-concentrated sectors (1,042). Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Identifying What methodology was used to define sectors as male dominated or female concentrated? male-dominated Because the data are representative, we define a sector as male dominated if at least 70% of the and female- owners of businesses in the sector are men. concentrated How many sectors are female concentrated? How many are male dominated? Out of 19 sectors, sectors 11 are classified as female concentrated and 8 as male dominated. The profitarchy How do profits of women in MDS compare to profits of men in MDS, and of women in FCS? In Cambodia, male-dominated sectors are less profitable than female-concentrated sectors. Women crossovers have past year profits (11,934,000 KHR) that are lower than women operating in FCS (14,649,000 KHR), and men in male-dominated sectors have past year profits (10,698,000 KHR) that are lower than women operating in female-concentrated sectors. However, within MDS and FCS, men have higher profits than women: for example, men operating in female-concentrated sectors have past year profits (19,444,000 KHR) that are higher than women operating in those sectors. Factors associated • Living in an urban area with being a • Having fewer female helpers in the business female crossover • Having paid workers in the firm Constraints • Women in male-dominated sectors have less education than men in those sectors. to business • Women crossovers are more likely than men in these sectors to have children under 2 in their performance faced household and more likely to have elderly household members, which may lead to challenges by women who related to domestic work. They are also less likely to be married than men in these sectors, cross over which means they may face additional challenges without the support of a partner. Given these In Cambodia, female microentrepreneurs in male-dominated sectors make less profit than constraints, what women in female-concentrated sectors. Therefore, it is not recommended to design policies in policies could Cambodia to support a switch into male-dominated sectors. help women to cross over into more profitable, male-dominated sectors? 88 Appendix D: Details on Each Individual Study Ethiopia Our sample Number of enterprises: 790 female microentrepreneurs across six cities in Ethiopia: Adama, Addis Ababa, Bahir Dar, Dire Dawa, Hawassa, and Mekelle. How and where were the data collected, and what are they representative of? The sample came from women who were registered with the Women Entrepreneurship Development Project (WEDP). Baseline enterprise survey data were collected from a sample of 2,369 female WEDP entrepreneurs, and an additional survey module was administered to a subset of 800 female WEDP entrepreneurs that included more questions about sector choice. Categories of business owners: Two categories of business owners: (1) women in male- dominated sectors (164); and (2) women in female-concentrated sectors (626). Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Identifying What methodology was used to define sectors as male dominated or female concentrated? male-dominated We drew on a question administered in the WEDP baseline survey that asked the entrepreneur, and female- “Are most enterprises in your business sector owned by men or women?” If more than 75 concentrated percent of responses from the entire baseline sample (2,369 female entrepreneurs) were that sectors men owned most enterprises in their business sector, we defined that sector as male dominated. We used the entire core baseline sample of 2,369 entrepreneurs to determine this definition of a male-dominated sector and defined a woman who had a business in any of these sectors as a crossover. How many sectors are female concentrated? How many are male dominated? Out of 24 sectors, 11 are classified as male dominated and 13 are classified as female concentrated. The profitarchy Women operating in male-dominated sectors (crossovers) have, on average, double the profits of women in female-concentrated sectors (noncrossovers). Monthly profits of women in MDS were (10,678 Ethiopian birr), which is statistically significantly higher than those of women operating in FCS (5,932 Ethiopian birr). We also observe that women in male-dominated sectors are able to create larger firms in terms of number of employees and capital levels. Factors associated • Opportunity entrepreneurs: started a business venture because of a market opportunity with being a • Having a husband who is in business himself strongly predicting crossing over female crossover • Education and skills do not seem to predict crossing over • Digit span recall and having a male role model growing up important for the top performing crossovers • Good support networks, or assistance of a husband or a male role model Constraints • Women crossovers are significantly more likely to face difficulty in building networks in their to business sector of operation and do not seem to differentially benefit from networks of women in the performance faced same industry, with crossovers reporting that they are more likely to feel despised by other by women who women business owners. cross over • Crossovers are more likely to face clients who prefer to do business with male business owners and more likely to face problems with male employees. • Harassment outcomes are similar for crossover and noncrossover entrepreneurs, with as many as 11 percent of the women reporting being sexually harassed within the past 12 months and 22 percent experiencing some form of other abuse in the past 12 months, suggesting that female entrepreneurs in Ethiopia face harassment problems when attempting to operate businesses in general. 89 Appendix D: Details on Each Individual Study Ethiopia (continued) Given these • Access to finance: The higher capital requirements for crossover businesses mean that constraints, what dedicated lending initiatives for female entrepreneurs are critical to easing financial policies could constraints and helping these businesses grow. help women to • Targeting: Policy efforts to encourage women to enter nontraditional sectors should establish cross over into which women are committed to operating a business (measured as a preference for business more profitable, rather than a necessity for money or inability to find wage work) as a first step in targeting the male-dominated appropriate women for these programs. sectors? • Training: Salaries for employees in crossover sectors are approximately double those of workers in noncrossover sectors. Programs that train women on the skills needed to operate in male-dominated sectors could help them compete for these higher-paying, salaried jobs. Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors • Networks: Facilitating access to networks and providing training on how to overcome discrimination and improve negotiation skills could give women a collective voice and improve some of the challenges women face when operating in male-dominated sectors. • Spousal support: With the finding that women with supportive husbands and male role models are more likely to perform well in a male-dominated sector, programs could encourage men to introduce their wives to their own business networks, pass on key technical skills, and help them obtain start-up capital. A better understanding of how husbands support their wives in business would help inform policy to replicate the support or advice structures that they provide. © Stephan Gladieu / World Bank 90 Appendix D: Details on Each Individual Study Guinea Our sample Number of enterprises: 465 microenterprises in Conakry How and where were the data collected, and what are they representative of? Face-to-face interviews were conducted in Conakry. To locate women entrepreneurs in male-dominated sectors, the study used a snowball sampling approach. Women entrepreneurs in non-male- dominated sectors and male entrepreneurs in male-dominated sectors were located using a systematic random selection in areas where the crossovers that we were able to interview were operating. Categories of business owners: Three categories of business owners: (1) women in male- dominated sectors (123); (2) women in female-concentrated sectors (213); and (3) men in male- Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors dominated sectors (129). Identifying What methodology was used to define sectors as male dominated or female concentrated? We male-dominated used data from a survey of small and medium-size enterprises that was conducted by the World and female- Bank in Conakry in 2014, as well as qualitative data collected in 2016 to identify male-dominated concentrated sectors. Specifically, the surveys collected information about the sector of operation and the sectors gender of business owners. Sectors were classified as male dominated if 70% or more of the owners of businesses in the sector were male. How many sectors are female concentrated? How many are male dominated? Using the definition of male dominated as 70% or more of businesses in a sector that are owned by men, 18 out of the 25 sectors we identified are classified as male dominated. In nine of these sectors, more than 90% of businesses were owned by men, including trades such as baking, brickmaking, mechanics, and carpentry. The profitarchy Women operating in male-dominated sectors have mean monthly profits (3,851,639 Guinean francs) that are statistically similar to those of men in male-dominated sectors (3,538,036 Guinean francs) and significantly higher than those of women operating in female-concentrated sectors (2,024,608 Guinean francs). It is important to note, though, that the differences in the mean profits per month should be interpreted with caution because the samples we use are not randomly drawn from the population of firms in Conakry, as it was extremely challenging to find female business owners in male-dominated sectors that were willing to participate in our survey. Furthermore, a small number of the respondents in our samples refused to respond (or indicated that they did not know the answer) to the questions on the monthly profits of their businesses. Factors associated • Education: Women crossovers are more likely to be literate and have higher levels of with being a education. They are more likely to have learned about the business at school or university. female crossover They are also more likely to have taken a training course in a male-dominated sector. • Business knowledge: Women crossovers self-report higher levels of knowledge with regard to practices such as managing workers, bookkeeping and budgeting, accessing credit, and regulatory compliance. They are also more likely to market their businesses. • Agency: Women crossovers, when compared to noncrossovers, report higher levels of empowerment in terms of being able to make personal decisions or making decisions within the household. • Prior experience in the sector: Women who own businesses in traditionally male-dominated sectors are more likely to have previous experience working in the sector or previous experience as entrepreneurs. • Family support and inheritance: Women crossovers are more likely to have inherited the business or to have had their spouse/partner start the business (they are less likely to start the business by themselves). Furthermore, they are more likely to indicate that their father had suggested the idea to start the business. For those with a spouse/partner, crossovers were more likely to receive assistance with and advice on running the business from their spouse/ partner. 91 Appendix D: Details on Each Individual Study Guinea (continued) Constraints Women crossovers were more likely than noncrossovers to list the following constraints to to business starting their businesses: performance faced • Accessing credit: The businesses run by women crossovers were more capital intensive than by women who those run by women noncrossovers, which may be one of the reasons why accessing credit is cross over specifically mentioned. • Social norms: Women crossovers were more likely to agree that if women wanted to work in a male trade, their family/friends would discourage them, to believe that women in male trades would have difficulty finding a husband, and to experience male employees looking down on them. They were also less likely to believe that people in the community took female business owners seriously. Finding staff to work for a woman was, for instance, listed as a constraint by Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors crossovers. • Household responsibilities: Women crossovers were also more likely to have to take care of household responsibilities than male business owners in male-dominated sectors. • Finding a location Given these • Education: Policy makers should implement measures that further increase access to higher constraints, what levels of education for girls and women in Guinea. This includes training that develops the policies could general business knowledge, as well as the specific knowledge required to operate businesses help women to in male-dominated sectors. cross over into • Social norms: Policies should be designed to address discriminatory social gender norms more profitable, in Guinea. First, this constraint should be addressed in the education system. Another male-dominated approach would be to use specifically designed media programming or other campaigns sectors? at a community level. Training interventions can be used to target interiorized gender-role norms among women entrepreneurs (or women who would like to start businesses), or to help crossovers cope with the psychological pressure of working in male-dominated sectors. Training can also be developed that targets the spouses/partners or other male relatives who would support women to establish businesses in male-dominated sectors. Further, training can be developed that targets the succession planning, which includes female family members or other relatives of male business owners in male-dominated sectors. • Access to credit: Promoting business plan competitions that specifically target women who have proposals for business opportunities in male-dominated sectors, with grants or low-interest loans as prizes, can support women crossovers. Other interventions such as psychometric testing can also be used to facilitate access to credit among women who want to establish or grow their businesses in male-dominated sectors. In addition, it may be possible to introduce business insurance products that are specifically tailored to any disruptions that are more likely to affect female entrepreneurs operating in male-dominated sectors (which may constrain their business’ performance and survival, and may inhibit women from crossing over). 92 Appendix D: Details on Each Individual Study Indonesia Our sample Number of enterprises: 14,019 enterprises How and where were the data collected, and what are they representative of? The data come from the Indonesian Family Life Survey for the years 2000, 2007, and 2014. The survey sample represented about 83% of the Indonesian population. Information on businesses is taken from the household business module. Categories of business owners: Four categories of business owners are included in the analysis: (1) women in male-dominated sectors (262); (2) women in female-concentrated sectors (6,519); (3) men in male-dominated sectors (1,939); and men in female-concentrated sectors (5,299). Identifying What methodology was used to define sectors as male dominated or female concentrated? Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors male-dominated Because the survey data are from a fairly representative sample of the population, we use them and female- to define male-dominated sectors as those in which more than 70% of businesses in the sector concentrated are owned by men. sectors How many sectors are female concentrated? How many are male dominated? Out of 17 sectors, 10 are classified as female concentrated and 7 as male dominated. On average, businesses in male-dominated sectors have almost twice the amount of start-up capital as businesses in female-concentrated sectors. The profitarchy In Indonesia, there is a premium for women crossing over into male-dominated sectors: those who cross over earn yearly profits (5,320,000 Indonesian rupiah) that are higher than those of women who operate in female-concentrated sectors (3,188,000 Indonesian rupiah). However, women still earn less than men, regardless of the sector in which they operate. Men in female- concentrated sectors earn yearly profits (6,273,000 Indonesian rupiah) that are higher than women in those sectors, and men in male-dominated sectors earn yearly profits (5,804,000 Indonesian rupiah) that are higher than women in these sectors. For men, male-dominated sectors are not more profitable than female-concentrated sectors: they actually have slightly higher profits when they work in female-concentrated sectors (although this difference is not statistically significant). Factors associated • Education: Female crossovers have slightly more education than women operating in female- with being a concentrated sectors. female crossover • Wage job: Female crossovers are more likely to also have a wage job, which could enable them to have access to more capital to invest in their businesses or help broaden their networks. • Networking: Consistent with a hypothesis of the importance of networks for crossing over, crossovers are also more likely to participate in community service volunteering, through which women may benefit from expanded networks. Capital from wage work or from networks may be particularly important for enabling women to work in male-dominated sectors. Women operating in male-dominated sectors have start-up capital that is more than four times that of women operating in female-concentrated sectors. Constraints • Access to finance: Crossovers may face more challenges securing adequate financing for to business their businesses, as crossovers have much larger capital requirements at start-up. In addition, performance faced crossovers’ profits are strongly correlated with the availability of credit in the village. by women who • Time: Childcare constraints affect crossover women more than noncrossover women. Having cross over a two-year-old or younger child in the household is associated with profits that are 49 percent lower for crossover women, whereas the profits for noncrossover women are not lower when young children are in the household. Given these • Access to finance: Programs that enable women to have access to greater capital may constraints, what support entry into male-dominated sectors, which have higher start-up capital requirements. policies could Facilitating access to credit may also help women stay and thrive in these sectors. help women to • Skills: Programs that support the development of women’s skills through formal training could cross over into broaden the sectoral opportunities for female entrepreneurs. more profitable, • Education: Formal education is positively correlated with both the likelihood of operating a male-dominated business in a male-dominated sector and also with the profits of female crossovers. sectors? • Networks: Programs that expand women’s networks and provide opportunities to learn from others may also support female crossovers. An expanded network may enable women to learn about opportunities in male-dominated sectors and overcome barriers to entry. Mentorship programs connecting crossovers to knowledgeable individuals may also enable them to be more successful once they cross into the sector. 93 Appendix D: Details on Each Individual Study Lao PDR Our sample Number of enterprises: 2,015 enterprises How and where were the data collected, and what are they representative of? The data come from the Lao PDR Expenditure and Consumption Survey (LECS) for the years 2012–2013. The survey is a nationally representative household survey, and the information on businesses comes from the household business module. Categories of business owners: Four categories of business owners: (1) women in male- dominated sectors (40); (2) women in female-concentrated sectors (1003); (3) men in male- dominated sectors (311); and men in female-concentrated sectors (661). Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Identifying What methodology was used to define sectors as male dominated or female concentrated? male-dominated Because the survey is nationally representative, we define a sector as a male-dominated sector if and female- at least 70% of the owners of businesses in the sector are men. concentrated How many sectors are female concentrated? How many are male dominated? Out of 14 sectors, sectors 8 are classified as female concentrated and 6 as male dominated. Male-dominated sectors have more paid workers and may be more physically demanding (such as construction, transportation, or agriculture). The profitarchy Women crossovers have monthly sales (8,557,000 Lao kip) that are higher than the sales of women operating in female-concentrated sectors (4,768,000 Lao kip). They also have higher sales than men operating in any sector; however, while substantial in size, the difference in profits between female crossovers and men is not statistically significant. Due to the small sample of female crossovers, the results must be interpreted with caution. Factors associated • Fewer domestic constraints: Crossovers are more likely to have water in the home during with being a the dry season, which can reduce the time needed for fetching water, and have fewer elderly female crossover household members that may require care. • Education: Skills may play a role in helping women enter male-dominated sectors, as crossovers have slightly more education than women in female-concentrated sectors. Constraints • Time: Women operating in male-dominated sectors may need to be able to spend more time to business to fully dedicate to their businesses. When crossovers also hold a wage job, their profits are performance faced much lower, whereas the same is not true for women in female-concentrated sectors. by women who • Couple dynamics: Being married is also negatively associated with profits, only for crossover cross over women. If married women spend more time on domestic tasks than unmarried women, this may signal a challenge with balancing multiple tasks. Aligned with this possibility, crossover women tend to have higher profits when they also have access to cooking fuel that requires less time to collect and ignite. Given these • Domestic work: Promoting a more equal sharing of domestic tasks across household members constraints, what of both genders or a reduction of domestic work overall could help women enter and be policies could more profitable in male-dominated sectors. Facilitating affordable access to care services, help women to such as child and elder care, could, for instance, be helpful, as well as increasing affordable cross over into access to modern technology that lowers the time needed for cooking and cleaning. Similarly, more profitable, encouraging men and boys to participate more in domestic work could also alleviate time male-dominated constraints that may orient female entrepreneurs’ choices. sectors? • Skills: Supporting women’s skill development, for example, through training programs, may also help them cross over into male-dominated sectors. 94 Appendix D: Details on Each Individual Study Mexico Our sample Number of enterprises: This study employs two sources of information: first, a nationally representative survey of microenterprises (ENAMIN), which contains data for 13,798 male and 13,784 female microentrepreneurs; and second, a baseline survey conducted as part of a randomized controlled trial to evaluate the impacts of a personal initiative training program paired with business skills. The sample consists of 3,907 informal and formal female entrepreneurs operating in five states of Mexico (México DF, State of Mexico, Guanajuato, Querétaro, and Aguascalientes). How and where were the data collected, and what are they representative of? The ENAMIN Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors survey was conducted in 2014 by the Institute of Geography and Statistics (INEGI) and it is a representative sample of Mexican microentrepreneurs at the national and state level. The sample frame for ENAMIN comes from the previous year’s Employment Survey, which identifies respondents, who are classified as self-employed or microentrepreneurs (owners of firms with less than 10 employees). The impact evaluation survey (“the experiment baseline survey” hereafter) is a nonrandom sample of female entrepreneurs interested in the training program implemented by CREA (a Mexican nongovernmental organization specialized in programs and support for female entrepreneurs). Researchers launched a communication campaign in five states to seek female- owned businesses from the formal and informal sectors interested in improving and growing their businesses. Only entrepreneurs who already had a business which had been working for 12 months or more were eligible to participate; in this way a total of 3,907 entrepreneurs were admitted and at registration had to fill a baseline survey between November 2014 and August 2015. Categories of business owners: The ENAMIN data include four categories of business owners: (1) men in male-dominated sectors (8,079); (2) women in male-dominated sectors (988); (3) men in female-concentrated sectors (5,719); and (4) women in female-concentrated sectors (12,796). The experiment baseline survey includes two categories of female business owners: (1) women in male-dominated sectors (368); and (2) women in female-concentrated sectors (3,539). Identifying What methodology was used to define sectors as male dominated or female concentrated? We male-dominated use the distribution of self-reported ownership by sex in ENAMIN to classify sectors into male and female- dominated or female concentrated. If more than 70 percent of businesses from the ENAMIN concentrated sample were owned by men, we defined that sector as male dominated. This definition was sectors applied to the sample of female entrepreneurs in the experiment baseline survey. Alternative definitions of male-dominated sectors at 65 and 75 percent were analyzed, and the results are robust across the different classifications. How many sectors are female concentrated? How many are male dominated? Out of 22 sectors, 17 are classified as male dominated. Overall, male-dominated sectors in Mexico are capital intensive (automotive repair, land transportation, and mining), followed by sectors that require ICT-type skills, finance, and stock market activities. The profitarchy Excluding outliers, women operating in male-dominated sectors have weekly profits (2,563 Mexican pesos) that are significantly higher than those of women operating in female- concentrated sectors (1,391 Mexican pesos). Based on the ENAMIN data, which also include male entrepreneurs, men in male-dominated sectors have the highest profits (6,223.3 Mexican pesos), followed by men in female-concentrated sectors (5,473.9 Mexican pesos). However, female-owned firms in male-dominated sectors have monthly profits (3,826 Mexican pesos) that are statistically higher than those of women operating in female-concentrated sectors (2,541.6 Mexican pesos). 95 Appendix D: Details on Each Individual Study Mexico (continued) Factors associated • Having a male mentor (but not a husband mentor): The subset of crossover women who had a with being a male mentor reported that they help them improve or solve problems related to the business female crossover (39 percent), help them with starting or planning their business (25 percent), finding clients (20 percent), finding suppliers (10 percent), or accessing financing (7 percent). Having a male mentor is found to have a positive effect on key business performance indicators, even after controlling for the effect of crossing over. Out of the female entrepreneurs who reported having a mentor, those with a male mentor have on average 17 percent higher profits and 19 percent higher revenues per week. • Being exposed to a male role model: The businesses of crossover women who report having an inspiring male role model also have better performance, compared to those who do not have one. • Having higher levels of education: A female entrepreneur with secondary education is 3 Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors percentage points more likely than one with primary education to enter a male-dominated sector. However, noncognitive skills do not seem to affect the probability of choosing a male- dominated sector. • Having higher levels of cognitive ability: Women entrepreneurs with a high level of cognitive ability are also 0.8 percent more likely to cross over. • Having a father with higher levels of education: One additional year in the father’s education increases the probability that a woman crosses over to a male-dominated sector by 0.3 percent. The mother’s level of education is, however, not statistically significant. • Having a husband who makes decisions about the business Constraints Our survey did not include a review of constraints; however, the National Survey on Productivity to business and Competitiveness of Micro, Small, and Medium Businesses (ENAPROCE 2015) provides performance faced non-gender-disaggregated data on the topic. In their survey, 23 percent of microentrepreneurs by women who identify the lack of credit as the main challenge for business growth, followed by competing with cross over informal businesses (19.3 percent) and low demand of their goods or services (17 percent). Only 10 percent mention high or complex taxes as the main challenge for business growth. Given these • Mentorship opportunities: Networks are often less available and less diverse for women constraints, what entrepreneurs, which puts them at a disadvantage. Male mentors can be valuable as business policies could partners or for support. They can also help women gain market or customer information. help women to • Engaging men: Skills training programs can engage husbands to offset gender norms or cross over into attitudes that might constrain women from successfully participating in programs. Men could more profitable, be encouraged to introduce female entrepreneurs to their own business networks, pass on male-dominated key technical skills, or help them access financing opportunities. sectors? • Providing information about sector-specific profitability: Showcasing higher-return businesses in male- compared to female-dominated sectors could change beliefs about profitability and encourage women to enter male-dominated fields. It can also motivate women to enroll in skills training in nontraditional trades. Information can be provided through career guidance in schools, informational sessions accompanying skills training programs, or through ‘edutainment.’ • Skills: Training programs offer opportunities to improve outcomes for female entrepreneurs. However, it is often more difficult for women than for men to access and complete these programs due to family responsibilities, movement restrictions, and gender norms. Programs can integrate “smart design” aspects such as childcare options, holding training in accessible and safe locations, and making transportation easy and safe, to help women overcome these constraints. 96 Appendix D: Details on Each Individual Study Peru Our sample Number of enterprises: The baseline survey included 1,905 female PDM clients, 1,148 of whom reported running a microbusiness. How and where were the data collected, and what are they representative of? The team worked with women clients of the Program Palabra de Mujer – PDM (Word of a Woman) of Financiera Confianza in Pucallpa, a city in the eastern region of Peru. Financiera Confianza is a regulated microfinance firm that has half a million clients, and the program Palabra de Mujer targets women microentrepreneurs using a group-based lending methodology. The survey was not intended to be representative of women microentrepreneurs in Pucallpa or Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors of the microfinance clients of Financiera Confianza. As part of a randomized controlled trial, the sample was selected to be able to show, with some degree of confidence, the impacts of three different interventions: training in hard skills, training in hard as well as soft skills, and training in soft skills that promote employability. The clients were all initially contacted during the monthly group payment meeting, and appointments were arranged for the participants to be later visited by the surveyor. The mean age of the clients was 36 years old, 67% were married or had a civil partner, and their average business tenure was about 4 years. Categories of business owners: The survey includes two categories of female business owners: women in male-dominated sectors (93); and women in female-concentrated sectors (1,055). Identifying What methodology was used to define sectors as male dominated or female concentrated? male-dominated We use the distribution of self-reported ownership by sex in ENAMIN (Mexico’s nationally and female- representative survey of microentrepreneurs) to classify sectors into male dominated or female concentrated concentrated. In this study, if more than 70 percent of businesses from the ENAMIN sample sectors were owned by men, we defined that sector as male dominated. This definition was applied to the sample of female entrepreneurs in the experiment baseline survey. Alternative definitions of male-dominated sectors at 65 and 75 percent were analyzed, and the results are robust across the different classifications. How many sectors are female concentrated? How many are male dominated? Out of 46 sectors, 26 are classified as male dominated. Overall, male-dominated sectors in Peru are primary activities (farming and forestry) and capital-intensive activities (automotive repair). On the other hand, female-concentrated sectors are related to services (grocery sales, food) and labor-intensive activities such as clothing. The profitarchy Women in MDS have monthly profits (330 US dollars) significantly higher than those of women operating in female-concentrated sectors (193 US dollars). We found similar results on monthly sales, total number of workers, and paid workers. Factors associated • Women crossovers have higher soft skills as well as higher levels of locus of control. with being a • They are also more likely to have a dwelling of good quality: having walls of brick or concrete female crossover block or having a ceiling of reinforced concrete increase the probability that a woman crosses over. There is, however, no effect of access to water, electricity, or the level of overcrowding. • We found an increase in the probability to cross over to a male-dominated sector if the woman had fewer children. • Having an entrepreneur partner increases the probability of crossing over. Constraints • This country survey did not include information on constraints to business performance. to business performance faced by women who cross over 97 Appendix D: Details on Each Individual Study Peru (continued) Given these • Improving hard and soft skills: Provide women entrepreneurs with the training they need to constraints, what develop the right skills and a growth-oriented mindset. Psychology-based trainings, such as policies could personal initiative training, that enhance women’s noncognitive skills and foster a proactive, help women to resilient, and entrepreneurial mindset, can help women introduce new innovative products in cross over into their businesses and increase their earnings. more profitable, • Reduce domestic work: Promote an equal sharing of domestic tasks across household male-dominated members of both genders to allow women to invest more time in their businesses. Facilitating sectors? affordable access to care services, such as childcare, could also be helpful to increase women’s economic participation and agency while also stimulating early childhood development. • Engaging men: Engage men to provide direct support to their wives. This support can be in Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors the form of economic empowerment or business support that leverages the skills, knowledge, and networks of male family members. For example, a gender transformation and joint training intervention in Côte d’Ivoire showed that male export crop farmers who filled out a two-year action plan with their wives shared more agricultural decisions, and enabled women to manage more cash-crop tasks. 98 Appendix D: Details on Each Individual Study Uganda Our sample Number of enterprises: 326 women and 409 male entrepreneurs in urban Uganda within and just outside Kampala. How and where were the data collected, and what are they representative of? We used quantitative data from a 2011 sampling of 735 entrepreneurs, most of whom belonged to the Katwe Small Scale Industry Association (KASSIDA). In addition, a quantitative and qualitative survey was administered in 2012 to 63 crossovers and to 120 women working in traditionally female sectors. Of the latter, half of the participants were randomly sampled, and half were matched to the crossovers based on a number of pre-business characteristics, such as similar age and completion of primary school. To get a better sense of communitywide perceptions Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors about female entrepreneurs, we also conducted 17 focus group discussions with crossovers, noncrossovers, clients, suppliers, and male employees, and interviewed 12 community leaders and credit providers. Categories of business owners: Three categories of business owners: (1) women in male- dominated sectors (30); (2) women in female-concentrated sectors (296); and (3) men in male- dominated sectors (321). Identifying What methodology was used to define sectors as male dominated or female concentrated? male-dominated Sectors were classified as male dominated if, based on the baseline survey data, at least 75% and female- of enterprises in that sector were owned by men, otherwise the sector was classified as female concentrated concentrated. sectors How many sectors are female concentrated? How many are male dominated? Out of 9 sectors, 4 are classified as female concentrated, and 5 as male dominated. The profitarchy Firms owned by crossovers make double the profits, on average, of firms owned by noncrossovers, with profits in the past month at 494,000 Uganda shillings compared to 205,000 Uganda shillings respectively. Within male-dominated sectors, businesses owned by women who cross over are just as profitable as businesses owned by men (494,000 Uganda shillings and 515,000 Uganda shillings respectively). Factors associated • Information about sectors: Many female entrepreneurs are simply not aware that they could with being a be earning higher profits in male-dominated sectors. About 75% of the noncrossovers we female crossover interviewed incorrectly believe that they make the same or more than crossovers, when in fact they do not. • Role models: Most crossovers do not come up with the idea of working in a male-dominated sector by themselves. Rather, that decision originates from someone else’s suggestion, observing others, or being offered a job in the sector by a friend or family member. Women who reported having a male role model in their youth were 20%–28% more likely to be a crossover. Fathers and politicians are particularly strong role models for crossovers, either in introducing women to the sectors where they work, or by providing important contacts or financial support. On the other hand, noncrossovers are more likely to have been introduced to traditionally female sectors by mothers, and especially teachers. This suggests that the current education system actually reinforces the gender segregation of labor. Moreover, once women engage in a traditionally female sector, they are unlikely to make the switch to a male- dominated sector. Therefore, early influence by a male role model is very important in shaping women’s professional path to a more profitable sector. • Apprenticeships: A significant intermediary step in becoming a crossover is active exposure to the sector by becoming an apprentice, engaging in actively learning the trade, or being taken to observe the trade. Constraints • Low technical skills: This is the most common constraint mentioned by crossovers, even to business though these women do not report making significantly lower profits than their male performance faced counterparts, nor do they have any trouble finding customers. But clients do acknowledge by women who that crossovers have limited technical skills, which could influence their decision about who to cross over engage for a large contract. • Access to credit: Both crossovers and noncrossovers commonly cite access to credit as a primary business issue. Crossovers are more likely to obtain credit from a bank or from a spouse, whereas noncrossovers most frequently borrow from a female friend or community member. 99 Appendix D: Details on Each Individual Study Uganda (continued) Given these • Information on earnings: Provide information early to youth about the profitability of certain constraints, what sectors, perhaps through informational campaigns or career guidance in schools. However, policies could given teachers’ current strong role in preventing women from crossing over, any program help women to using teachers requires significant training and sensitization of teachers. cross over into • Networks and support: Offer supportive engagement with individuals who can guide female more profitable, entrepreneurs as they seek to operate a business in a male-dominated sector. This is ideally male-dominated done by drawing from the entrepreneur’s existing network of friends and family, perhaps sectors? within the context of a youth mentorship program. • Early exposure to the sector: Facilitate active exposure to the sector through apprenticeships or other work experience programs. It is especially important to target young women who are Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors just entering the labor force, as well as older women without previous experience in a female- dominated sector. • Social norms: Engage figures of influence within communities to avoid potential opposition and to gain support in changing social perceptions of which sectors are appropriate for women. • Support crossovers in maintaining their businesses, such as by facilitating access to networks or by creating business organizations dedicated to crossovers. © Stephan Gladieu / World Bank 100 Appendix D: Details on Each Individual Study Vietnam Our sample Number of enterprises: 2,406 enterprises How and where were the data collected, and what are they representative of? Our data from from the Vietnam Access to Resources Household Survey (VARHS) for the years 2008, 2010, 2012, and 2014. The survey is representative for 12 provinces in rural areas. Categories of business owners: Four categories of business owners: (1) women in male- dominated sectors (106); (2) women in female-concentrated sectors (1170); (3) men in male- dominated sectors (354); and men in female-concentrated sectors (776). Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Identifying What methodology was used to define sectors as male dominated or female concentrated? male-dominated Because the data are representative, we define a sector as male dominated if more than 70% of and female- the owners of businesses in the sector are men. concentrated sectors How many sectors are female concentrated? How many are male dominated? Out of 18 sectors, 13 are classified as female concentrated and 5 as male dominated. The profitarchy Both men and women operating in male-dominated sectors have higher profits than individuals of the same gender who operate in female-concentrated sectors. For women, crossing over into a male-dominated sector is associated with higher annual profits (15,038,000 Vietnamese dong for female crossovers compared to 11,405,000 Vietnamese dong for noncrossovers). While female crossovers have profitability comparable to men who operate in female-concentrated sectors (20,876,000 Vietnamese dong), their profits are still substantially lower than those of men who operate in similar sectors (26,166,000 Vietnamese dong). Although there is a premium to crossing over in Vietnam, a within-sector gender profit gap remains. Factors associated • Fewer time constraints related to domestic tasks: Women with fewer young children in the with being a household and who use cooking fuel that requires less time to gather and ignite are more female crossover likely to operate in male-dominated sectors. • Access to finance: Aligned with the higher start-up capital needed in male-dominated sectors, female crossovers are more likely to be in households that have borrowed money and have fewer household savings. • Being married • Living in an urban area Constraints • Access to finance: Accessing finance with acceptable terms may be a binding constraint for to business female crossovers. Businesses operating in male-dominated sectors had start-up capital that performance faced was more than twice that of businesses operating in female-concentrated sectors. While by women who they are more likely to have borrowed money than women operating in female-concentrated cross over sectors, having a loan is negatively associated with profits for female crossovers. To secure the capital needed to operate in a more capital-intensive male-dominated sector, crossovers may feel pushed to accept loans with usurious terms, even if these terms can make it harder to turn a profit under the weight of repayment. • Competition: Crossovers may also face challenges with competition. Although crossovers are more likely to live in urban areas, their profits are higher when they live in rural areas, where they may face less competition from other entrepreneurs in the sector. • Time: Unpaid care work may also pose a greater constraint for female crossovers. While women with fewer young children and easier sources of cooking fuel are more likely to work in male-dominated sectors, having elderly household members who may require care has a stronger negative relationship with the profits of female crossovers than those of women operating in female-concentrated sectors. 101 Appendix D: Details on Each Individual Study Vietnam (continued) Given these • Fewer domestic responsibilities: Policies that support a reduction in the time needed for constraints, what domestic tasks or a more equal sharing of domestic work across household members of both policies could genders could support female crossovers. Women are more likely to cross over into male- help women to dominated sectors if they have fewer young children and easier sources of cooking fuel, and cross over into their profits are more adversely correlated with the presence of elderly household members more profitable, that may need care. To enable other women to cross over, policies could facilitate affordable male-dominated access to care services, such as child and elder care. Increasing affordable access to modern sectors? technology that lowers the time needed for cooking and cleaning and encouraging men and boys to participate more in domestic work could also alleviate time constraints that may orient female entrepreneurs’ choices. Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors • Access to finance: Improving access to capital and the terms of credit could also support women in transitioning into more capital-intensive male-dominated sectors and support their performance once there. Interventions could promote savings or seek to improve the terms of loans that women entrepreneurs can receive through tools such as credit guarantee schemes or psychometric testing in absence of collateral. • Facing competition and access to markets: Policies could also support women in facing competition in urban areas and having a greater access to markets in rural areas. A tailored needs assessment could help identify specific interventions that would support female crossovers in facing competition in urban areas or finding ways of opening businesses in more remote areas. © Simone D. McCourtie / World Bank 102 Appendix D: Details on Each Individual Study Future of Business Survey Our sample Number of enterprises: The total sample size on which we base the analysis in the paper comprises 55,932 observations, with different sample sizes depending on the outcome being analyzed. How and where were the data collected, and what are they representative of? The survey was implemented on the Facebook mobile platform to administrators of Facebook Business Pages in the specified countries for a period of two weeks in early to mid-December 2018. The data were collected on the Facebook platform and the final clean and aggregated, anonymized data are now published on the World Bank Open Data portal. The sample of Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors respondents of the Future of Business survey are limited to those business owners/managers with internet access and Facebook accounts with a business page. The sample should not be considered representative of all business owners but a select group of global micro, small, medium, and large enterprises (0 to 5,000 employees). Categories of business owners: Four categories of business owners: (1) women in male- dominated sectors (3,237); (2) women in female-concentrated sectors (17,264); (3) men in male- dominated sectors (12,362); and (4) men in female-concentrated sectors (23,069). Identifying What methodology was used to define sectors as male dominated or female concentrated? In male-dominated our paper we define male-dominated sectors based on the question, “Who owns most of the and female- businesses in your sector? Men or women?” as reported by the subset of female respondents. If concentrated more than 70 percent of women report that men own most of the businesses within their sector, sectors we define that sector as male dominated. How many sectors are female concentrated? How many are male dominated? Of the 42 sectors, 18 are classified as male dominated by this definition. The profitarchy How do profits of women in MDS compare to profits of men in MDS, and of women in FCS? Men in male-dominated sectors are the top earners and make statistically significantly higher profits than all other categories. Men in male-dominated sectors, men in female-concentrated sectors, and female crossovers have 116, 82, and 67 percent statistically significantly higher profits than female noncrossovers, respectively. While different in magnitude, the difference between profits for males in female-concentrated sectors and female crossovers is not statistically different. After winsorizing, the mean in profits for the reference category (women in female-concentrated sectors) is equivalent to 126,489.20 US dollars for the pooled sample, 166,592.70 US dollars for the developed countries’ samples, and 112,189 US dollars for samples of developing countries. Factors associated • Starting early: Age is negatively associated with being a female crossover. While statistically with being a significant, these effects are very small: one year or more implies an increase in the probability female crossover of being a crossover by 0.1 percent. • Spousal support: Female crossovers are more likely to be married than female noncrossovers. Once we control for the set of socioemotional skills variables, though, this correlation disappears, partly because we lose a significant proportion of the sample when including these variables (about 35% of the observations). • Family support: Female crossovers were less likely to have started their business themselves and a chief reason for crossing over seems to be related to the fact that the business was inherited from the family. • Male role models: Having a male role model while growing up positively affects the likelihood of crossing over into a male-dominated sector. • Self-efficacy: Higher levels of self-efficacy are positively associated with the probability of being a female crossover. Conversely, women with higher levels of entrepreneurial identity and those more committed to staying in the sector are less likely to cross over. 103 Appendix D: Details on Each Individual Study Future of Business Survey (continued) Constraints • Access to finance: Female crossovers are significantly more likely to have a line of credit to business or loan from a financial institution than noncrossovers. Interestingly, female-owned firms in performance faced male-dominated sectors are also more likely to have credit than male-owned firms in the by women who same sector. This could be due to lower assets to draw upon in capitalizing the business. cross over Additionally, women in the higher-return crossover sectors rely more on family inheritance and savings provided by their spouse to start the business. Men in male-dominated sectors, on the other hand, are more likely to be able to draw from their own savings and loans from relatives and friends to start the business, relative to women. • Time: Compared to their male peers in male-dominated sectors, female crossovers work 1.5 hours fewer, and this pattern is more pronounced in developed countries. Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Given these • Male mentorship: Women in male-dominated sectors are more likely to have had a male role constraints, what model while growing up. policies could • Apprenticeships: Such programs could help provide the same kind of support that a role help women to model does. A key programmatic aspect would be to target younger women, ideally before cross over into they choose a career. When introducing potential career paths to young women, highlighting more profitable, the earnings potential and the growth in female representation in nontraditional sectors could male-dominated also motivate them to pursue a career in that industry. sectors? • Socioemotional skills training: We find that crossovers have higher self-efficacy, a socioemotional skill often associated with entrepreneurial success. • Improving household dynamics: We find that certain types of spousal support can be helpful for breaking into male-dominated trades, which suggests that improving intrahousehold dynamics could support women to cross over. 104 Aderemi, H. O., M. O. Ilori, W. O. Siyanbola, S. A. Adegbite and I. O. Abereijo 2008. “An Assessment of the Choice and Performance of Women Entrepreneurs in Technological and Non-Technological Enterprises in Southwestern Nigeria.” African Journal of Business Management 2 (10): 165–176. Adikaram, Arosha. 2017. “Not So Silent: How Social Support Is Given and Received in Dealing with Sexual Harassment in Sri Lankan Work-Places.” NSBM Journal of Management 1 (94). 10.4038/nsbmjm.v1i2.13. Alibhai, Salman, Niklas Buehren, Rachel Coleman, Markus Goldstein, and Francesco Strobbe. 2018. “Disruptive Finance: Using Psychometrics to Overcome Collateral Constraints in Ethiopia.” World Bank, Washington, DC. Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Alibhai, Salman, Niklas Buehren, Michael Frese, Markus Goldstein, Sreelakshmi Papineni, and Kathrin Wolf. 2019. “Full Esteem Ahead? Mindset-Oriented Business Training in Ethiopia.” Policy Research Working Paper 8892, World Bank, Washington, DC. Alibhai, Salman, Niklas Buehren, Sreelakshmi Papineni, and Rachael Pierotti. 2017. “Crossovers--Female Entrepreneurs Who Enter Male Sectors: Evidence from Ethiopia.” Policy Research Working Paper 8065, World Bank, Washington, DC. Aneke, Evelyn, Evelyn Derera, and Mapeto Bomani. 2017. “An Exploratory Study of Challenges Faced by Women Entrepreneurs in the Construction Industry in South Africa.” International Journal of Business and Management Studies 9 (2). Anna, A., G. Chandler, E. Jansen, and N. Mero. 1999. “Women Business Owners in Traditional and Non-Traditional Industries.” Journal of Business Venturing 15: 279–303. REFERENCES Asia Foundation. 2013. “Access to Trade and Growth of Women’s SMEs in APEC Developing Economies: Evaluating the Business Environment in Indonesia.” The Asia Foundation, Indonesia. Aterido, Reyes, Mary Hallward-Driemeier, and Carmen Pages. 2009. “Big Constraints to Small Firms’ Growth? Business Environment and Employment Growth across Firms.” Policy Research Working Paper WPS 5032, World Bank, Washington, DC. Babbitt, Laura G., Drusilla Brown, and Nimah Mazaheri. 2015. “Gender, Entrepreneurship, and the Formal–Informal Dilemma: Evidence from Indonesia.” World Development 72: 163–174. Bahramitash, R., and S. Kazemipour. 2011. “Veiled Economy: Gender and the Informal Sector.” In Veiled Employment: Islamism and the Political Economy of Women’s Employment in Iran, edited by R. Bahramitash and H. S. Esfahani, 226–255. Syracuse, New York: Syracuse University Press. Bandura, Albert. 1977. “Self-Efficacy: Toward a Unifying Theory of Behavioral Change.” Psychological Review 84 (2): 191–215. Bandura, Albert. 1999. “A Social Cognitive Theory of Personality.” In Handbook of Personality, edited by L. Pervin and O. John, 2nd edition, 154–196. New York: Guilford Publications. Banerjee, Abhijit, Eliana La Ferrara, and Victor Orozco. 2019. “Entertainment, Education, and Attitudes toward Domestic Violence.” AEA Papers and Proceedings, 109: 133–37. 105 REFERENCES Baranov, V., R. de Haas, and P. Grosjean. 2018. “Men Roots and Consequences of Masculinity Norms.” Discussion Paper 2018-041, Tilburg University, Center for Economic Research. Bardasi, Elena, and Abay Getahun. 2008. “Unlocking the Power of Women.” Toward the Competitive Frontier: Strategies for Improving Ethiopia’s Investment Climate, World Bank, Washington, DC. Bardasi, Elena, Shwetlena Sabarwal, and Katherine Terrell. 2011. “How Do Female Entrepreneurs Perform? Evidence from Three Developing Regions.” Small Business Economics 37: 417–441. Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Barnir, Anat, Warren Watson, and Holly Hutchins. 2011. “Mediation and Moderated Mediation in the Relationship among Role Models, Self‐Efficacy, Entrepreneurial Career Intention, and Gender.” Journal of Applied Social Psychology 41: 270–297. Barsoum, Ghada, Bruno Crepon, Drew Gardiner, Bastien Michel, and William Pariente. 2017. “Evaluating the Effects of Entrepreneurship Edutainment in Egypt.” AEA RCT Registry. http://www.socialscienceregistry.org/trials/370/history/17880. Bell, Alex, Raj Chetty, Xavier Jaravel, Neviana Petkova, and John Van Reenen. 2019. “Who Becomes an Inventor in America? The Importance of Exposure to Innovation.” The Quarterly Journal of Economics 134 (2): 647–713. https://doi.org/10.1093/qje/qjy028. Benhassine, Najy, David McKenzie, Victor Pouliquen, and Massimiliano Santini. 2018. “Does Inducing Informal Firms to Formalize Make Sense?” Journal of Public Economics 157: 1-14 http://dx.doi.org/10.1016/j.jpubeco.2017.11.004. Bernhardt, Arielle, Erica Field, Rohini Pande, and Natalia Rigol. 2019. “Household Matters: Revisiting the Returns to Capital among Female Microentrepreneurs.” AER: Insights 1(2): 141–160. https://doi.org/10.1257/aeri.20180444. Breda, Thomas, Julien Grenet, Marion Monnet, and Clementine Van Effenterre. 2020. “Can Female Role Models Reduce the Gender Gap in Science? Evidence from Classroom Interventions in French High Schools.” IZA Discussion Papers No. 13163. Brouwers, Symen A., Fons J.R. Van de Vijver, and Dianne A. Van Hemert. 2009. “Variation in Raven’s Progressive Matrices Scores across Time and Place.” Learning and Individual Differences 19 (3): 330–338. Buse, K., D. Bilimoria, and S. Perelli. 2013. “Why They Stay: Women Persisting in US Engineering Careers.” Career Development International 18 (2): 139–154. Buvinic, M., J.C. Knowles, and F. Witoelar. 2018. “Unequal Ventures: Results from a Baseline Study of Gender and Entrepreneurship in East Java, Indonesia.” CGD Evidence- Based Report Series on Women’s Economic Empowerment No. 2, Washington, DC: Center for Global Development. Campos, Francisco, Michael Frese, Markus Goldstein, Leonardo Iacovone, Hillary Johnson, David McKenzie, and Mona Mensmann. 2018. “Personal Initiative Training Leads to Remarkable Growth of Women-Owned Small Businesses in Togo.” Gender Innovation Lab Policy Brief 22, World Bank, Washington, DC. 106 REFERENCES Campos, Francisco, Markus Goldstein, Laura McGorman, Ana Boudet, and Obert Pimhidzai. 2017. “Breaking the Metal Ceiling: Female Entrepreneurs Who Succeed in Male-Dominated Sectors.” WIDER Working Paper 2017/166, The United Nations University World Institute for Development Economics Research (UNU-WIDER), Helsinki. http://dx.doi.org/10.35188/UNU-WIDER/2017/392-9. Campos, Francisco, Leitao Moraes, Markus Goldstein, Laura McGorman, Ana Maria Munoz Boudet, and Obert Pimhidzai. 2015. “Breaking the Metal Ceiling: Female Entrepreneurs Who Succeed in Male-Dominated Sectors.” Policy Research Working Paper WPS 7503, World Bank, Washington, DC. Carranza, Eliana, Chandra Shekar Dhakal, and Inessa Love. 2018. “Female Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Entrepreneurs: How and Why Are They Different?” Jobs Working Paper Series 20. World Bank, Washington, DC. Chang, Wei, Lucia Diaz-Martin, Akshara Gopalan, Eleonora Guarnieri, Seema Jayachandran, and Claire Walsh. 2020. “What Works to Enhance Women’s Agency: Cross-Cutting Lessons from Experimental and Quasi-Experimental Studies”. Boston, MA. Chen, M. A. 2001. “Women and Informality: A Global Picture, the Global Movement.” SAIS Review 21 (1): 71–82. https://doi.org/10.1353/sais.2001.0007. Cherchi, Ludovica, and D. Kirkwood. 2019. “Crossovers in Botswana: Women Entrepreneurs Who Operate in Male-Dominated Sectors: Output for Women Entrepreneurship Study (P164089).” World Bank, Washington, DC. Chhair, S., and L. Ung. 2016. “Economic History of Industrialization in Cambodia.” Learning to Compete Working Paper No. 7. The Brookings Institute. https://www. brookings.edu/wp-content/uploads/2016/07/L2C_WP7_Chhair-and-Ung-v2-1.pdf. Colom, R., and O. Garcia-Lopez. 2002. “Sex Differences in Fluid Intelligence among High School Graduates.” Personality and Individual Differences 32: 445–451. Croke, Kevin, Markus Goldstein, and Alaka Holla. 2018. “Can Job Training Decrease Women’s Self-Defeating Biases? Experimental Evidence from Nigeria.” Gender Innovation Lab Policy Brief 28, World Bank, Washington, DC. Das, Smita, Clara Delavallade, Ayodele Fashogbon, Wale Ogunleye, and Sreelakshmi Papineni. 2021. “Occupational Sex Segregation in Agriculture: Evidence on Gender Norms and Socio-Emotional Skills in Nigeria.” Policy Research Working Paper WPS 9695, World Bank, Washington, DC. Delecourt, Solene, and Odyssia Ng. 2019. “Does Gender Matter for Small Business Performance? Experimental Evidence from India.” https://cega.berkeley.edu/wp-content/uploads/2020/03/Ng_PacDev2020.pdf. de Mel, Suresh, David McKenzie, and Christopher Woodruff. 2009. “Are Women More Credit Constrained? Experimental Evidence on Gender and Microenterprise Returns.” American Economic Journal: Applied Economics 1 (3): 1–32. Demenet, Axel, Mireille Razafindrakoto, and Francois Roubaud. 2016. “Do Informal Businesses Gain from Registration and How? Panel Data Evidence from Vietnam.” World Development 84: 326–341. 107 REFERENCES de Walque, D. 2005. “Selective Mortality during the Khmer Rouge Period in Cambodia.” Population and Development Review 31 (2): 351–368. de Walque, D. 2006. “The Socio-Demographic Legacy of the Khmer Rouge Period in Cambodia.” Population Studies 60 (2): 223–231. Doyle, Kate, Ruti Levtov, Gary Barker, Gautam Bastian, Jeffrey B. Bingenheimer, Shamsi Kazimbay, Anicet Nzabonimpa, Julie Pulerwitz, Felix Sayinzoga, Vandana Sharma, and Dominick Shattuck. 2018. “Gender-Transformative Bandebereho Couples’ Intervention to Promote Male Engagement in Reproductive and Maternal Health and Violence Prevention in Rwanda: Findings from a Randomized Controlled Trial.” PLoS ONE 13(4): Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors e0192756. https://doi.org/10.1371/journal.pone.0192756. Evans, Alice. 2014. “‘Women Can Do What Men Can Do’: The Causes and Consequences of Growing Flexibility in Gender Divisions of Labour in Kitwe, Zambia.” Journal of Southern African Studies 40 (5): 981–998. doi: 10.1080/03057070.2014.946214. Fairlie, Robert, and Alicia Robb. 2009. “Gender Differences in Business Performance: Evidence from the Characteristics of Business Owners Survey.” Small Business Economics 33: 375–395. Gassier, Marine, Léa Rouanet, and Lacina Traore. Forthcoming. Addressing gender- based segregation through information: Evidence from a randomized experiment in the Republic of Congo. World Bank, Washington, DC. Goldin, Claudia. 2014a. “A Grand Gender Convergence: Its Last Chapter.” American Economic Review 104 (4): 1091–1119. Goldin, Claudia. 2014b. “A Pollution Theory of Discrimination: Male and Female Differences in Occupations and Earnings.” National Bureau of Economic Research Working Paper 8985 https://www.nber.org/papers/w8985. Goldstein, Markus, Paula Gonzalez Martinez, and Sreelaksmi Papineni. 2019. “Tackling the Global Profitarchy: Gender and the Choice of Business Sector.” World Bank Policy Research Working Paper 8865. https://openknowledge.worldbank.org/ handle/10986/31747. Grosjean, P., and R. Khattar. 2019. “It’s Raining Men! Hallelujah? The Long-Run Consequences of Male-Biased Sex Ratios.” The Review of Economic Studies 86 (2): 723–754. Guizzo, Francesca, and Mara Cadinu. 2020. “Women, Not Objects: Testing a Sensitizing Web Campaign against Female Sexual Objectification to Temper Sexual Harassment and Hostile Sexism.” Media Psychology. https://www.tandfonline.com/doi/abs/10.1080/15 213269.2020.1756338. Hallward-Driemeier, Mary. 2013. “Enterprising Women: Expanding Economic Opportunities in Africa.” Africa Development Forum. Washington, DC: Agence Francaise de Developpement and the World Bank. Hardy, Morgan, and Gisella Kagy. 2018. “Mind the (Profit) Gap: Why Are Female Enterprise Owners Earning Less Than Men?” American Economic Association Papers and Proceedings 108: 252–255. 108 REFERENCES Hardy, Morgan, and Gisella Kagy. 2020. “It’s Getting Crowded in Here: Experimental Evidence of Demand Constraints in the Gender Profit Gap.” The Economic Journal 130 (631): 2272–2290. https://doi.org/10.1093/ej/ueaa040. Haupt, Theo and Kolosa Madikizela 2009. “Why Do South African Women Choose Careers in Construction?” Acta Structilia 16 (2). Haupt, Theo, and Jabulile Ndimande. 2019. “A Review on Why Women-Owned Construction Firms Fail in KwaZulu-Natal?” Journal of Physics: Conference Series 1378022070. Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Heuveline, P., and B. Poch. 2007. “The Phoenix Population: Demographic Crisis and Rebound in Cambodia.” Demography 44 (2): 405–426. Hicks, Joan Hamory, Michael Kremer, Isaac Mbiti, and Edward Miguel. 2011. “Breaking Gender Barriers: Training Vouchers and Kenyan Youth.” Africa Region Gender Practice Policy Brief 3, World Bank, Washington, DC. Hoffmann, A., M. Junge, and N. Malchow-Moller. 2015. “Running in the Family: Parental Role Models in Entrepreneurship.” Small Business Economics 44: 79–104. Hsieh, Chang-tai, Erik Hurst, Charles I. Jones, and Pete Klenow. 2019. “The Allocation of Talent and U.S. Economic Growth.” Econometrica 87 (5): 1439–1474. Hsieh, Chang-tai, and Pete Klenow. 2016. “Productivity and Misallocation.” NBER Reporter 1: 21–23 National Bureau of Economic Research (NBER), Cambridge, MA. Hundley, G. 2001. “Why Women Earn Less Than Men in Self-Employment.” Journal of Labor Research 22: 817–829. ICRW (International Center for Research on Women). 2019. “Women Entrepreneurs Need More Than Capital: What Do Women’s Businesses Really Need to Grow and Thrive?” ICRW. https://www.icrw.org/publications/women-entrepreneurs-need- more-than-capital/. IFC. 2019. “Insuring Women Ensures Sustainable Economic Development.” IFC Blog. https://www.ifc.org/wps/wcm/connect/news_ext_content/ifc_external_corporate_ site/news+and+events/news/women-insurance. International Labour Organization. 2018. “Care Work and Care Jobs for the Future of Decent Work.” Geneva: ILO. https://www.ilo.org/global/publications/books/ WCMS_633135/lang--en/index.htm. Jayachandran, Seema. 2020. “Social Norms as a Barrier to Women’s Employment in Developing Countries.” NBER Working Paper w27449. https://ssrn.com/ abstract=3637751. Kalleberg, Arne L., and Kevin T. Leicht. 1991. “Gender and Organizational Performance: Determinants of Small Business Survival and Success.” The Academy of Management Journal 34 (1): 136–161. www.jstor.org/stable/256305. Kelly, D.J, B.S. Baumer, C. Brush, P.G. Greene, M. Mahdavi, M. Majbouri, M. Cole, M. Dean, R. Heavlow, Babson College, Smith College, and the Global Entrepreneurship Research Association (GERA). 2017. “Global Entrepreneurship Monitor 2016/2017 109 REFERENCES Report on Women’s Entrepreneurship.” https://gemconsortium.org/report/gem- 20162017-womens-entrepreneurship-report. Kelley, D.J., S. Singer, and M.D. Herrington. 2016. “Global Entrepreneurship Monitor 2016: Global Report.” London: Global Entrepreneurship Research Association, London Business School. Klapper, Leora F. and Simon C. Parker. 2010. “Gender and the Business Environment for New Firm Creation.” The World Bank Research Observer 26: 237–257. Kiefer, Tina. 2018. “What Does an Effective Leader Look Like?” Warwick Business School. https://www.wbs.ac.uk/news/what-does-an-effective-leader-look-like/ Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors and https://www.nytimes.com/2018/03/16/health/women-leadership-workplace. html. La Ferrara, Eliana, Alberto Chong, and Suzanne Duryea. 2008. “Soap Operas and Fertility: Evidence from Brazil.” Working Paper 633, Inter-American Development Bank, Research Department, Washington, DC. Lafortune, Jeanne, Julio Riutort, and Jose Tessada. 2018. “Role Models or Individual Consulting: The Impact of Personalizing Micro-Entrepreneurship Training.” American Economic Journal: Applied Economics 10 (4): 222–45. Lancaster University. 2020. “Role Models Have Major Influence on Female University Choices: Women Exposed to Successful and Charismatic Role Models Are More Likely to Follow Them in Choosing a University Major.” ScienceDaily. http://www.sciencedaily. com/releases/2020/07/200701084643.htm. Makarova, Elena, Belinda Aeschlimann, and Walter Herzog. 2016. “Why Is the Pipeline Leaking? Experiences of Young Women in STEM Vocational Education and Training and Their Adjustment Strategies.” Empirical Research in Vocational Education and Training 8 (2). https://doi.org/10.1186/s40461-016-0027-y. Manning, C. 1998. Indonesian Labour in Transition: An East Asian Success Story? Cambridge, MA: Cambridge University Press. Mansour, Hani, Daniel I. Rees, Bryson M. Rintala, and Nathan N. Wozny. 2020. “The Effects of Professor Gender on the Post-Graduation Outcomes of Female Students.” NBER Working Papers 26822, National Bureau of Economic Research, Inc. Martin, P. and A. Barnard. 2013. “The Experience of Women in Male-Dominated Occupations: A Constructivist Grounded Theory Inquiry.” SA Journal of Industrial Psychology/SA Tydskrif vir Bedryfsielkunde 39 (2), Art. #1099. Mayoux, L. 1995. “From Vicious to Virtuous Circles? Gender and Micro-Enterprise Development.” Occasional paper, 3rd ed. Geneva: United Nations Research Institute for Social Development, UNDP. McKenzie, David. 2017. “Identifying and Spurring High-Growth Entrepreneurship: Experimental Evidence from a Business Plan Competition.” American Economic Review 107 (8): 2278–2307. McKenzie, David, and Anna Luisa Paffhausen. 2017. “Small Firm Death in Developing Countries.” Policy Research Working Paper 8236, World Bank, Washington, DC. 110 REFERENCES Meunier, Frederic, Yulia Krylova, and Rita Ramalho. 2017. “Women’s Entrepreneurship: How to Measure the Gap between New Female and Male Entrepreneurs?” World Bank Policy Research Working Paper 8242, World Bank, Washington, DC. Nissan, E., M. Castano, and I. Carrasco. 2012. “Drivers of Non-Profit Activity: A Cross- Country Analysis.” Small Business Economics 38 (3): 303–320. Okamoto, Dina and Paula England. 1999. “Is There a Supply Side to Occupational Sex Segregation?” Sociological Perspectives 42 (4): 557–582. Pike, Isabelle. 2018. “‘Work is Work’: Crossing Gendered Lines in Central Kenya.” Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors African Studies Association UK conference, Birmingham, UK. Rauch, A., and M. Frese. 2007. “Let’s put the person back into entrepreneurship research: A meta-analysis on the relationship between business owners’ personality traits, business creation, and success.” European Journal of Work and Organizational Psychology 16 (4). Reed, Debbie, Albert Yung-Hsu Liu, Rebecca Kleinman, Annalisa Mastri, Davin Reed, Samina Sattar, and Jessica Ziegler. 2012. “An Effectiveness Assessment and Cost- Benefit Analysis of Registered Apprenticeship in 10 States.” Mathematica Policy Research. Rijkers, Bob, and Rita Costa. 2012. “Gender and Rural Non-Farm Entrepreneurship.” World Bank Policy Research Working Paper 6066, World Bank, Washington, DC. Rosa, Julio, and Daouda Sylla. 2016. “A Comparison of the Performance of Majority Female-Owned and Majority Male-Owned Small and Medium-Sized Enterprises.” Ottawa, Canada: Statistics Canada Centre for Special Business Projects. Rose, Stephen J., and Heidi Hartmann. 2004. “Still a Man’s Labor Market: The Long- Term Earnings Gap.” Institute for Women’s Policy Research, Washington, DC. Sabarwal, Shwetlena, and Katherine Terrell. 2008. “Does Gender Matter for firm performance? Evidence from Eastern Europe and Central Asia.” IZA Discussion Papers, No. 3758, Institute for the Study of Labor (IZA), Bonn. http://nbn-resolving. de/urn:nbn:de:101:1-20081126115. Sappleton, N. 2009. “Women Non‐Traditional Entrepreneurs and Social Capital.” International Journal of Gender and Entrepreneurship 1 (3): 192–218. Schomer, Inka, and Alicia Hammond. 2020. “Stepping Up Women’s STEM Careers in Infrastructure: An Overview of Promising Approaches.” ESMAP Paper. World Bank, Washington, DC. Shankar, A., M. Onyura, and J. Alderman. 2015. “Agency-Based Empowerment Training Enhances Sales Capacity of Female Cookstove Entrepreneurs in Kenya.” Journal of Health Communications 20 (Suppl 1): 67–75. Singh, Surendra, Ruthie Reynolds, and Safdar Muhammad. 2001. “A Gender-Based Performance Analysis of Micro and Small Enterprises in Java, Indonesia.” Journal of Small Business Management 39 (2):174–82. 111 REFERENCES Tophoven, Silke, Jean-Baptist du Prel, Richard Peter, and Veronika Kretschmer. 2015. “Working in Gender-Dominated Occupations and Depressive Symptoms: Findings from the Two Age Cohorts of the LidA Study.” Journal for Labour Market Research 48: 247–262. United Nations, Department of Economic and Social Affairs, Population Division. 2019. “World Population Prospects 2019.” Custom data acquired via website. Vanek, J. 2013. “Entrepreneurship in Informal Employment: Gender Perspective.” WIEGO. https://unstats.un.org/unsd/gender/Events/5-6%20Dec%202013/ Session%203_Vanek.ppt. Breaking Barriers: Female Entrepreneurs Who Cross Over to Male-Dominated Sectors Vu, N. T. H, R. van Velzen, R. Lensik, and E. Bulte. 2015. “The Impact of Gender and Business Training for Female Microfinance Clients in Vietnam.” 3ie Grantee Final Report. New Delhi: International Initiative for Impact Evaluation (3ie). WeXchange and Inter-American Development Bank. 2020. “The Rise of Women STEMpreneurs: A Study on Women Entrepreneurs in STEM in Latin America and the Caribbean.” IDB Lab, Inter-American Development Bank. https://publications.iadb.org/ en/wx-insights-2020-rise-women-stempreneurs-study-women-entrepreneurs-stem- latin-america-and-caribbean. Wolf, Kathrin, and Michael Frese. 2018. “Why Husbands Matter: Review of Spousal Influence on Women Entrepreneurship in Sub-Saharan Africa.” Africa Journal of Management 4 (1): 1–32. World Bank. 2015. Adolescent Girls Initiative. https://www.worldbank.org/en/ programs/adolescent-girls-initiative#8. World Bank Group. 2019. “Profiting from Parity: Unlocking the Potential of Women’s Business in Africa.” World Bank, Washington, DC. World Bank. 2020. “Top Policy Lessons in Agriculture.” Gender Innovation Lab, World Bank, Washington, DC. 112