Profiting from Parity Unlocking the Potential of Women’s Businesses Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa This report was produced jointly by the World Bank’s Africa Region Gender Innovation Lab (GIL) and the Finance Competitiveness & in Africa Innovation (FCI) Global Practice. The World Bank’s Africa Region Gender Innovation Lab (GIL) conducts rigorous research to support the design of innovative, scalable interventions to address gender inequality across Africa. The Finance, Competitiveness & Innovation Global Practice (FCI) combines expertise in the financial sector with expertise in private sector development to foster private-sector led growth and help create markets in client countries. in Africa Businesses of Women’s the Potential Unlocking Parity from Profiting Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Table of Contents 4 Acknowledgments 5 Foreword 6 Executive Summary 26 Introduction 44 Part 1: Who is a female entrepreneur in Africa? 52 Part 2: Gender gap explained 62 Part 3: Decisions and constraints: A blueprint for closing the gender gap in business performance 63 Introduction Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 65 Strategic decisions 82 Constraints 100 Part 4: Deep dives 128 Part 5: Path forward 132 Overview of existing programs 135 What works and why 164 Designing programs to support female entrepreneurs 170 Expanding the knowledge base 176 Appendix 186 References 197 Endnotes Acknowledgments Foreword The Profiting from Parity: Unlocking the Sub-Saharan Africa has the highest rate psychology to encourage women to act with Potential of Women’s Businesses in Africa of entrepreneurship in the world, with an entrepreneurial mindset; secure savings Flagship Report is a joint production of the approximately 42% of the non-agricultural labor mechanisms that provide a layer of privacy and World Bank’s Africa Region Gender Innovation force classified as self-employed or employers. security in the management of funds; and large Lab and the Finance, Competitiveness and Yet most entrepreneurs are unable to grow their grants as part of business plan competitions to Innovation (FCI) Global Practice. It was prepared businesses beyond small-scale subsistence address capital constraints of growth-oriented by a team led by Francisco Campos, and operations, impeding their contribution to firms. included Rachel Coleman, Adriana Conconi, poverty reduction and shared prosperity. Aletheia Donald, Marine Gassier, Markus As part of its Maximizing Finance for Goldstein, Zenaida Hernandez, Joanna This is particularly so for women. As Development (MFD) approach, the World Bank Mikulski, Annamaria Milazzo, Maliheh Paryavi, governments continue their efforts to improve Group continues to support policymakers in Rachael Pierotti, Michael O’Sullivan, and Julia the conditions for private investment, a focus their efforts to improve the business enabling Vaillant. on the specific constraints faced by women environment and crowd in private sector entrepreneurs can help maximize the returns development support. However, we also The team gratefully acknowledges the guidance to limited public resources and ensure the recognize that the results of these efforts will and leadership of Markus Goldstein, Head of benefits of private sector development are felt be limited if they are gender-neutral and are not the Africa Gender Innovation Lab. The team by everyone. informed by evidence on what works to support thanks the leadership and support from FCI women entrepreneurs to profitably run and grow management, including David Bridgman, Dahlia Women play a key role in the economies of their businesses. Khalifa, Douglas Pearce, Klaus Tilmes, and sub-Saharan Africa. In fact, sub-Saharan Sebastian-A Molineus. The peer reviewers were Africa is the only region where women make up We are hopeful that the findings of this report Benedicte De La Briere, Steven Dimitriyev, the majority of those who are entrepreneurs. will drive action to ensure that women can David McKenzie and Noa Gimelli. Administrative However, a range of impediments render fully access the benefits of improved business support was provided by Nenette Santero. women’s businesses less productive and having opportunities. The World Bank Group will strive fewer employees than those owned by men. to operationalize this evidence in policy dialogue The team thanks Joanna Mikulski for guidance and program design with our clients and on narrative, and Aline Menden, Fabio Venturini, This new report, “Profiting from Parity: partners. The continent is full of promise and, by and Barbara Serfozo for project management Unlocking the Potential of Women’s Businesses addressing the difficulties women entrepreneurs and editing. The report’s design was guided in Africa”, produced by the World Bank Group’s face, we can unlock the potential of women’s by Vito Raimondi. The team is also grateful to Africa Gender Innovation Lab and the Finance, businesses to become a major driver of economic Endeva, Assemblyfor and Without Violence for Competitiveness and Innovation Global Practice, growth and poverty reduction. their organizational support to this project. seeks to focus attention on the challenges that All photos included in this report are by Stephan Africa’s women entrepreneurs face and identify Gladieu. The team is grateful for Stephan’s practical solutions. dedication to capturing inspiring portraits of women entrepreneurs across the continent. The report draws on new, high-quality, household and firm level data to present the The World Bank team would like to clearest evidence to date about the barriers Hafez Ghanem Ceyla Pazarbasioglu acknowledge the generous support of the to growth and profitability faced by women Umbrella Facility for Gender Equality (UFGE), entrepreneurs. It goes beyond looking at Vice President, Vice President, a multi-donor trust fund administered by contextual, endowment and household Africa Equitable Growth, the World Bank to advance gender equality restrictions in isolation, and, through deep-dive Finance and and women’s empowerment through analysis, uncovers new evidence on how social Institutions experimentation and knowledge creation to norms, networks and household-level decision help governments and the private sector focus making contribute to business performance. It policy and programs on scalable solutions analyzes how they are linked to each other and with sustainable outcomes. The UFGE is to women’s strategic business decisions. supported with generous contributions from Australia, Canada, Denmark, Finland, Germany, The report offers policy makers evidence- Iceland, Latvia, the Netherlands, Norway, based guidance on designing programs to Spain, Sweden, Switzerland, United Kingdom, target multiple obstacles and improve the United States, and the Bill and Melinda Gates performance of women entrepreneurs. The Foundation. policies recommended by the report include training programs that apply lessons from 4 5 Female Increasingly, national government leaders and other stakeholders across Africa are recognizing that women are a force for growth – but could be even more so. Policymakers need to act to expand opportunities for female entrepreneurs to be agents of growth and job creation - particularly in the entrepreneurs context of a large youth population with high expectations for quality employment. In Africa, the performance of female-owned businesses consistently lags behind that of male-owned in Africa: businesses. They have fewer employees, lower average sales, and less value-added.4 Drawing on survey data from 14 countries,5 this report finds wide gaps in average profits between male- and female-owned firms (Figure 1). A force for growth Figure 1 Gender gap in profits wide on average, but varies across countries Gender gaps in monthly profits (%) Summary Executive WOMEN MEN Average 34% Benin micro 12% DRC census 49%*** Enterprise development is a crucial engine of Ethiopia manufacturing 45% economic growth and jobs creation. Without entrepreneurship, there would be little innovation, Ghana (1) micro 36%* Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa little productivity growth, and few new jobs.1 Over three-quarters of the African population believe Ghana (2) tailoring 20% that entrepreneurs are admired in their societies. Entrepreneurship is also seen by 76% of Africans Ghana (3) manufacturing 82%*** as a good career choice. This is the highest rate in the world.2 Malawi micro 31%*** Advancing gender equality is smart economics, Mozambique SMEs 16% sound business practice, and essential development policy. When women and men have Nigeria (4) SMEs 52%** equal opportunities to shape their own lives and contribute to their families, communities, Nigeria (5) SMEs 8% and countries, it leads to enhanced productivity, improved development outcomes, and better South Africa SMEs 65% performance by businesses and institutions.3 Togo micro -7% The countries in sub-Saharan Africa (referred to as Africa in this report) have already made significant Uganda (6) micro 30% progress in fostering the economic empowerment of women and girls. Women in Africa are more Uganda (7) micro 31% likely to be working than women in other regions, and almost 50% of women in the non-agricultural Source: Authors using IE database. Ghana (1): Grants for micro-enterprises survey; Ghana (2): Tailoring survey; Ghana (3): Enterprise Census; labor force are entrepreneurs. It is the only Nigeria (4): Growth and employment survey; Nigeria (5): Business plan competition survey; Uganda (6): Kassida survey; Uganda (7): Loans, grants, region in which women are more likely to be and training survey. entrepreneurs than men. 7 For countries seeking to harness the full potential of female entrepreneurs, the central question is why female-owned businesses underperform. The report draws on a wealth of new, high-quality What’s driving the gender gap household and firm-level data in Africa to paint a more detailed picture of female entrepreneurs and the constraints to growth and profitability they face. It first explores the strategic decisions of female in business performance? entrepreneurs, and then explores the underlying constraints. It presents new deep-dive analysis in three areas: social norms, networks, and household-level constraints. Finally, based on the findings Wage job opportunities are limited in Africa. This drives men and women who and prior research, it makes recommendations for policymakers and other stakeholders on how to might not otherwise be inclined to start a business to become entrepreneurs. close the gap. According to the Global Entrepreneurship Monitor, Africa features the world’s lowest share of entrepreneurs who started a business in order to pursue an opportunity.7 Many women who become entrepreneurs out of economic necessity do not intend or have the skills to build large and successful companies. Their decision to start a business instead of seeking wage work is influenced by important constraints such as differences in skills, capital, networks, time and Box 1 family formation, occupational opportunities, and safety.8 What’s new about this report? It is important to note that if women face greater constraints than men in pursuing alternative job opportunities, this can lead to a relatively higher share 1 2 3 of women taking up self-employment, which can affect their opportunities. Recent evidence from Ghana suggests that self-employed women operate in more crowded markets than do self-employed men.9 The gaps in economic opportunities are a primary, and significant, driver of the gender gap in business performance. Updated analysis: Uses Deeper understanding: Evidence-informed solutions: Furthermore, this report examines the status of already existing businesses new high-quality data to Presents a novel way Offers policymakers and which, in the absence of alternative labor-market opportunities, are likely update earlier analyses of of understanding other stakeholders research- to persist in Africa. The report also examines the conditions influencing the gender gap in business how gender-specific based guidance on designing female entrepreneurs’ decision-making, in particular those that perpetuate performance. constraints affect female programs and policies to their lower performance rates. Women overwhelmingly choose to enter entrepreneurs’ decisions improve the performance of sectors with reduced opportunities for growth; they also have lower levels and outcomes. female entrepreneurs. of available assets and capital to invest into their businesses; and they show less willingness to compete. They are more likely to operate in the informal economy and less likely to adopt advanced business practices. Drawing on extensive data analysis and earlier research, this report argues that women make or are obliged to make different decisions than men because Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa they are constrained by gender-specific factors that hinder the growth of their businesses. These constraints, related to the contexts in which women operate, their endowments, and household-related factors, influence the strategic decisions that female entrepreneurs make – which, in turn, lead to less productive outcomes. This report presents a new, clear and illuminating guide to help further explain the factors that give rise to the gender gaps in productivity (Figure 2). Executive Summary 8 9 Figure 2 Strategic decisions Constraints, decisions, and outcomes: As illustrated in Figure 2, female entrepreneurs’ strategic decisions in key areas diverge from those of A blueprint for closing the gender gap in business performance male entrepreneurs in important ways. These areas include: 1. the sector they choose; in Africa 2. differences in capital and labor; 3. differences in firm capabilities; and, 4. differences in their willingness to compete. OUTCOMES 1 Gender Sector of operations differences in firm performance • Female entrepreneurs build larger and more profitable companies when they operate in male-dominated sectors. While female (sales, profits) entrepreneurs tend to cluster in sectors dominated by other women, such as retail and hospitality, evidence suggests a potential dividend to “crossing over” into male-dominated sectors. Female-owned enterprises operating in male-dominated industries are as large and just as STRATEGIC DECISIONS profitable as their male-owned counterparts. They are also larger than those in female-dominated sectors.10 Studies in specific settings suggest that a female entrepreneur’s decision to work in higher-return sectors 1 2 3 is not driven by differential access to education or finance but by social 4 factors, particularly the influence of male role models and exposure to the sector by family and friends.11 Sector Differences Differences in Differences segregation in capital firm capabilities in and labor (business willingness 2 practices / to compete Capital and labor innovation), and formalization • Female entrepreneurs in Africa have systematically lower levels of business capital – including equipment, inventory and property – relative to their male peers. Drawing on data from 14 impact-evaluation datasets from 10 countries in Africa, the typical male-owned firm has Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa over six times the capital investment of female-owned enterprises. UNDERLYING CONSTRAINTS • Female entrepreneurs’ lower levels of labor contribute to gender productivity gaps in six of the 10 countries analyzed. Relative to male entrepreneurs, female entrepreneurs employ fewer workers and use fewer labor hours in operating their businesses – but in most countries, CONTEXTUAL FACTORS women obtain returns on their business labor that are on par with those obtained by men. I Legal discrimination ENDOWMENTS HOUSEHOLD-LEVEL • The gender differences in capital and labor are associated with men CONSTRAINTS having higher profits than women. Although it may also be the case that II Social norms male entrepreneurs use more capital and labor due to gender differences Allocation of factors IV Education/Skills VIII in returns to the initial units of investment, the analyses in this report of Production III Risk of GBV suggest that providing women with the same amounts of investment Confidence/Risk resources as men is promising. V IX Time constraints/Care preferences Executive Summary VI Finance and assets Networks and VII information 10 11 Firm capabilities: Business practices, innovation, Nine factors holding back women’s 3 and formalization • Differences in the adoption of good business practices help explain the business performance in Africa gender gap in productivity for some countries. Consistent with earlier research,12 this report finds that male entrepreneurs are more likely to Female entrepreneurs do not make business decisions in a vacuum. Rather, their business decisions adopt advanced business practices in most countries analyzed. However, differ systematically from those of male entrepreneurs because they are constrained by factors in a the magnitude of the gender gaps on an index of business practices is on way that men are not. This report presents nine underlying constraints and explores the evidence on average less than half of the capital investment gaps. why the factor matters and the extent to which it contributes to the gap in business performance. In some cases, the report recommends further analytical work to allow for a deeper understanding of • This report shows some differences in innovation between male and a constraint’s relevance. The report leverages new data to dive deep on three factors – adding new female-owned enterprises. Women in Togo are half as likely as men insights for policymakers and other stakeholders looking to close the gap. to have introduced a new product in their operation. In Mozambique, women are less likely to have introduced a new process – 31% of women and 39% of men have taken innovative steps in the past 12 months. In Contextual factors Nigeria, female-owned firms are 20% less likely than male-owned firms to have improved existing products or to have introduced a new product I design. Legal discrimination: Female entrepreneurs cannot have equal economic opportunity if a country’s laws restrict a woman’s ability to own and run • This report’s analysis does not find a clear pattern between individual a business. Women often face barriers from customary law.17 In addition, business formalization and the gender gap in enterprise performance. formal laws still do not ensure a level playing field. Although many African Many female-owned firms are in the informal economy. While this countries have made progress in removing legal barriers – including laws report’s analysis finds that female-owned businesses enjoy higher that deny women the same rights as men to register a business, sign a productivity returns when formal in three countries, recent studies13,14 contract, open a bank account, or own and inherit property – only three indicate a lack of impact on business performance and access to finance African countries have formal laws that prohibit gender discrimination.18 from helping entrepreneurs obtain formal status. Bottom line: Recent progress in regulatory reforms means legal discrimination may not be constraining female entrepreneurs as much 4 as in the past. However, even when laws are gender neutral, they are not Decision to compete applied equally. Work is needed to ensure gender-equal laws are both introduced and appropriately implemented. • A lab experiment in Kenya15 finds that women were less than half II as likely as men to compete with others. Numerous laboratory Social Norms: Social norms exert strong influence over the strategic experiments from outside of Africa confirm this finding. However, limited choices that female entrepreneurs make and can constrain their ability data on willingness to compete makes it difficult to weigh the importance to grow their businesses. They can shape how women view themselves,19 of gender differences in entering competition relative to other drivers of perceive their abilities,20 impact their aspirations, and can lead to the gender gap in business performance.16 discriminatory treatment by others.21 Because social norms in many African Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa countries do not align with a woman striving for business growth, women’s choices are limited. Women who contradict social norms may also face retaliation.22,23 Bottom line: The influence of gender norms on women’s business decisions is likely to be pervasive and significant, affecting major choices such as the sector of operations. For more on this, see “deep dive” 1. III Risk of gender-based violence: Widespread gender-based violence (GBV) likely takes a toll on women’s health and well-being, which hinders their ability to run their businesses effectively. Working outside of the home may put women at risk,24 while some women may view self-employment as a way to avoid sexual harassment at the workplace.25 In Malawi, 14% of female entrepreneurs have been subject to physical or emotional violence by their domestic partner; 32% say their husband insists on knowing where they are at all times; and 7% say their husbands force them to have sexual intercourse. Executive Summary Bottom line: GBV can hinder women’s managerial capacity. Testing and evaluating more solutions to overcome GBV is critical to understanding its relevance in addressing the business-performance gender gap. Research is also needed to determine whether the risks of being exposed to GBV influences a woman’s tendency to shy away from growing her business. 12 13 Endowments Household-level constraints Education and skills gaps: While most African countries have achieved Household allocation of productive resources: Women often lack authority IV VIII gender parity in access to primary education,26 a persistent gap in over the allocation of household assets and may face more pressure educational and skill attainment between male and female entrepreneurs to share resources, which restricts both their willingness and ability to – particularly at the secondary level and beyond – may help explain gender invest in their businesses. Women’s lack of control over the allocation of differences in strategic business decisions. Evidence points to gaps household resources may be a source of inefficiency if it means that assets between male and female entrepreneurs in three areas: formal education, are invested in male-supported enterprises irrespective of managerial management skills, and socio-emotional skills. This report finds that ability or the value of the business opportunity.32 Research shows that self-employed women have overall completed fewer years of education female entrepreneurs struggle to direct capital to their business, which than self-employed men. Male entrepreneurs often have higher technical can be a function of either their own or others’ needs. Inefficiency in skills; sometimes have higher financial literacy; and are sometimes more intra-household allocation is compounded when female entrepreneurs are likely to participate in training or offer training. Data from Togo shows that compelled to share resources derived from social connections outside of while male and female entrepreneurs are comparable on some important the home. socio-emotional measures, male entrepreneurs score higher than their female counterparts on measures of ambition, creativity, innovation, and Bottom line: Household resource allocation is likely a major factor imagination. influencing women’s decisions regarding their businesses. The next stage of research should include identifying scalable mechanisms to encourage Bottom line: Education and skill gaps in Africa are wide and persistent, and couples to think differently about the role of women’s contribution to likely have a strong influence on women’s business decisions. household decision-making, or advocating that female entrepreneurs simultaneously achieve both their business and non-business goals. For Confidence and risk preferences: Women business owners in Africa more on this, see “deep dive” 3. V IX frequently show less confidence than their male counterparts. Among entrepreneurs in Ghana, women are 14% less likely than men to think Time constraints and care: Women in Africa spend more time than men they would make a good leader. Female entrepreneurs demonstrate less on domestic chores.33 This limits the amount of time they can dedicate confidence in their abilities,27 which may make them less willing to compete to their business and requires them to stay home at times of the day that (and win)28 – especially in stereotypically male domains.29 Women’s lack of are best for conducting business. This report finds that women in Uganda, confidence relative to men could be related to risk aversion, but analysis for Togo, and Malawi are much more likely than men to be taking care of this report do not show a clear pattern on this issue. others while running their business – a task that can take up to twice as much of their time as men’s.34 This report finds that men spend on average Bottom line: Women’s lack of confidence relative to men may keep them 10% more time per week working in their business than women. Being from taking big risks that lead to high returns. More experimental work married increases the gender gap in time spent on the business in three on mechanisms designed to enhance confidence can prove important for countries, while the gap is lower in male-dominated sectors where women policy. may have to work the same hours as men to participate. Emerging evidence suggests that childcare programs may have a positive impact on women’s Finance and assets: Female entrepreneurs continue to control fewer assets employment outcomes,35 but these studies do not analyze the impact of VI Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa than men, affecting their capacity to invest in their business and access childcare on women’s business outcomes. large enough loans. While the gender gap in obtaining loans from financial institutions is smaller in Africa than in any other region of the world,30 this Bottom line: Women’s time constraints associated with domestic chores report’s analysis shows consistent and large gender gaps in the size of the and care are likely a strong constraint on their business activities, but loans outstanding for various target groups of entrepreneurs in Africa. further rigorous evidence and analysis is needed on whether increasing childcare services and other types of interventions produce lasting effects Bottom line: The gender divide in access to credit is not as strong as it once on business performance. was, but with smaller asset ownership, women still struggle to get loans of the same size as men – a factor that likely fuels the capital investment gap. Access to networks and information: Women often do not have the same access as men to large and diverse social networks that can support VII the growth and competitiveness of their business. This report’s analysis suggests that men’s and women’s networks vary in important ways. Both men’s and women’s networks are largely segregated by gender. Women’s networks command fewer resources than men’s and include more “strong” family and kin relationships that are less valuable than new connections in creating business opportunities.31 Executive Summary Bottom line: With growing evidence on the importance of networks, understanding how the networks of female entrepreneurs vary from those of men – and how those differences may impact their success – is vital. For more on this, see “deep dive” 2. 14 15 Diving deep on three constraints  Deep dive 1: Do gender-biased social norms dampen business performance? Drawing on data from Togo, Ghana, and Malawi, this report presents a deep-dive analysis assessing the prevalence of gender-biased beliefs and their potential contributions to the gender gap in business performance in Africa.36 Key findings 1. Gender-biased beliefs are prevalent among surveyed entrepreneurs in Africa. 2. Both men and women have largely internalized gender-biased norms,37 though there are some important gender differences. Women are less likely than men to be biased toward the pursuit of business opportunities, but they are more likely to prioritize household needs. 3. Entrepreneurs with comparatively high levels of education are less likely to espouse gender-biased beliefs. 4. Holding gender biases is associated with lower investment levels and lower business performance, but this relationship is not robust in all settings. There is suggestive evidence in some settings that businesswomen that hold more progressive views and operate in male- dominated sectors face discrimination in their operations. Way forward for combined research and policy • A research agenda is needed to test the mechanisms and importance of social norms as a constraint to business development. Solutions to tackle embedded social norms in enterprise development can be drawn from a number of promising areas of research on norms, including: large-scale institutional changes,38 changes in regulations,39 addressing mobility issues,40 overcoming sector sex-segregation issues,41 and community and couple dialogues.42 Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa  Deep dive 2: Do gender differences in entrepreneurial networks fuel the gender performance gap? This deep-dive analysis draws on data from Ghana, Malawi, Togo, and Uganda to assess gender differences in African entrepreneurs’ business networks and their potential contributions to the gender gap in business performance. Key findings 1. Female entrepreneurs’ business networks are mostly comprised of other women. 2. Men have larger business networks than women. 3. Business networks with “strong ties,” and family members in particular, are more likely to be part of the process of business creation and development for women than for men. Executive Summary 4. Women rely on their networks when starting a business and for financial support, but men more intensively use their business networks to share information, equipment, and supplies. 5. The relationship between networks and business performance is likely not straightforward and depends on aspects such as the depth of the relationships and their influence on various areas of business development. 16 17 Way forward for combined research and policy Path forward: How policy and decision makers can act to • Additional experiments creating networking opportunities for different categories of entrepreneurs would be helpful in identifying the type of businesses who stand to benefit the most from diversifying their connections. Is networking more important in early or later stages of business development? To what extent do other gendered constraints eliminate the gender gap dampen the benefits of expanding one’s networks? Why do business networks remain so strongly segregated by gender? Are there opportunities to loosen the networks’ “ties”? What is the effect of network diversification on entrepreneurs’ choice of activity?  Deep dive 3: Do intra-household relationships affect the strategic choices made by female entrepreneurs? This deep dive draws on an analysis of micro-entrepreneurs in urban Ghana, in-depth qualitative research with the same population to explore how women’s businesses contribute to the household, and of how household demands and power dynamics impact their business decisions. Though specific to urban Ghana, the findings suggest ways that intra-household dynamics might be influential in women’s business decisions elsewhere. Gender-neutral policies that seek to create a designed to support firms, innovation, and more conducive business environment while investment in Africa. Whenever applicable, the supporting entrepreneurs will not be sufficient solutions recommended in this report are also Key findings to address the constraints identified in this sound for male-owned firms. Implementing them 1. Women’s businesses are important for meeting household needs. report. Indeed, they may instead widen existing at scale with thorough mechanisms of delivery 2. Spouses have incomplete information about each other’s earnings. gaps. But the good news is that with targeted can support female entrepreneurship while also 3. Women generally maintain control of their business income, but that strategies, the international community, national expanding all types of businesses in the economy. does not always mean they have flexibility in determining how to spend governments, NGOs, and other stakeholders it. can help address the challenges that female  Promising strategies 4. Women and men have incentives to hide their income due to pressure entrepreneurs face and thereby unleash their to share it and to increase contributions to household needs. productive potential. The best evidence today suggests that these 5. Women’s independence in business management is associated three strategies are likely to be effective for with higher profits, but it is still unknown whether this is a causal This report reviews the existing evidence, female entrepreneurs: relationship. classifies strategies based on the strength of the 6. Households manage their finances in a range of different ways and evidence, and makes recommendations based 1. Training programs that apply lessons from future research and programs should explore links between intra- on the evidence. The results are summarized in psychology to encourage women to act with an household relationships and the success of women’s businesses. Table 1. As much as possible, the report identifies entrepreneurial mindset. Emerging evidence success factors, potential risks, and mitigating from impact evaluations demonstrates the Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa measures. This evidence is typically based on importance of strengthening socio-emotional Way forward for combined research and policy a set of countries and generalizing to others skills for female entrepreneurs in Africa. In is not necessarily straightforward. Therefore, Togo, personal initiative training has positive and • Future research should examine the extent to which different financial local adaptation will be important as a means significant effects on sales and profits of male management practices within households can be encouraged and lead of ensuring sound implementation of the most and female-led micro-enterprises and generated to business investments by female entrepreneurs. Additional research promising ideas. a 91% return on investment.43 Similarly, a is also needed to identify possible policy responses either to think training program emphasizing self-esteem differently about the role of women’s businesses in the household, or The evidence presented regarding the most and entrepreneurship in Ethiopia increased the to support female entrepreneurs to simultaneously achieve both their promising interventions is typically based on business performance of female-owned firms.44 business and non-business goals. the results for the average female-owned A video to increase aspirations of market vendors businesses. However, there are often female in Mozambique (45% of them women) led to entrepreneurs who can benefit more than others very large impacts on business performance.45 from each viable or emerging solution. Finally, a training program for female micro- entrepreneurs in Kenya led to sustained While the set of constraints and solutions are increases in firm profits, survival, and growth.46 listed individually, female entrepreneurs in Africa are affected by the interplay of several Summing It up: Training programs addressing Executive Summary gender-specific constraints. Increasingly, socio-emotional skills and gender-specific programs for female entrepreneurs recognize content – as opposed to standard managerial this and combine interventions targeting multiple training programs – have proven effective constraints. in numerous contexts in Africa, and pay for themselves in increased profits over the long- The strategies proposed are to a large extent term. a subset of the most promising interventions 18 19 2. Supporting women with secure savings In addition, the following set of policies and mechanisms. Unequal bargaining power within interventions that draw on emerging evidence the household and domestic expenditure needs offer strong potential and would benefit from can affect women’s ability to finance their further assessment of impact: business activities. Therefore, providing women with mechanisms to set aside money for their 1. Removing legal constraints to gender business can help insulate these funds from equality and regulatory implementation household demands. Providing female market gaps; vendors in Kenya with access to savings accounts led to a 45% increase in business investment, 2. Strengthening land tenure rights for while no impact was found in providing such women; accounts to male motorbike drivers.47 Targeting seems important, as providing these accounts 3. Expanding women’s linkages to new in rural settings in Uganda and Malawi led to business networks; limited bank account usage.48 On the other hand, adding access to business bank accounts to 4. Offering women-friendly training support formalization led to significant increases designs, including peer support; in women’s usage of business bank accounts and insurance, and it also resulted in more 5. Providing in-kind grants to female- women separating household and business owned firms; funds.49 This drove large impacts on sales and profits for female entrepreneurs.50 Mobile money 6. Introducing financial innovations and other digital payments make it easier to that reduce collateral requirements, target women specifically and provides them including psychometric scoring; with greater privacy and control over household expenditures.51 In turn, this leads to potential 7. Facilitating access to childcare services; increases in women’s consumption and savings,52 and supports new economic activities 8. Engaging men to provide a more and empowerment.53 supportive environment for female entrepreneurs; Summing It up: Rigorous evidence from more than one study in Africa shows the positive 9. Incentivizing women to cross over to effects of savings mechanisms on business male-dominated sectors by sharing investment and performance of female-owned information on expected returns in those firms. The interventions need to be well-targeted sectors, and through early exposure in to ensure appropriate uptake from the relevant the form of apprenticeships and male target group. role models. Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 3. Providing large cash grants to female- owned businesses as part of business plan competitions. Providing large cash grants as part of a business plan competition can help address the capital constraints of growth- oriented firms, including those owned by women. A business plan competition in Nigeria providing cash grants averaging $50,000 increased the likelihood that women would operate a firm by reducing capital constraints. It also helped trigger hiring and led to large increases on sales and profits.54 This is backed up by another study of business plan competitions in Ethiopia, Tanzania, and Zambia.55 Additional impact evaluations of business plan competitions are under way in Africa, and will address how results are influenced by the size of different grants and Executive Summary the selection process. Summing It up: Studies show positive impacts of large grants under business plan competitions on employment, sales, and profits among female-owned firms. 20 21 Table 1 LEGEND CREDIBLE EVIDENCE OF POSITIVE IMPACT ON BUSINESS OUTCOMES What works to support female-owned firms in Africa? EMERGING EVIDENCE OF IMPACT ON BUSINESS OUTCOMES EVIDENCE OF NO/LOW IMPACT ON BUSINESS OUTCOMES (NOT PROMISING) Key findings from rigorous impact evaluations POLICY AREA CONSTRAINT ADDRESSED TYPOLOGY OF FIRMS MAIN CONCLUSIONS All firms Removing legal gender biases and gaps in the implementation of laws increase women’s agency Legal discrimination and intra-household bargaining power 1. Removing regulatory and institutional Legal discrimination All firms Strengthening land rights for women increases their time and effort in entrepreneurship constraints for female entrepreneurs Easing constraints to formalization by itself is not sufficient to help female-led micro-enterprises Informality Micro-enterprises grow Providing traditional managerial training alone does not typically improve the business Skills Micro-enterprises performance of small female-owned firms Skills; confidence/risk preferences; Micro-enterprises and Training addressing socio-emotional skills and gender-specific content leads to high levels of social norms small-business owners impact on business performance 2. Improving skills and Micro-enterprises and Expanding firms’ access to new networks may, in the right settings, have positive impacts on Networks and information networks small-business owners business performance Providing mentoring on top of traditional business training has limited additional value to micro- Skills; networks and information Micro-enterprises entrepreneurs Skills; confidence/risk preferences; Micro-enterprises Complementing delivery of training programs with direct peer support may be promising social norms Finance and assets Micro-enterprises Microcredit has only limited effects on business outcomes for women Finance and assets; allocation of factors Providing in-kind grants can lead to higher profits for more successful women micro- Micro-enterprises of production entrepreneurs Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 3. Improving access to Finance and assets; confidence/risk Start-up or existing Large cash grants for growth-oriented firms selected through a business plan competition can capital and assets preferences; skills businesses help overcome capital constraints for women Finance and assets; allocation of factors Providing women with access to secure mechanisms for savings – including mobile savings – can Micro-enterprises of production increase business investment Alternative credit scoring technologies using psychometric tests offer the promise of easing Finance and assets Micro-enterprises women’s access to larger business loans Time constraints/care All Providing childcare can increase female participation in the workforce 4. Easing household constraints GBV; time constraints/care; allocation Micro-enterprises Engaging men can potentially foster a more supportive environment for female entrepreneurs of factors of production 5. Addressing social Providing information on earnings in traditionally male-dominated sectors and early exposure norms regarding Skills; networks and information; Executive Summary Young entrepreneurs through apprenticeships and male role models can encourage female entrepreneurs to enter women’s occupational confidence/risk preference these sectors decisions 6. Facilitating access to Training does not eliminate harassment by guards at border crossings, but can make female Social norms; GBV Micro-enterprises markets traders aware of ways to minimize harassment 22 23  Challenges in designing effective programs  Steps for development partners, for female entrepreneurs corporations, and civil society When designing programs for female 1. Provide funding to test innovative approaches entrepreneurs, policymakers and other and to research that contributes to closing stakeholders need to consider the following the gender entrepreneurship gap while factors: expanding global knowledge of what works. 2. Use findings from this report to inform the 1. Getting targeting right: Growth-oriented design of programs. female entrepreneurs often benefit the 3. Continue advocacy efforts by highlighting most from skills-development programs, the business and economic case for but will only participate if they understand removing the obstacles constraining female the value-added. Some interventions can entrepreneurs’ growth. only be successful for groups of female 4. For larger firms, consider the gender entrepreneurs with access to satisfactory dimensions in their supply strategies and infrastructure such as roads, energy, and opportunities to integrate female-owned communications. firms. 2. Building gender-sensitive aspects of program design: Programs need to address the constraints discussed in the report, such as women’s lower willingness to compete and greater constraints on time.  Steps for African policymakers 1. Support concrete policy actions that demonstrate strong commitment to female entrepreneurs, such as eliminating existing legal barriers and fostering women’s participation in public life, including the promotion of female role models. 2. Ensure private sector development strategies and policies include a gender focus that addresses the specific challenges faced by female entrepreneurs. 3. Scale up policies that have shown credible results. 4. Support the testing and evaluation of Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa promising approaches – and share the findings widely. 5. Invest in promoting the systematic collection of gender-differentiated data that captures the performance, endowments, and preferences of female-owned firms. 6. Involve men in policy advocacy and in the implementation of solutions. Because men are husbands, but are also more likely to be bankers, inspectors, trainers, and policymakers, it is important to engage them at multiple levels in efforts to provide better opportunities for female entrepreneurs. Executive Summary 24 25 Unleashing the hidden potential of female entrepreneurs in Africa Introduction Advancing gender equality is smart economics, sound business practice, and essential development policy. A growing body of evidence demonstrates that when women and men have equal opportunities to shape their own lives and contribute to Following increases in access to infrastructure, communications, education, and health systems, their families, communities, and young people are demanding jobs. Ensuring that half the population has appropriate opportunities to Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa become employers and create such jobs is critical for the long-term political and economic well-being countries, it leads to enhanced of countries in Africa. productivity, improved development Enterprise development is a crucial engine of economic growth and job creation. Without outcomes, and better performance entrepreneurship, there would be little innovation, little productivity growth, and few new jobs.57 Over by businesses and institutions. three-quarters of the African population believes that entrepreneurs are admired in their societies. Entrepreneurship is also seen by 76% of Africans as a good career choice. This is the highest rate in Increasingly, national government the world.58 The top 10 African countries in the Global Entrepreneurship Index are Botswana, South Africa, Namibia, Gabon, Swaziland, Rwanda, Ghana, Nigeria, Zambia and Senegal. leaders and other stakeholders across Africa are recognizing that Over the past five years, there has been a tenfold increase across the region in the supply of new entrepreneurship-support solutions such as incubators, accelerators and tech hubs, which collectively female entrepreneurs are a force provide a wide range of services for budding entrepreneurs and startups. These intermediaries for growth – but could be even offer the opportunity to decentralize the implementation of entrepreneurship initiatives. Today more so. Policymakers need to act numbering more than 200 across the region as a whole, these entrepreneurship intermediaries are multifunctional one-stop shops that connect nascent startups with key investors and raise the overall in order to expand opportunities profile of entrepreneurship in the eyes of the public.59 for women as agents of growth Entrepreneurs in Africa are different than the average member of the population. This report’s analysis and job creation, particularly in of data from almost 1 million Africans across the region, 30% of which are entrepreneurs, show that the context of a developing young business owners are on average three years older and more likely to be married – especially in the population in Africa with high case of men – than the rest of the adult population (Table 2). Entrepreneurs have lower average levels of education than non-entrepreneurs, although the differences are small in magnitude. Business expectations in terms of finding owners typically work more hours. They are less likely to live in households with access to water, quality employment.56 electricity, bathrooms, and computers. 26 27 Table 2 Female entrepreneurs in Female and male entrepreneurs show some differences relative to the rest of the population Africa can drive growth Sub-Saharan Africa (referred to as Africa in this report) has already made significant progress in Sub-Saharan Africa fostering the economic empowerment of women and girls. Women in Africa are more likely to be in the workforce than is the case in any other region in the developing world (Figure 3). Half of Africa’s Female Female non- Male Male non- working women are active in the agricultural sector, 38% work in identified non-agriculture sectors, Variables entrepreneurs entrepreneurs Difference entrepreneurs entrepreneurs Difference and the remainder are in undefined sectors or unemployed. mean mean mean mean INDIVIDUAL CHARACTERISTICS Age 38.8 36.0 2.8*** 39.8 36.9 2.9*** Figure 3 Married 70.3 58.4 12.0*** 76.9 49.5 27.4*** Women in Africa more likely to be in labor force Ever attended school 62.3 63.6 -1.3*** 76.3 80.9 -4.6*** than women in other regions Literate 52.3 55.8 -3.5*** 71.8 76.5 -4.7*** Years of education 4.2 5.1 -0.8*** 5.4 7.4 -2.0*** Labor force participation by region HOUSEHOLD CHARACTERISTICS Rural area 62.5 62.8 -0.4** 66.7 59.8 6.8*** Female AFR House owner 75.9 76.9 -1.0*** 79.4 74.2 5.2*** Male Water 9.0 13.7 -4.7*** 8.4 14.6 -6.2*** MENA AFR= Africa Electricity 35.2 39.7 -4.5*** 28.4 44.6 -16.2*** MENA= Middle East and North Africa SAR= South Asia Bathroom 6.1 9.1 -3.0*** 5.4 9.9 -4.5*** SAR ECA= Europe and Central Asia EAP= East Asia and Pacific Phone available 1.3 3.6 -2.3*** 1.6 3.6 -2.0*** LAC= Latin America and the Caribbean ECA Cell phone 62.5 62.7 -0.2 55.2 68.0 -12.7*** Computer available 5.4 6.9 -1.5*** 4.4 8.0 -3.6*** EAP LABOR CHARACTERISTICS Hours worked last week 39.1 34.6 4.6*** 41.2 40.1 1.1*** LAC Introduction: Unleashing the Hidden Potential of Women Entrepreneurs in Africa Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 0% 10% 20% 30% 40% 50% 60% 70% 80% Source: Authors using household surveys. Source: Authors using household surveys. Women’s strong non-agricultural labor-force participation rate is driven in large part by their participation in entrepreneurship. About 50% of the women who are in the labor force but not working in agriculture are entrepreneurs. Africa is the only region where women are more likely to be self- employed than men. In non-agricultural sectors, 58% of the self-employed are women, and 45% of employers are women, the highest rate alongside Europe and Central Asia. 58% 45% Self-employed Employers 28 29 Figure 4 • Lower average sales. Average sales are 38% lower in female-owned formal firms than in male- owned formal firms.65 • Less value-added.66 Enterprises run by men add approximately 38% more value than those owned Women in Africa more likely than men to be entrepreneurs by women.67 Activity of non-agricultural labor force by gender and region While these gender gaps vary widely across countries, female-owned businesses consistently perform worse than male-owned businesses (Figures 5 and 6). 100% 90% 80% Figure 5 Wide gender gap in monthly profits 70% 60% Gender gaps in monthly profits (%) Women 50% Men 40% 30% Unpaid work 20% Wage work WOMEN MEN Employer 10% Average 34% Self employed 0% Benin micro 12% AFR MENA SAR ECA EAP LAC DRC census 49%*** Source: Authors using household surveys. AFR= Africa Ethiopia manufacturing 45% MENA= Middle East and North Africa Figure 4 shows the activity among those in the labor force: SAR= South Asia in Africa, 61% of men and 55% of women participate in ECA= Europe and Central Asia Ghana (1) micro 36%*** the labor force. Taking into account this difference, women EAP= East Asia and Pacific are still more likely than men to be entrepreneurs in Africa. LAC= Latin America and the Caribbean Ghana (2) tailoring 20% Introduction: Unleashing the Hidden Potential of Women Entrepreneurs in Africa Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Ghana (3) manufacturing 82%*** Malawi micro 31%*** Mozambique SMEs 16% But gender gaps persist Nigeria (4) SMEs 52%*** in business performance Nigeria (5) SMEs 8% South Africa SMEs 65%** While female entrepreneurs are already a vital and vibrant source of economic growth in Africa, prior research60 and this report’s analysis of data Togo micro -7% from enterprise, household and impact-evaluation studies demonstrate that they have yet to fully realize their tremendous potential. Compared to male- Uganda (6) micro 30% owned businesses, female-owned businesses have: Uganda (7) micro 31% • Lower profits. In the datasets analyzed for different groups of entrepreneurs,61 firms62 run by women have profits that are on average -100% 0% 100% 34% lower than those run by men (see Figure 5).63 • Fewer employees. Firms with majority-female ownership account Source: Authors using IE database. Ghana (1): Grants for micro-enterprises survey; Ghana (2): Tailoring survey; Ghana (3): on average for 20% of firms in the formal economy with 10 or fewer Enterprise Census; Nigeria (4): Growth and employment survey; Nigeria (5): Business plan competition survey; Uganda (6): employees, but only 10% of firms with 100 to 500 employees, and 7% of Kassida survey; Uganda (7): Loans, grants, and training survey. firms with more than 500 employees64 (see Figure 6). 30 31 32 Introduction: Unleashing the Hidden Potential of Women Entrepreneurs in Africa % female majority ownership % female majority ownership 0% 0% 10% 10% 40% 40% 20% 20% 50% 50% 30% 30% Figure 6 Central Africa Rep. Lesotho Rwanda Botswana Burundi Zambia Botswana Benin Cape Verde Rwanda Ethiopia Cape Verde 0-10 Employees Zambia Madagascar Source: Authors using ES database. Namibia Mali 101-500 Employees Nigeria Namibia Cameron Côte d’Ivoire Mauritius Ghana Congo Kenya Mali Angola Sudan Burundi Kenya Liberia Sierra Leone Uganda Uganda Senegal Côte d’Ivoire Malawi Lesotho Ethiopia Angola Central Africa Rep. Burkina Faso Cameron Mauritania Nigeria Tanzania Tanzania Malawi Gabon DRC Mauritius Proportion of firms with female majority among those with 0-10 employees Madagascar Togo Proportion of firms with female majority among those with 101-500 employees Benin DRC Ghana Congo Liberia Eritrea Senegal Burkina Faso Average number of employees in Female-owned businesses Gabon South Sudan Togo Mauritania Eritrea Niger South Sudan Sierra Leone Niger Chad Chad Sudan AVERAGE: 10% AVERAGE: 20% Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 33 Despite these average differences in business size and performance, a number of Figure 8 individual enterprises owned by women perform as well as those owned by men. Some do better than many of those owned by men. There is some overlap in the distribution of profits (Figure 7). But the gender gap in profits is persistent across the profit distribution The gender gap in profits is persistent across profit distributions 69 (Figure 8). Similar patterns occur for other measures of business performance. Female-owner average profits per decile as % of male-owner profits: Thus, the main question is how to bring more female-owned enterprises to the level of performance of male-owned firms. If women and more generally Africa are to profit from parity, policies need to be well designed to eliminate the gender gaps currently Benin affecting the various levels of business development and types of firms. 200% 150% 100% 50% 0 Ghana 200% Figure 7 150% Partial overlap in distribution of profits by gender 68 100% 50% DRC Mozambique 0 Malawi 200% 150% Introduction: Unleashing the Hidden Potential of Women Entrepreneurs in Africa Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 100% 50% 0 South Africa Uganda Uganda 200% 150% 100% 50% 0 DECILE 1 DECILE 2 DECILE 3 DECILE 4 DECILE 5 DECILE 6 DECILE 7 DECILE 8 DECILE 9 DECILE 10 Source: Authors using IE database. Female owner Male owner Source: Authors using IE database. Female owner Male owner 34 35 Understanding how What to expect to close the gap from this report For countries seeking to harness female the gender gap in business performance in the This report focuses on gender gaps displayed by Part 4 presents deep dive analyses focusing entrepreneurs’ full potential, the central question interim. existing non-farm enterprises in Africa. It does specifically on three of the constraints for which is why female-owned businesses lag behind not address the agriculture sector. Moreover, research is currently weakest. male-owned firms. In that vein, this report investigates the status it does not focus per se on labor-market of existing businesses – a sector that is likely participation, including the broad area of youth Finally, Part 5 provides an overview of the First and foremost, the pool of women and men to continue to exist in Africa in the absence employment, or why women and men choose evidence to date regarding what has been who end up owning businesses may be different. of alternative labor-market opportunities to enter into entrepreneurship versus wage successful in eliminating the constraints facing Women in Africa face disadvantages as they – and examines how female entrepreneurs employment versus household work. It does not female entrepreneurs, and in helping them seek to enter the labor market.70 Their choice systematically make or are obliged to make concentrate on livelihood programs or graduation improve the performance of their businesses. It to have a business instead of engaging in wage strategic decisions that contribute to the policies. It does not make a detailed study of analyzes opportunities for the World Bank Group work is influenced by important factors such as comparatively lower performance of their property rights, or of land and financial assets and other actors to adapt efforts to support differences in skills, capital availability, networks, businesses. Women overwhelmingly choose beyond the finance provided to businesses. female entrepreneurs so as to better align time availability and family duties, constraints on to enter sectors with constrained growth These are separate areas of study pursued by the programming with the research on the decisions occupational choices, and safety.71 opportunities,75 have lower levels of available World Bank Africa Gender Innovation Lab and and constraints presented in this report. It also assets and capital to invest into their businesses76 other researchers. However, as these areas are describes areas where evidence gaps remain, It is important to note that if women face greater than men, and show less willingness to connected with enterprise development, they will and where more analytical work is required. limitations than men in regards to alternative job compete.77 They also are more likely than men to be referenced when applicable in the report. opportunities, this can lead to a relatively higher operate in the informal economy,78 and less likely Closing the gender gap in business performance share of women engaging in self-employment to adopt advanced business practices.79 This report updates previous regional and requires knowledge, innovation, and investment. which can in turn have an effect on their relative cross-regional analyses of gender differences in Although many countries and the development options. Recent evidence from Ghana suggests This report argues that female entrepreneurs firm ownership participation and performance, community have mobilized in support of women’s that self-employed women operate in more make different business decisions than men and analyzes new data collected from impact economic empowerment, investing in gender crowded markets than do self-employed men.72 because they are constrained by gender-specific evaluations and other research studies. It equality within the financial and productive Introduction: Unleashing the Hidden Potential of Women Entrepreneurs in Africa Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Women compete with a relatively large number factors that influence their strategic decisions analyzes the main contributors to the gaps sectors lags behind other priorities.80 African of other women in small markets typically and hinder the growth of their businesses. in business performance. It then outlines countries can show global leadership by making consisting of mostly female clients. the decisions and constraints facing female gender equality in enterprise development These underlying constraints are related to entrepreneurs and provides action-oriented a priority. The challenges faced by female Moreover, the pool of people who own businesses the contexts in which these firms operate, the recommendations for removing barriers and entrepreneurs in Africa are shared by women is influenced over time by imperfections in the endowments possessed by female entrepreneurs, allowing female entrepreneurs to become more throughout the world. By promoting evidence- labor market; this implies that women with and additional household-level factors. powerful drivers of economic growth. based policymaking and designing new solutions comparatively fewer employment alternatives – Understanding how these constraints influence to support female entrepreneurs, countries in that is, with limited access to wage-labor – are female entrepreneurs’ strategic decisions is key Part 1 provides a profile of female entrepreneurs Africa have a rare and unique opportunity to more likely to persist in operating subsistence- to developing effective policies and interventions in Africa using demographic data. provide a model for unleashing women’s unmet level businesses.73 This has additional to improve their business performance. To date, potential worldwide. implications with regard to the type of women there have been relatively few evidence-based Part 2 presents an analysis of the main factors (relative to men) who own businesses at any solutions aimed at closing the gender gap in observed to contribute to gender differences in given point. business performance. business performance. While labor-force participation is not the Part 3 presents an overarching blueprint focus of this report, expanding labor- indicating how women’s strategic decisions and market opportunities beyond small-scale underlying constraints impact their performance entrepreneurship could potentially help women as entrepreneurs. It then presents a detailed in self-employment. However, given Africa’s analysis of these decisions and constraints, with fast-growing young population,74 this appears to the goal of understanding how different factors be a challenging prospect in the short to medium affect the size of the gender gap in business term. If anything, women who are not at the performance. bottom of the pyramid may be able to find first wage jobs, which could potentially even increase 36 37 Box 2 Box 3 Entrepreneurship and women’s economic empowerment Understanding the data Women’s empowerment is the increased opportunity for women to make strategic life choices where such A wealth of new, high-quality household and firm-level data allows this report to paint a more detailed picture choices were previously denied to them.81 Economic empowerment is about expanding a woman’s ability to make of female entrepreneurs and their constraints to growth and profitability. Specifically, this report draws on three decisions and achieve outcomes important to her in the economic sphere. There are three central components of types of data: enterprise surveys, household surveys, and impact-evaluation surveys. Appendix 2 describes the empowerment: resources, agency, and achievements. datasets used in the report. When combined, these datasets provide rich data on female entrepreneurs and their firms. Due to differing methodologies, these surveys may sometimes reach divergent answers to the same question, such as “What percentage of female entrepreneurs are literate?” Below is an explanation of the types of surveys, their advantages and their limitations: Enterprise surveys: Household surveys: A household Impact-evaluation surveys: The World Bank Group’s survey is any survey administered Several experimental studies Enterprise Surveys, which at the household level. This assessing the effectiveness of are firm-level surveys of a report uses 38 nationally specific enterprise-development (PRE-CONDITIONS) representative sample of an representative household policies and interventions have Resources include material resources, but also human and social resources that RESOURCES economy’s private sector, surveys from the International been conducted in African enhance the ability to exercise choice. As compared with their male counterparts, can be used to assess firm Income Distribution Database countries in recent years. The female entrepreneurs generally have a lower degree of access to the key characteristics and gender (I2D2), which includes modules resulting database83 (referred resources necessary for business success. These include assets, time, networks, gaps in firm revenues. However, on labor-force participation. to in this report as the “IE skills, information, and legal rights. The relationship between the unequal they have important limitations These include data collected database”) includes surveys distribution of resources and social norms of gender inequality is a mutually with regard to analyzing female through the World Bank Living conducted by researchers from reinforcing one, with each condition exacerbating and perpetuating the other. entrepreneurs in Africa. First, Standard Measurement Studies the World Bank’s Africa Gender because they are designed to (LSMS), considered the “gold Innovation Lab; the Finance, survey formal firms with five standard” for multi-topic surveys Competitiveness and Innovation or more full-time employees, in developing countries, as well Global Practice; and the and because female-led firms as other household surveys. Development Research Group, are on average smaller than While these surveys provide as well as various academics. Agency is the ability to define one’s goals and act on them. It is a key component male-led firms, these surveys important data – including some Appendix 2 describes these of women’s empowerment, as it links underlying constraints with productivity might underrepresent female- information on the occupation datasets. While the samples are outcomes. This requires understanding three concepts: owned firms. As a result, the 36 of all household members – not nationally representative country-level Enterprise Surveys they have some limitations with and their structure varies (e.g., Goal-setting capacity: Women’s relatively more limited access to finance, time, have been supplemented in regard to the analyses pursued not all samples include male- and information can hamper their ability to set well-defined goals and develop some countries with additional by this report. Specifically, they owned firms), the types of firms strategies for achieving them. Moreover, explicit coercion from their families or surveys focusing on micro-firms lack deep firm-level data, and included in these experimental implicit gender norms may diminish their capacity to set goals in line with their with fewer than five employees focus on household enterprises studies capture the full range own values and preferences. For example, female entrepreneurs may struggle to (11 countries) and/or informal rather than corporations (e.g., of women’s entrepreneurship set clear business goals and engage in business activities that they themselves firms (13 countries). Second, foreign companies with no activities. The surveys provide (PROCESS) value (rather than simply doing what the community expects of them). these surveys do not provide a domestic ownership are not a depth of information about AGENCY Introduction: Unleashing the Hidden Potential of Women Entrepreneurs in Africa detailed breakdown of female captured). Given their sampling entrepreneurs and their Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Perceived control and ability (“sense of agency”): Internationally, women tend to ownership and control of firms. methodolody, they often over- businesses that is typically report a weaker sense of agency as compared to men.82 Female entrepreneurs Therefore, female control over represent the smallest of the lacking in standard enterprise may perceive themselves as less able than men to engage in successful actions assets and decision-making is firms, without fully capturing and household surveys – thus to increase business performance, such as coming up with ideas for new not typically captured in their businesses with a fixed location making it possible to measure products or persuading a bank to lend them money. Women with a comparatively data. Finally, low response in non-residential areas. gender gaps in performance in weak sense of agency might also believe that most outcomes are sealed by fate rates to some questions can a rigorous way. Moreover, these or determined by “powerful others.” undermine the quality of the surveys include richer measures analysis of the gender gap in of performance, firm capabilities, Ability to act on goals: Female entrepreneurs might be comparatively less able performance. skills, informal access to to take steps to achieve their goals. For example, pressure from family or the finance, networks, and women’s community might prevent a woman from entering a male-dominated sector or economic empowerment than investing profits back into her business instead of paying for household needs. have previous investigations. Achievements is the last component of empowerment. Economic empowerment ACHIEVEMENTS outcomes are typically associated with improvements in well-being across (OUTCOMES) the categories of employment, earnings, education, and consumption. For female entrepreneurs, these types of outcomes include an increase in business performance (profits or revenue), along with their level of control over the income they earn. 38 39 Table 3 Impact-evaluation and household surveys sample of different entrepreneurs Average from Democratic Republic of Congo, Ethiopia, Ghana, Kenya, Malawi, Mozambique, Nigeria, Togo, and Uganda Variables Female entrepreneurs Male entrepreneurs Impact All in percentages, unless Household Impact Household evaluation Difference Difference otherwise stated surveys evaluation surveys surveys surveys INDIVIDUAL CHARACTERISTICS Age 37.9 39.8 -1.9 35.9 40.7 -4.8 Married 66.4 72.2 -5.8 66.9 76.5 -9.6 Ever attended school 97.1 65.8 31.3** 97.4 83.1 14.3 Literate 90.9 55.1 35.9*** 95.3 75.7 19.6 Years of education 10.4 4.8 5.7* 9.6 6.3 3.3 LOCATION CHARACTERISTICS Owns house or space 33.9 72.4 -38.5*** 29.8 75.6 -45.8*** of business Water 20.5 8.5 12.1 21.2 8.1 13.1 Electricity 35.7 47.0 -11.3 55.0 39.2 15.8 Phone available 9.1 0.3 8.9*** 14.4 0.4 14.1*** Cell phone 86.6 68.0 18.7 89.5 60.1 29.3 Computer available 31.2 5.9 25.3 45.8 4.9 40.9*** Introduction: Unleashing the Hidden Potential of Women Entrepreneurs in Africa Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa LABOR CHARACTERISTICS Hours worked last 64.7 40.0 24.7*** 58.4 41.3 17.0** week Source: Authors using impact-evaluation and household surveys. This report seeks to present the evidence in clear and understandable terms, while acknowledging and reflecting on complexity where it exists. If possible, it presents the best available single data point in the main text and in infographics. This may be the average or median across several surveys, the figure provided by the impact-evaluation surveys, or the data provided by the only available source. In some cases, the difference between the surveys is illustrative, and the implications are addressed within the report. Wherever necessary, endnotes provide additional context on what the data show and how the report makes use of it. 40 41 Box 4 At a glance: Quantitative methods used This report employs a combination of quantitative methods, including summary descriptive analysis, regression analysis, and Oaxaca-Blinder decomposition, to better understand the drivers behind gender gaps in entrepreneurship across Africa. The report also uses qualitative assessments, mixed-methods and laboratory experiments to study specific aspects of interest. Within each section, the report provides a summary of the specific method used for that analysis. Oaxaca-Blinder Assessing the Testing for the Analyzing the gender decomposition analysis. productivity gap robustness of findings. gap, step-by-step. To The main decomposition between firms with When negative profits analyze the gender analysis is run for 10 male and female are common, additional gap, the analysis first countries using 14 owners. The logarithm measures of productivity uses a regression datasets from the IE of monthly firm profits are used, including framework to test for a database. Surveys were is used a proxy for the inverse hyperbolic simple (unconditional) carried out between productivity. All datasets sine transformation of difference in monthly 2003 and 2016, with the have some information monthly profits.86 The profits between male large majority carried about firm-level profit. analysis also considers and female business out in the last five years. Some surveys directly the firm’s value-added, owners. Conditional See Box 5 and Appendix ask for the previous which is the amount gender differences 1 for more information. month’s profits.84 obtained by subtracting are examined by In other cases,85 direct costs (costs of controlling for key the surveys record inputs, raw materials, firm-level factors (e.g., information about the and goods purchased capital, labor, inputs), firm’s revenues and for resale) from firm owner characteristics costs, allowing for a sales. In this case, (e.g., age, education calculation of profits by the logarithm of the level), and other subtracting total costs value-added instead firm characteristics (both direct and indirect) of the logarithm of such as sector and from total revenues. profits is used as a business practices. The proxy measure. In multivariate regression general, the results are analysis allows an consistent and do not estimation of whether vary significantly from the gender difference in the analysis of monthly profits can be explained firm profits. by the observable characteristics of the firm and the owner. Introduction: Unleashing the Hidden Potential of Women Entrepreneurs in Africa Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa For specific datasets, an extended analysis that controls for additional firm-specific characteristics is conducted where data is available. This includes the owner’s time and risk preferences, socio-emotional skills, place of business, spouse characteristics, household relationships, and market structure, among other items. The full set of variables included in the extended analysis varies based on the information available for each country. 42 43 01 Who is a female entrepreneur in Africa? Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Profile In the datasets analyzed for this report, the typical female entrepreneur is 37 years old, married or living with a partner, with low levels of education. About 71% of female entrepreneurs have completed primary school, 30% have finished secondary school, and only 9% have completed higher education. Schooling differences between female and male entrepreneurs are significant at the highest levels: 40% of male business owners have completed secondary school and 12% have finished tertiary education. Whether driven by economic Male business owners are more likely than female entrepreneurs to be considered the main provider of the household. About 91% of male business opportunity or acting out of owners consider themselves to be the main household providers, compared necessity, selling food in rural to 69% of female business owners. It is possible that male entrepreneurs overestimate their income’s importance when their wives also run businesses areas or running shops in urban and hide their returns. Female entrepreneurs are more likely to have a spouse centers, women business owners who is a salaried worker – 30% versus 5% for male entrepreneurs. are a visible force across Africa. Nonetheless, women’s business income is a critical source of household This section provides a snapshot earnings. In Kenya, firm-level profits earned by female entrepreneurs of these women, their businesses, represent on average 65% of their household income. In Ghana, women micro- entrepreneurs’ profits account for 33% of the household expenditures.88 In and how they compare to male Malawi, 38% of the money earned by women’s businesses goes toward daily entrepreneurs. household expenses. 44 45 Figure 9 Business Profile Profiles of African women as entrepreneurs vary across countries Among the firms analyzed, the typical Female-owned businesses in Africa have entrepreneur launched her or his own low levels of capital investment. One in five business and has been operating for female entrepreneurs own their place of % women owners % women owners that borrowed approximately 10 years, with female-owned business.93 The capital-stock value of male- married $ in the past for business businesses on average one year younger than owned firms is typically six times larger than those owned by men.89 About 27% of women that of female-owned firms.94 100% 100% business owners report having a mother who 90% 90% runs (or ran) a business, compared with 18% of Female and male entrepreneurs vary in the male business owners. their use of and access to financial services. 80% 80% About half of female business owners have 70% 63% 70% Business is a full-time job for female a bank account, compared to 59% of male 60% 60% entrepreneurs in Africa, who work on average business owners.95 Female entrepreneurs are 50% 50% 22 days per month.90 The typical woman- less likely than male entrepreneurs to have 40% 40% 34% owned enterprise has only two workers, financial accounts (50% vs. 63%). There is a 30% 30% compared with four workers for male-owned great variation in access to loans across the 20% 20% businesses.91 datasets analyzed. For instance, in the case of 10% the Democratic Republic of Congo (DRC), 11% 10% Women and men tend to run different types of female owners have borrowed in the past, 0% 0% of business. For example, while most female while this proportion climbs to 74% in Togo.96 Benin DRC Ghana (1) Ghana (2) Kenya Malawi Mozambique Nigeria (4) Nigeria (5) South Africa Togo Uganda (6) Uganda (7) Ghana (1) Ghana (2) Kenya Malawi Mozambique Nigeria (4) Nigeria (5) South Africa Togo Uganda (6) Uganda (7) entrepreneurs operate in the retail sector, male entrepreneurs are more likely to be in manufacturing. Women are also less likely than men to run enterprises with access to electricity.92 Number of people working in Proportion of female workers in female business (median) female business 20 100% Impact 90% 80% 75% 15 70% 60% Female entrepreneurs are much more likely than male entrepreneurs to employ women, Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 10 50% suggesting that they can be a catalyst for bringing more women into the workforce. About 75% 40% of the workers in female-owned enterprises are women (when excluding the business owner), 5 30% while in male-owned businesses, only 20% of employees are women. This difference stays this 5 20% large even after controlling for the sector of operations. 10% 0 0% Benin DRC Ghana (1) Ghana (2) Ghana (3) Kenya Malawi Mozambique Nigeria (4) Nigeria (5) South Africa Togo Uganda (6) Uganda (7) Benin Ghana (3) Malawi South Africa Togo Part 1 | Who is a female entrepreneur in Africa? Source: Authors using IE database. Ghana (1): Grants for micro-enterprises survey; Ghana (2): Tailoring survey; Ghana (3): Enterprise Census; Nigeria (4): Growth and employment survey; Nigeria (5): Business plan competition survey; Uganda (6): Kassida survey; Uganda (7): Loans, grants, and training survey. Average 46 47 Figure 10 Profiles of female entrepreneurs Female and male entrepreneurs in Africa: How do they differ? WOMEN MEN 64% Married or living 66% with someone 52-84% 72-95% Literate Average: 5 Number of Average: 9 Median: 2 employees Median: 4 Average: 4 Number of female Average: 1.5 Median: 1 employees Median: 0 Owns space where 16% 26% business operates Business has 41% 56% access to electricity Took a loan for 35% 34% business purposes Average amount in 437 477 loans outstanding97 (US$, winsorized at 99th %ile) Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa High-level of 23% 33% financial literacy Has a bank 49% 59% account Retail: 48% Retail: 27% Sewasew Hailu Manufacturing: 32% Sector Manufacturing: 48% Services: 15% Services: 16% Part 1 | Who is a female entrepreneur in Africa? As a little girl, Sewasew Hailu’s grandmother taught her how to make handicrafts. This experience sparked a lifelong passion and led to a profession in fashion design. Now a shop owner and clothing designer at 39, Sewasew is part of a growing fashion industry in Addis Ababa, Ethiopia. Over the last seven years, she has built up her business, winning over loyal customers from other African countries and from Europe who come to her Note: All statistics were obtained from the impact-evaluation surveys except for literacy rates, where two figures are presented: one based on impact-evaluation surveys and the other on household surveys: 52% of female owners and 72% of men are on average for their wedding dresses, suits, and graduation outfits. Together literate based on the household surveys; these proportions rise to 84% for women and 95% for men when using impact-evaluation with her five employees, Sewasew produces all the clothing by surveys. hand in her Addis Ababa shop. Despite this success, Sewasew faces ongoing challenges as a divorced, single mother of three. Source: Authors using impact-evaluation and household surveys. In particular, collateral requirements have prevented her from getting enough financing to grow her business further. 48 49 Merharriet Hailemariam Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Merharriet Hailemariam, 26, originally studied to be a journalist, but today she is an electrician working at a 200-acre condominium development outside of Addis Ababa, Ethiopia. Merharriet switched professions when she learned she could earn more money as an electrician, and convinced two sisters and two friends to follow her example. Now she would like to pursue Part 1 | Who is a female entrepreneur in Africa? a degree in construction technology and management. Merharriet says it is better to be an entrepreneur than to work as an employee of an organization. Betty Ajio Although there is a lot of opportunity for women in construction, few women work in the field due to stereotypes regarding gender roles. Many people think the construction sector is only for men; but women can install electric wiring as well as any man, she says. Betty Ajio, 43, pours molten metal into dirt-covered molds to make cooking pots in the Kisenyi slum of Kampala, Uganda. She makes 120 to 230 pots a day as part of a sweltering, open-air production line. Preparing the molds requires a lot of shoveling, and this physically demanding work is typically done by men, but Betty is proud of her job. A metal fabricator for 23 years, she was encouraged by her father to get into the business to improve her earning potential. Today, the mother of seven has four employees and makes enough money to support a nine-person household and pay school fees. She says she would not trade this work for an easier but lower-paying job, and believes other women should learn the trade. 50 51 02 Gender gap explained First take: How large is the gender gap?  Methodology This analysis draws on data from 14 impact-evaluation datasets from 10 countries in Africa. These countries represent more than 55% of sub-Saharan Africa’s population and West, East, Central and Southern Africa – Benin, the Democratic Republic of Congo, Ethiopia, Ghana, Malawi, Mozambique, Nigeria, South Africa, Togo, and Uganda.98 Controlling for potential core differences in individual, firm and sector characteristics allows for an “apples-to-apples” comparison of the gender gap across contexts. The core set of factors includes: • Firm characteristics: labor,99 capital stock,100 business practices,101 formalization,102 innovation, financial services,103 firm has a loan, firm has at least one additional worker, and sector. • Owner characteristics: age, tenure,104 marital status, household size, number of children per adult, and education. Figure 11 Map of Africa with 10 countries used in impact-evaluation analysis Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa This section seeks first to quantify the extent of the gender gap in profits and then to pinpoint the main contributors to the gender gap in each country. Later sections will then seek to understand what gives rise to the differences between female- and male-owned businesses. Source: Authors / World Bank Group cartography unit. 52 53  Key finding Looking deeper: What is As presented in Figure 1, the unconditional gender profit gap is 34% for the datasets analyzed. Profits in female-owned firms are on average 23% lower than in male-owned firms, after accounting for the driving the gender gap? core differences (Figure 12).105 The gender gaps in business profits range from 1% in Benin to 52% in the manufacturing census from Ghana, when accounting for firm and owner characteristics.106 Differences in firm and individual characteristics account for about one-third of the gender gap in business performance – which suggests that much of the gap cannot be explained just by these core differences.  Methodology To further understand the factors that help explain the gender gap in productivity among those with existing enterprises, an Oaxaca-Blinder Figure 12 decomposition approach is used. This tool allows researchers to predict the main contributors to the gender gap in each country (see Box 5 and Gender gap in profits after controlling for key factors is wide on Appendix 1 for an explanation of the methodology). While the decomposition does not allow for a causal explanation of the differences in business average, but varies across countries performance, the method helps illuminate the answers to key questions: Gender gaps in monthly profits after controlling for differences in firm • Is the gender gap in profits associated with: and owner charactersitics (%) −− Differences in quantity (or levels) of resources used by male and female entrepreneurs, or, −− Differences in returns to those factors and resources used by male and female entrepreneurs WOMEN MEN This quantitative analysis is an important starting point in assessing the factors that help explain the gender gaps in business performance. The Average 23% analysis is relevant for structuring this report’s proposed blueprint of constraints impacting choices women make when operating a business in Benin micro 1% Africa. DRC census 19%** On average, 41% of the gender gaps in business profits can be explained with the Oaxaca-Blinder decomposition and the set of control variables Ethiopia manufacturing 9% available.107 The unexplained parts of the gender gaps in business performance can to some extent be a reflection of measurement issues, Ghana (1) micro 23%* particularly of underlying factors that are not well captured in firm-level Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa surveys (e.g., social norms, intra-household dynamics, confidence, etc.). In Ghana (2) tailoring 23% addition, they may also reflect the lower level of demand for female-owned businesses’ products and services. Other studies have shown that the Ghana (3) manufacturing 52%*** gender gaps in business performance are often not explained by extensive firm- and owner-level characteristics.108 Malawi micro 22%*** Mozambique SMEs 13% Nigeria (4) SMEs 39%** Nigeria (5) SMEs 7% South Africa SMEs 45% Part 2 | Gender gap explained Togo micro 20% Uganda (6) micro 43% Uganda (7) micro 5% Source: Authors using IE database. Note: Ghana (1): Grants for micro-enterprises survey; Ghana (2): Tailoring survey; Ghana (3): Enterprise census; Nigeria (1): Growth and employment survey: Nigeria (2): Business plan competition survey; Uganda (1): Kassida survey; Uganda (2): Loans, grants, and training survey. 54 55 Box 5 gains to female-owned enterprises – seems to Gender differences in taking a loan do not be explained by the frequency with which those contribute to the gap Oaxaca-Blinder decomposition resources are diverted away from the female- owned business due to other household needs114 The results also reveal that gender differences in or other opportunities in the household such as having ever had a loan do not explain productivity This report uses a common in endowments are at the entrepreneur, despite having the husband’s business.115 gaps in any of the studies analyzed – a finding method of breaking down margin also a response to the same level of education, that runs counter to the typical narrative the gender gap in business the returns earned by firms or may obtain greater sales outcomes: the Oaxaca-Blinder at any given moment, the with the same amount of Sector choice also contributes to the gap in around female entrepreneurs. However, male decomposition. It allows a decomposition provides labor. It may be the case that some countries entrepreneurs typically borrow higher amounts distinction to be made between suggestive evidence that men and women increase than female entrepreneurs – a factor that likely two categories of factors: the policymakers can help reduce the level of their investments In three countries, operating in male-dominated contributes to the large gaps in the measure of endowment effect and the the portion of the gender gap until they have equal marginal sectors or not helps explain the gender gap capital investment. structural effect. While this attributable to differences returns, or there may be in business profits.116 Operating in a male- method is widely used in labor in quantity used by ensuring differences in the treatment dominated sector is associated with a larger Other factors contribute to gap in isolated cases economics, it is applied in a that women have the same of men and women by formal gender productivity gap in the Democratic novel way in this report, shedding amounts of these resources and informal institutions, Republic of Congo. In both Nigeria and Uganda, In Nigeria and Ghana, male-owned firms benefit new light on the gender gap in as men – i.e., in the form of markets, programs, etc. African enterprise development. more labor, expanded access Providing women with more differences in returns between businesses from higher returns to the entrepreneur’s to finance, more education, resources may not necessarily operating in male-dominated sectors and those education than do female-owned firms, but in • The endowment effect and so forth. reduce this structural-effect that do not contribute to explaining differences in the remaining countries analyzed, the business refers to the portion of the portion of the gender gap. business performance in one of two datasets for owner’s education level does not explain the gender gap that results from • The structural effect refers Instead, policies may need each country. gender gap in productivity. In Benin and Togo, differences in the quantities to the portion of the gender to understand and target the female entrepreneurs tend to be younger than or levels of resources used in gap that is explained by constraints that contribute to Business practices are an additional issue in their male counterparts. This age difference running a business by male differences between men these differences in returns. some countries increases the gender profit gap, and this could entrepreneurs as compared and women with regard to be a proxy for overall experience. In Nigeria, with female entrepreneurs. the average returns to the Appendix 1 provides further These resources include resources used. For example, technical details on this Male entrepreneurs are more likely to adopt female entrepreneurs derive lower returns than labor, capital, years of a male entrepreneur may methodology. high-quality business practices (such as the use men from running their business for an extra education, and so on. achieve higher productivity of advertising and financial record-keeping). year, which suggests that the gender gap widens Although the differences from schooling than a female The better business practices of male-owned as businesses age. In Malawi, women face enterprises are associated with widening the structural disadvantages to entrepreneurship as gender gap in four countries studied. they themselves age. Evidence is mixed on whether business In addition to the core analysis, specific country formalization contributes to the gender gap factors available in individual datasets were studied (results not shown in Figure 13). These In the Democratic Republic of Congo and Ghana, include analyses of a rich set of information on  Key findings lower registration rates among female-owned socio-emotional traits in Togo; various metrics Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa businesses is associated with larger gender of time and risk preferences and household Female entrepreneurs use less capital and Differences in returns to capital and labor are gaps. In Ghana, Nigeria, and South Africa control of assets in Ghana, Malawi, Mozambique, labor – A key factor in explaining the gender not associated with differences in business formalization is associated with reducing the Nigeria, South Africa, and Uganda; and measures gap in business performance performance gender gap through differences in returns. In the of competition in several countries. Most do Democratic Republic of Congo and South Africa, not help explain the gender gap in productivity. Taken together, the results from the multi- In contrast, men- and female-owned businesses the fact that male-owned businesses are more Notable exceptions include differences in the country analysis point to two key factors in most of the countries analyzed show likely to pay taxes is associated with the gap in level of domestic responsibilities in Malawi and behind the gender productivity gap in African similar average returns to capital and labor. productivity. Mozambique; differences in returns to firm entrepreneurship (Figure 13). First, compared to One explanation for this finding is that male competition in Nigeria and South Africa; and male entrepreneurs, female entrepreneurs have entrepreneurs use capital and labor until their differences in the spouse’s control of earnings in lower levels of capital – which includes inventory, marginal productivity reaches the same level as Togo. equipment, property, and other firm assets. This that of women. A separate way of interpreting gender difference in capital is associated with this result is that when both men and women a wider productivity gap in half of the countries invest the same capital and labor resources examined.109 Second, female-owned businesses (up to the current levels), women produce the Part 2 | Gender gap explained typically use less labor than male-owned ones. same productive outcomes. Both interpretations These gender differences in quantities are accord with causal relationships identified in associated with men having higher profits than recent research where the value at the margin women. Although it may also be the case that of providing large grants under a business plan male entrepreneurs use more capital and labor competition leads women to perform at least due to gender differences in returns to the initial as well as men.110 A similar effect is evident in units of investment, these findings seem to the case of smaller grants allocated to existing suggest policies aimed at providing women with businesses through in-kind capital infusions.111 the same amounts of investment resources as The evidence from microfinance112 and small men. cash grants113 – which reveals limited to no 56 57 Figure 13 The factor significantly contributed to reducing the gender The factor significantly contributed to reducing the gender gap through differences in quantities gap differences in returns Factors contributing to Africa’s productivity gap in entrepreneurship The factor significantly contributed to a widening of the The factor significantly contributed to a widening of the gender gap through differences in quantities gender gap through differences in returns Contributors to the gap in the mean logarithm of monthly profits based on the Oaxaca-Blinder The factor was included in the country analysis but did not The factor was not included in the analysis for that country decomposition analysis. affect the gender gap FIRM’S CHARATERISTICS OWNER’S CHARATERISTICS COUNTRY SURVEY Labor Firm capabilities Formalization Finances Firm’s age Sociodemographic charateristics Sector of Business Log of Log of At least activity Business Business separates Took a loan value of total one Business is Firm pay Firm pay Owner’s Household Children per significantly practices Innovation has a bank business for business Firm’s age Owner age Married Education capital hours of additional registered license taxes tenure size adults dominated by score account money from (ever) stock labor worker men HH Entreprenant Benin Initiative •• •• •• • National DRC Household ••• •• ••• ••• ••• •• Survey Enterprise Ethiopia Census ••• • Grants for Ghana microenterprises •• Tailoring Ghana Enterprises ••• Enterprise ••• ••• Ghana Census • ••• • •• Business ••• Registration Malawi Impact ••• •• • •• Evaluation ••• Matching Grant Scheme Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Mozambique for Business •• Performance Growth and Nigeria Employment • ••• • •• Survey Business Plan Nigeria Competition •• •• •• • •• •• • Online Market South Africa Place • •• •• • Managerial •• Training for Togo Formal and ••• •• •• •• •• Part 2 | Gender gap explained Informal Firms Uganda Kassida •• • • ••• Loans, Grants, and Training Uganda Impact ••• • •• • Evaluation Survey Source: Authors using IE database. Note: ••• p<0.01 •• p<0.05 • p<0.1 58 59 60 Part 2 | Gender gap explained Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 61 03 Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa This section presents a new blueprint to help explain the factors that give rise to the gender gaps in productivity (Figure 14). Under this framework, gender differences in business outcomes are linked to gender differences in the strategic decisions that female and male entrepreneurs make in areas such as: • Sectoral choice: Which business sector should I choose? • Differences in capital and labor: How much should I invest? How much can I invest? • Differences in firm capabilities: What business practices can I employ? Should I formalize the The analysis in the last section business? provides an important starting • Differences in willingness to compete: Should I compete in certain markets? point for policy action. But it does These real-life decisions are then shaped by a set of underlying constraints that fall into three not fully explain the why behind categories: • Contextual factors, such as unequal legal frameworks and social norms; the findings. Why do female • Endowments, including differences in education, confidence, assets and networks; entrepreneurs invest less than • Household-level constraints that undermine women’s ability to control their resources and devote their male counterparts in capital time to their business. and labor? Why are they operating As laid out in Figure 14 the blueprint thus links the: (i) gender-specific underlying constraints; the in specific sectors? Why are they (ii) strategic decisions made by these entrepreneurs, which are in turn influenced by the underlying constraints; and (iii) the business outcomes that reflect these decisions. The rest of this section not borrowing the same amount as will examine how strategic decisions vary between male and female entrepreneurs, and how the male entrepreneurs? underlying constraints influence those decisions. 62 63 Figure 14 Constraints, decisions, and outcomes: Framing the gender gap in business performance in Africa OUTCOMES Gender Strategic decisions: differences in firm performance How do the constrained (sales, profits) choices made by female entrepreneurs influence STRATEGIC DECISIONS the gender gap? 1 2 3 4 Consistent with this report’s analysis, the literature on gender and enterprise development shows that female and male entrepreneurs Sector Differences Differences in Differences in Africa consistently make different strategic segregation in capital firm capabilities in decisions that impact business performance. In and labor (business willingness Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship Madagascar,117 a study shows that 42% to 51% practices / to compete of the gender gap in firm performance can be innovation), and explained by differences in the amount of capital and labor used. In Ghana, women channel formalization a lower share of their accessible capital into their business than do men, and this adversely affects their business performance.118 Female- Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa owned firms in Africa also adopt fewer advanced UNDERLYING CONSTRAINTS business practices – such as marketing strategies and human-resources management – that can be critical for success.119 Furthermore, the choice of the sector in which a business operates is a consequential one: Studies in Africa CONTEXTUAL FACTORS estimate that sector choice can help predict a firm’s profitability.120 This section exploits a wealth of new micro data from impact-evaluation I Legal discrimination ENDOWMENTS HOUSEHOLD-LEVEL and household surveys to assess the differences CONSTRAINTS in each of these strategic decisions and weigh II Social norms their importance in contributing to the gender Allocation of factors IV Education/Skills VIII gap. of Production III Risk of GBV Confidence/Risk V IX Time constraints/Care preferences VI Finance and assets Networks and VII information 64 65  Analyzing gender differences in strategic decisions Figure 15 1 Sector of operations Female and male entrepreneurs consistently choose to operate in different sectors Female entrepreneurs build larger and more profitable companies when they operate in male-dominated sectors. Women in Africa, both those in small-scale self-employment and those operating larger formal firms, are more involved in retail and less involved in the transport, WOMEN MEN B B manufacturing, and construction sectors than men. When formal-sector G Agricolture G female-owned firms are active in manufacturing, they tend to produce M & forestry M garments, textiles, or food. Women are also more active in the hospitality T T industry than in other types of services. This is reflected in data from B B Benin, Ghana, Togo, and Malawi (Figure 15). G Carpentry, G M metal work M However, a study across three regions shows that female-owned T T enterprises in Africa that operate in male-dominated industries are B B G G on average as large as their male-owned counterparts, while firms in Construction M M female-dominated sectors are 56% smaller than those in male-dominated T T sectors.121 This report’s analysis finds that one-quarter of the gender B B gap in profits in the Democratic Republic of Congo can be explained by G Cosmetic G women’s choices to operate in comparatively less profitable sectors. M & salons M T T Moreover, evidence from studies in Ethiopia122 and Uganda123 shows that B B female entrepreneurs who operate in male-dominated sectors (referred to G Food processing G M M as sectoral “crossovers”) perform much better than those in traditionally T T female-dominated sectors. In Uganda, for instance, only 6% of women B B operate in male-dominated sectors, but these “crossover” firms are just G Food: retail G as profitable as those owned by men in those sectors and larger than M & services M female-owned enterprises in traditional sectors. These studies suggest T T that social factors – rather than differential access to education or finance Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship B B – may play the most significant role in a female entrepreneur’s decision G G Other to work in a higher-return sector. Such factors include having male role M M T T models and exposure to the sector by family and friends. Experiments in multiple settings are now testing strategies based on these findings for B B encouraging female entrepreneurs to shift into high-productive sectors.124 G Retail: non G M food products M T T Crossing over to male-dominated sectors increases the risk of Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa B B harassment and gender-based violence, as well as unfair practices from G G male competitors. These risks should be assessed when encouraging M Services M women to enter non-traditional sectors. T T B B G Textil, leather G M & handicrafts M T T B B G G M Transportation M T T 60% 50% 40% 30% 20% 10% 0% 0% 10% 20% 30% 40% 50% 60% B Benin / G Ghana / M Malawi / T Togo micro-entrepreneurs: distribution by gender and sector Source: Authors using IE database. 66 67 2 Capital Figure 16 Female entrepreneurs in Africa have systematically lower levels of business capital – including equipment, inventory and property – than Female entrepreneurs have systematically lower levels of business do their male peers. This plays a central role in explaining the gender gap capital than do male entrepreneurs in business performance. The typical male-owned firm in Africa has over Gender gap in capital investment six times the capital investment of the typical female-owned enterprise. The large differences in capital investment are consistent across a wide range of business types.125 The large gender gap in the level of capital investment persists even after accounting for the sector of operation and 25%*** other firm and entrepreneur characteristics126 – indeed, this gap exceeds 23%** Benin 6% 70% in the Democratic Republic of Congo and in one of the datasets in Ghana (Figure 16). The remaining gap is as high as 30% in Ethiopia and 72%*** DRC 73%*** Togo, and persists at above 50% in South Africa, an environment with 68%*** relatively efficient credit markets. An analysis of the impact-evaluation 46%*** datasets does not detect a significant “reverse gender gap” in capital Ethiopia investment in any country. 38%*** 58%*** 79%*** Ghana (1) 68%*** 1% -1% Ghana (2) -1% 82%*** Ghana (3) 71%*** 37%*** 28%*** Malawi 13%** -6% Mozambique -14% Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship 47%** 34% Nigeria (1) 26% 17% Nigeria (2) 13% 19% 56%** South Africa 54%* Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 61%*** Togo 60%*** 33%*** 49%*** Uganda (1) 10% -12% -10% 0% 10% 20% 30% 40% 50% 60% 70% 80% Gender gap ... controlling for sector ... Core specification Source: Authors using IE database. Note: Ghana (1): Grants for micro-enterprises survey; Ghana (2): Tailoring survey; Ghana (3): Enterprise census; Nigeria (1): Growth and employment survey; Nigeria (2): Business plan competition survey; Uganda (1): Kassida survey. Gender gaps in capital investment are evident both in manufacturing-sector firms and in retail- sector firms, but are smaller in the former. On average, male-owned firms in the manufacturing sector invest in capital at levels 30% higher than female-owned firms. However, this gender gap in manufacturing capital varies significantly across countries. It stands at 77% in the Democratic Republic of Congo, 65% in Ghana, and nearly 50% in Togo and Ethiopia. By contrast, the capital-investment gap in Nigeria favors female-owned manufacturing firms. Excluding Nigeria, the average gender gap in capital in the manufacturing sector rises to 49%, which is close to the capital gap in the retail sector (51%). Gender differences in retail capital investment range from 36% in Malawi to 82% in Ghana. 68 69 Labor Figure 17 Female and male entrepreneurs in Africa differ starkly in their hiring decisions. On average, women make up almost 75% of the employees of Female entrepreneurs use less labor female-owned enterprises – a share nearly four times that of male-owned Gender gap in labor hours enterprises. This sex segregation among workers is not explained by the sector of operation for male- and female-owned firms. Male-owned enterprises in Africa employ more workers on average than do female-owned enterprises. While female-owned firms in the 6%*** 6%* Benin datasets analyzed have an average of two people working in the business, 5%* those male-owned have four people. These differences include the 15%*** entrepreneurs along with formal and informal workers alike, addressing 12%*** DRC the issue that smaller firms are more likely to include family and part- 0% time people working in the business. 25%*** Ethiopia Workers in male-owned firms spend more hours operating the business than do workers in female-owned firms. Female-owned firms 13%*** 18%*** Ghana (1) in Africa use on average about 40 hours less labor per month than do 10%** those owned by men, after accounting for the number of people working -7% in the business. The typical male-owned firm in Africa uses 25% more Ghana (2) labor hours than does the typical female-owned enterprise. The labor gap 1% is particularly large for the micro-enterprises in the impact-evaluation 12% Ghana (3) database.127 For example, in the outskirts of Kampala in Uganda, the total number of hours of monthly labor for male-owned firms, on average at 12%*** 827 hours, dwarfs the 361 hours of labor used by female-owned firms. 17%*** Malawi 14%*** These large labor disparities contribute to the gender gap in firm -2% Mozambique performance. Female-owned firms’ lower levels of labor contribute 1% to gender productivity gaps in six of the 10 countries analyzed: the 25%*** Democratic Republic of Congo, Ghana, Malawi, Nigeria, Togo, and Uganda. Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship 25%*** Nigeria (1) Even though female entrepreneurs have more household responsibilities 18%*** (and therefore fewer available hours to dedicate to their business), female 21%** entrepreneurs do not seem to compensate by hiring more workers. 22%** Nigeria (2) However, in only two of these countries (Ghana and Uganda) did women 12% obtain lower returns than men from their business labor – suggesting that 14% South Africa in most of the countries analyzed, women and men derive similar average 3% profit gains from their employees’ time.128 Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 38%*** 26%*** Togo The gender difference in labor is explained by a set of factors. Sector 16%*** allocation alone does not explain the differences between the size of men 41%*** Uganda (1) and female-owned firms, but the gap is lower in most countries when 13% 2% other firm and personal characteristics are taken into account (Figure 17). For example, after accounting for these factors,129 the gender labor gap -10% 0% 10% 20% 30% 40% 50% is eliminated in the Democratic Republic of Congo; Togo’s gap shrinks by more than half to 16%; the gap observed in one survey in Nigeria drops from 25% to 18%; and the gender gap in Uganda is reduced from 41 to 2%. Gender gap ... controlling for sector ... Core specification Source: Authors using IE database. Note: Ghana (1): Grants for micro-enterprises survey; Ghana (2): Tailoring survey; Ghana (3): Enterprise census; Nigeria (1): Growth and employment survey: Nigeria (2): Business plan competition survey; Uganda (1): Kassida survey. 70 71 3 Firm capabilities: Business practices, innovation, Figure 18 and formalization Female entrepreneurs less likely to adopt good business practices Business practices Gender gap in an index of business practices Differences in the adoption of good business practices help explain the gender gap in productivity for some countries. Recent research on small firms in developing countries130 shows that female entrepreneurs are less likely to adopt advanced business practices including: (i) marketing 15%** practices such as advertising and price comparisons with competitors; 25%*** Benin 17%** (ii) purchase- and inventory-control practices, including negotiating with 23%*** suppliers and techniques designed to avoid running out of stock; (iii) 28%*** DRC cost-analysis and record-keeping practices, for instance keeping written 22%*** business records or identifying which products are most profitable; and 3% (iv) financial-planning practices, such as setting sales targets and creating Ethiopia budgets. 32%*** 40%*** Ghana (1) Using a standardized business-practices score, this report finds gender 31% differences in favor of male-owned firms that are large and significant 4% in most of the datasets analyzed (Figure 18). Only in Benin did female- Ghana (2) owned firms score higher than male-owned businesses with respect to 3% these practices. The gender gap in business practices remains in 10 of 13% Malawi 13 datasets in Africa, even after accounting for personal and firm-level 11% 10% factors. Yet, these adjusted differences are statistically significant in only 3% four of these datasets. And in contrast, the gap in business practices Mozambique reverses in favor of female-owned businesses in Togo after controlling for 6% other factors. 9% 7% Nigeria (1) The gender gaps in business-practices indexes are on average less than 3% half the size of the capital-investment gaps. This may help explain the 6% Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship 7% Nigeria (2) relatively muted impact deriving from efforts to teach business skills on 8% reducing business-performance gender gaps. 21% South Africa 16% 11%** 18%*** Togo -17% Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 24%*** 11% Uganda (1) 8% -20% -10% 0% 10% 20% 30% 40% Gender gap ... controlling for sector ... Core specification Source: Authors using IE database. Note: Ghana (1): Grants for micro-enterprises survey; Ghana (2): Tailoring survey; Nigeria (1): Growth and employment survey: Nigeria (2): Business plan competition survey; Uganda (1): Kassida survey. 72 73 Innovation Figure 19 The analysis for this report shows some differences in innovation Gaps in innovation more pronounced in introducing new processes practices between male and female-owned enterprises, notably on the introduction of new processes within the firm. Innovation relates to the Gender gap in innovation (pp) ability of a business “to use knowledge to develop and apply new ideas that result in changes in the production and organizational structure of the firm.”131 This includes (i) the introduction of a new product or a modification -4.0 -2.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 to an existing product, (ii) a new process or technology, (iii) the discovery of a new market, (iv) the development of new sources of supply for inputs and -1.9 Malawi raw materials, and (v) changes in industrial organization. Data on innovation in Africa is scarce, and gender-gap analysis focusing -1.9 Mozambique on innovation (before this report) is even slimmer.132 In a survey of female- 8.0*** owned firms in Kenya, 82% of female entrepreneurs indicated that they -3.1 ask customers if there are other products or services the clients would 3.5 Nigeria buy from them. Almost 20% of women business owners said also that they were planning to introduce new products or improve existing products 4.0** Togo in the coming two years.133 However, 31% of the female entrepreneurs in 11.0*** Kenya suggested that they were unwilling to try something new unless 10.3*** they were 100% certain it would succeed. Uganda 9.1*** A survey in Togo elicited information on innovation practices among female and male-owned businesses (Figure 19). An analysis finds that female-owned businesses are less likely than their male-owned peers to have introduced new products and services or to have implemented a new Product innovation Process innovation business process. Women are half as likely as men to have introduced a new product in their neighborhood of operation. This may be because there is little space to innovate in the specific sectors in which women typically operate. Source: Authors using IE database. Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship Note: *** p<0.01, ** p<0.05, * p<0.1. In Mozambique, about 18% of female entrepreneurs and 16% of male entrepreneurs have introduced a new product in the past 12 months. However, women are less likely to have introduced a new process – 31% of women and 39% of men have carried out innovation of this kind in the past 12 months. In Nigeria, female-owned firms are 20% less likely to have improved existing products or to have introduced a new product design. Female-owned businesses in Nigeria also spend less on R&D,134 and their Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa owners are on average less confident than men in their ability to come up with an idea for a new business product or service. An analysis of these differences indicates that they can be explained by differences in firm and individual characteristics of the entrepreneurs. In Uganda, 18% of female entrepreneurs introduced a new product or service in the past 12 months, and 16% introduced a new process in Furthermore, gender gaps in innovation do not arise when examining larger the same period. But both of these are 37% below male-owned firms’ firms in Africa. An analysis for this report of Enterprise Survey data from likelihood of having introduced a new product or new process. Women in 11 countries135 did not reveal significant differences in innovation measures Uganda are also less likely to develop a new product or service in the next between female and male-owned formal firms with at least five employees. 12 months. These differences are explained by a range of factors, chiefly This suggests that women who have overcome growth constraints and the differences in the entrepreneurs’ education levels. reached such a scale adopt innovative practices at rates similar to men.136 74 75 Formalization Figure 20 Women own approximately half of the firms in Africa, but many are in the informal economy. Only a third of formal firms with five or more Rate of female ownership of formal firms in Africa lags that in East employees have any female ownership; even fewer (17%) have at least Asia and Pacific and Latin America,140 but compares favorably to other 50% female ownership (Figure 21).137 This clearly indicates differences regions in business-formalization rates, which are likely associated with the gaps measured in the size of the businesses. % of formal firms with female ownership by region In theory, business formalization could help increase women’s use of formal financial services and facilitate their access to formal business institutions such as private-sector associations and government- 32% development programs. AFRICA 25% 17% However, this report’s analysis does not find a clear pattern between 55% individual business formalization and the gender gap in enterprise EAST ASIA AND PACIFIC 53% 26% performance. In datasets in Ghana, Nigeria, and South Africa, female- owned businesses enjoy higher productivity returns to formalizing than 31% male-owned businesses. This finding may relate to a small subset of EUROPE AND CENTRAL ASIA 20% 17% female-owned firms that are operating comparatively large formal businesses – and whose owners may be less constrained than the 41% average female business owner in Africa. Recent studies in Benin138 LATIN AMERICA 36% 0% and Malawi139 with a broader set of entrepreneurs indicate a lack of impact on business performance and access to finance from helping 21% entrepreneurs obtain formal status. MIDDLE EAST AND NORTH AFRICA 9% 6% 18% SOUTH ASIA 13% 11% 35% High Income: OECD Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship 24% 20% 38% High Income: non OECD 43% 25% 0% 10% 20% 30% 40% 50% 60% Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa % Female % Female ownership % Female majority ownership sole propietors only ownership Source: Authors using World Bank Enterprise Surveys, most recent years. Note: Data considering firms with five or more employees. 76 77 4 Figure 21 Decision to compete Women’s reluctance to compete in stereotypically male environments Rates of female participation in the formal economy vary widely, but may help explain some of the gender gaps in entrepreneurial no countries have achieved gender parity with regard to majority performance in Africa. There is limited research on women and men’s competitive behavior within Africa. When studying the Maasai in Tanzania, ownership rates researchers141 found that men were twice as likely as women to enter into competition during an experiment. A lab experiment in Kenya conducted % of formal businesses with majority female ownership for this report,142 found that women were less than half as likely to participate in a competition than men, no matter whether the stakes associated with winning were low or high. Women were much more likely 50% to enter into competitions when facing female competitors than when they faced male competitors. Box 6 describes this set of experiments in Kenya. 40% Numerous laboratory experiments from outside of Africa confirm these findings.143, 144 Women opt out of engaging in competition even when they 30% are qualified.145 This leads to different economic outcomes, with more high-ability women missing out on significant earnings as compared to 20% high-ability men. This difference in attitudes between men and women REGIONAL AVERAGE can explain some of the gaps with regard to engaging in competition. 10% It could be that “while the prospect of engaging in a future competition [with men] may cause women to anticipate a psychic cost and deter them from tournaments, men may anticipate a psychic benefit and instead be 0% drawn to them.”146 This may stem from different upbringings as children, Lesotho Botswana Namibia Cape Verde Benin Zambia Rwanda Côte d’Ivoire Madagascar Mali Ghana Kenya CAR Mauritius Liberia Ethiopia Burundi Cameroon Uganda Angola Malawi Senegal Nigeria Tanzania Gabon Niger Togo Congo Burkina Faso DRC South Sudan Eritrea Sierra Leone Chad Sudan Mauritania with boys being the ones who are typically encouraged to be assertive.147 The expectations set by social norms could further exacerbate these differences. In fact, in the same study in Tanzania, researchers found that when applying the experiment to the Khasi society, a matrilineal society as opposed to the Maasai which is patrilocal, women are found to be more competitive than men.148, 149 Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship Source: Authors using World Bank Enterprise Surveys, most recent years However, limited data150 on willingness to compete makes it difficult to weigh the importance of gender differences in engaging in competition in explaining the gender gap in business performance. Data on this issue is rarely present in impact-evaluation, household, or enterprise surveys, not least because the most effective ways to measure this are through laboratory experiments similar to those described in Box 6.151 Planned lean evaluation studies in Kenya testing multiple mechanisms Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa for encouraging women to apply to coding bootcamps and a business plan competition will complement these lab experiments. 78 79 Box 6 Kenya: Exploring gender differences in the willingness to compete To examine gender differences in competition behavior in  Key findings traditional male domains and sectors, a series of laboratory The results reveal a significant gender gap in willingness to compete: experiments in Kenya were conducted.152 Recent research in the region has shown that the willingness to compete in lab (1) Women were much less likely to compete in a mixed-gender environment. In the standard group, only 31% of women chose to compete as compared to 69% of settings is associated with competitive choices, including men. This is very similar to Niederle and Vesterlund’s findings in the United States, decisions resulting in higher capital- and labor-investment where they found 35% of women and 73% of men entering the competition. rates.153 (2) Women were significantly more likely to enter into the competition when faced with only female competitors: 53 percent of women entered the competition in the The experiments followed the basic design of the Niederle and Vesterlund (2007) female-only competition environment (Treatment 1) as compared to 31 percent of competition experiments in the United States, but used a stereotypical male women in the mixed-gender competition (standard group).154 performance task that was appropriate for the Kenyan context and the population of interest. (3) Women were significantly less likely to compete in a mixed-gender competition environment with high stakes: in Treatment 2, only 28% of women Due to sample size limitations, the study did not analyze the results by marital chose to compete as compared to 61% of men. status or any other metric and focused primarily on a comparison between men and women in terms of the willingness to compete in different situations. (4) Regardless of the group, past performance does not appear to be a significant predictor of the choice to enter the competition.  Methodology Figure 22 The experiments included three groups, each with 80 study participants: Women less likely to compete in mixed-gender than Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship in female-only contexts 1 2 3 Gender differences in competition entry, Kenya The standard group (control A female-only group A high-stakes groups 80% group) examined gender (Treatment 1) replicated the (Treatment 2) replicated the Female Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa differences in competition standard group, but with only standard group with a mixed-gender behavior in a stereotypical male female study participants facing competition environment; however, 69% Male 70% domain and in a mixed-gender other female competitors. This the financial stakes associated with competitive environment that allowed an examination of whether winning were increased by 10 times. 61% included two men and two women change their competition This allowed an examination of 60% women in each competition behavior depending on who they whether the stakes associated with 53% group. are competing against. the competition change behavior. 50% 40% 31% 28% 30% 20% Participants reflected the typical pool of African entrepreneurs. They were between 18 and 40 years of age and had completed schooling at least through the primary 10% level, with a maximum secondary-level education. To ensure that groups were balanced, participants were randomly assigned to different session times. 0% The study relied on touch-screen computers. Study participants were seated in Standard Female only Mixed-gender/ randomly assigned groups of four in the lab. Each participant had his or her own mixed-gender group group high stakes group cubicle, with a computer terminal that was separated by partition walls. While individuals could see each other, they were not able to see each other’s computer screens, performance, or decision information. In addition, the topic of gender was not explicitly mentioned during the experiment. 80 81 Nine factors holding back women’s In recent years, many African countries have made strides in removing formal legal barriers that have prevented female entrepreneurs from owning and running successful businesses. In 2012, Mali removed legal business performance in Africa restrictions that prevented married women from registering a business. Côte d’Ivoire now allows both spouses the legal right to choose where to live, whether and where to work outside the home – and to be designated Understanding the constraints “head of household.” In 2016, Rwanda granted both spouses equal rights to choose where to live, and eliminated a provision designating the husband as “head of household.” That same year, the Democratic Republic of Female entrepreneurs do not make business decisions in a vacuum. Rather, their business decisions Congo eliminated provisions that prevented married women from signing differ systematically from those of male entrepreneurs because they are constrained in a way that contracts, getting jobs, registering companies, opening bank accounts, or men are not. obtaining loans without the permission of their husbands. The obligation for a married woman to obey her husband, as well as gender-based This section presents nine of these underlying constraints and explores the evidence on why the discrimination by creditors in financial transactions were also prohibited. In factors matter and the extent to which they contribute to the gap in business performance. Some of Zambia and Guinea, discrimination based on gender and marital status in these factors are more established in the literature, while others, such as the risk of gender-based access to credit is now forbidden by law. violence or risk preferences, are more limited in terms of evidence regarding their relevance for explaining gender gaps in business performance. While analyzing each case individually, the report Some countries have sought to eliminate gender discrimination arising will highlight where the evidence is strongest. from customary law. Zimbabwe approved important constitutional reforms in 2013, and Botswana has also made significant strides with regard to The section is followed in the next section by deep-dive analyses of three factors for which the eliminating discrimination in customary law. evidence on its impact on business performance is relatively limited, or where the mechanisms behind the evidence remain unclear. The deep dives include studies of social norms, networks and However, and despite the significant regulatory advances in Africa, information, and household allocation of factors of resources. many countries grapple with challenges in enforcing these laws. Gender discrimination is still a factor in the implementation of these laws, and judicial systems are not responding fast enough to these concerns.  Contextual factors Bottom line: Progress means legal discrimination may not be constraining female entrepreneurs as much as in the past. Continuous work is needed Legal discrimination to ensure that laws avoid gender discrimination, and that they are I appropriately implemented. Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship Female entrepreneurs cannot have equal economic opportunity if a country’s laws restrict a woman’s ability to own and run a business. In some countries, formal business and family laws deny women rights equal to those of men to register a business, sign a contract, open a bank account, or own and inherit property. For example, in Cameroon, Chad, Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa the Democratic Republic of Congo, and the Republic of Congo, the law gives husbands sole control over marital property, making it difficult for married women to obtain loans to finance their business, since these loans Social norms II often require property as collateral.155 In Chad, Guinea-Bissau, and Niger, married women need their husbands’ permission to open a bank account. Social norms exert strong influence over female entrepreneurs’ strategic In Equatorial Guinea, a wife needs her husband’s permission to sign a choices, and can constrain their ability to grow their businesses. contract. A review in the World Bank’s Women, Business and the Law 2018 (WBL) Social norms are informal rules about what behaviors are typical and publication finds that formal law treats women and men equally in their desirable that are widely accepted and shared within a given social group. pursuit of economic opportunities in only three African countries156 – South No formal sanction system exists to support and enforce such rules. Yet Africa, Zimbabwe, and Rwanda. Countries such as Mozambique, Ghana, because such norms can become part of a community’s system of values, and Liberia comply with 21 out of 22 measures of equality with regard to they retain a strong influence on behavior, and people may feel a strong economic opportunities. Three economies (Sudan, Guinea-Bissau, and obligation to obey them.159 Gender norms are shared expectations about Niger) had at least 10 significant differences in the legal status accorded differences between men and women, as well as what behaviors are to women and men that made it more difficult for female entrepreneurs common and appropriate for men and women. to operate. Another nine countries had at least eight differences.157 Furthermore, women in 65% of African economies face legal challenges in Female entrepreneurs make decisions shaped by norms that establish building credit ratings, and 19% of countries in Africa receive a score of zero expectations regarding accepted roles and responsibilities for men and with regard to protecting women from violence. Even where statutory law women, and widely shared conceptions of masculinity and femininity.160 seeks to promote gender equality, female entrepreneurs may face barriers For example, women perform the vast majority of housework in Africa deriving from customary law – the uncodified traditional legal system of a (described in the time constraints/care section), which affects the amount community or locale – which often takes precedence in settings with legal of time they have for their businesses. Non-conforming behavior can result pluralism. The influence of customary law continues to be stronger in sub- in discrimination or other sanctions. For example, women who strive to Saharan Africa than in any other region.158 grow their businesses can face reduced support or violence from a spouse 82 83 who feels threatened, ostracization from social networks of peers who disapprove, or discrimination from suppliers or buyers. When women seek Risk of gender-based violence III to contradict societal prescriptions for female behavior, they may face retaliation.161 And fear of retaliation could deter women from being assertive Experiencing gender-based violence can take a toll on women’s in work and competitive negotiation settings.162 health and well-being that hinders their ability to run their businesses effectively. Research demonstrates that women (as well as men) internalize these social norms through gendered self-conceptions, self-assessments, and Gender-based violence (GBV) – which includes physical violence, emotional aspirations: violence, sexual violence and economic violence – is prevalent in many African countries and has the potential to shape women’s business • Self-conceptions, or how women imagine themselves and their employment decisions. Two-thirds of young female apprentices in Ibadan, responsibilities in life, can influence their career choices.163 Research in Nigeria’s third-most-populous city, reported having experienced physical Ghana finds that some women view entrepreneurial success as core to violence. Among this group, 39% indicated that their employer was the their identity, while others approach their businesses more as generators most recent perpetrator.171 A study conducted in Peru, Haiti, and Zambia of petty cash.164 Gender norms that emphasize women’s role in the home finds that women who were victims of GBV were more likely to be employed and men’s role as providers discourage women from developing an than women who had not been abused – suggesting that engaging in work entrepreneurial identity. outside the home may itself put women at risk for GBV.172 Women may • Women’s self-assessments, or the specific beliefs about their ability prefer self-employment because it helps them avoid sexual harassment, as to perform specific tasks, reflect gender differences that affect their reported by 20% of participants in a study in Liberia.173 confidence in their ability to engage in technical occupations.165 If women perceive that they are less able to compete in and pursue In recent impact-evaluation surveys, about 11% and 14% of female business activities considered to be “male,” female entrepreneurs entrepreneurs in Malawi and Uganda respectively reported that they had may avoid entering higher-value sectors that present greater business experienced sexual harassment while running their businesses in the opportunities.166 preceding 12 months. Moreover, in Malawi, 14% and 11% of female and • Gendered differences in aspirations, or the shared mental male entrepreneurs respectively indicated that a man has a good reason representations of the proper roles, behaviors, and goals for men and to hit his wife for not completing the housework, disobeying, refusing women, may lead to substantial differences in aspirations between men to have sexual relations with him, or being unfaithful. Similarly, 14% of and women.167 When female entrepreneurs conform to social norms, they female entrepreneurs in Malawi have experienced physical or emotional may be less likely to aspire to growing a large business if that does not violence from their domestic partner; 32% say the husband insists on align with how women in their community behave. knowing where they are at all times; 7% say their husbands force them to have sexual intercourse; and 8% say their husbands try to keep them from Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship Direct quantitative evidence regarding the degree to which social norms seeing their friends.174 contribute to the gender gap in business performance is limited. Recent research in Nigeria suggests that effective training can help women In Tanzania, 11% of female entrepreneurs say they have received a request overcome unconscious self-defeating biases. It shows that promoting skills for sex in exchange for support for their business in the past 12 months.175 in the information- and communications-technology sector was more Asked the same question in terms of other women – “out of 10 women, how than three times as effective (with regard to having women switch their many do you think have received requests for sex in exchange for help with professional activities to that sector) for women who had implicit biases their business in the past 12 months?” – the proportion reaches 40%.176 Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa against women’s professional work than it was for unbiased women.168 There is little rigorous evidence on how the risk of GBV impacts the Furthermore, because of their greater power and influence, men can play decisions made by female entrepreneurs or their capacity to run a a significant role in either perpetuating or changing existing gender norms productive business. Existing research suggests that GBV has complex and how they impact women’s economic decisions and empowerment. In consequences. Taking time off work due to injury may lead to losses in Ethiopia, women with spousal support were found to be significantly more productivity and revenues. Furthermore, exposure to GBV can take a toll likely to cross over into the more profitable traditionally male-dominated on women’s mental health, which may manifest as insomnia, anxiety, and sectors.169 In Uganda, women who had male role models in their youth (e.g., social dysfunction,177 and may hinder the victims’ managerial capacity. A fathers or politicians) were 20% to 28% more likely to cross over into these study in Colombia finds that the monthly earnings of women exposed to sectors than women who did not have these role models in their lives.170 GBV are on average 70% lower than the earnings of non-victims.178 A study in Rwanda suggests that engaging men in group discussions on sensitive Bottom line: The influence of gender norms on women’s business issues regarding sexual health and gender-based violence can lead to large decisions is likely pervasive and strong, including through its influence on and significant impacts on women’s reports of violence perpetrated by major choices such as the sector of operations. More research is needed intimate physical and sexual partners, as well increased decision-making on mechanisms by which social norms can be circumvented or changed as power for women.179 This growing evidence is suggestive of the importance a means of reducing the gap in business performance in Africa. For more, of the constraint and the need to identify effective solutions to reduce its see “deep dive” 1. prevalence. Bottom line: GBV likely has complex consequences on women’s mental health, which may hinder their managerial capacities and lead to losses in productivity and revenues. More rigorous evidence is needed in order to understand how GBV impacts female entrepreneurs and their economic performance, including on whether women shy away from growing their businesses because of the risk180 of increasing their exposure to GBV. 84 85  Endowments 3. Differences in socio-emotional skills. For taking advantage of economic opportunities, cognitive and noncognitive skills are equally important.187 A growing body of evidence188 shows that socio-emotional skills such as self-starting behavior, future orientation and persistence have a positive bearing on business success.189 Even if there are no average differences in socio-emotional skills between male and female entrepreneurs, they Education and skills gaps can be used differently for business-development purposes according to IV While most African countries have achieved gender parity in access to need. primary education,181 a persistent gap in educational attainment between male and female entrepreneurs in older cohorts and in secondary Data from Mozambique suggests limited average differences in socio- education for younger cohorts may help explain gender differences in emotional skills between male and female entrepreneurs. Data from Togo strategic business decisions. points to a gender gap in some socio-emotional skills.190 Men and women in Togo are just as likely to search for new and better solutions, seek new Traditionally, the gender gap in educational and skill attainment in Africa is experiences, explore novel ideas, and demonstrate emotional stability and driven by a range of factors, including greater parental investment in sons’ empathy. However, male entrepreneurs in Togo score higher on measures educations182 and the prevalence of early marriages for girls. A systematic of ambition, creativity, innovation, and imagination than do female review of more than 100 empirical studies indicated that education has a entrepreneurs. Men are also more likely than women to enjoy gathering substantial effect on self-employment earnings,183 and found that this effect the right people for their business. On the other hand, women have greater tends to be larger for women than for men. verbal-communication skills than men. Data from Togo suggests a gender gap of 24 percentage points on a combination of socio-emotional skills. Three types of education and skills gaps matter most for entrepreneurs across Africa: Bottom line: Education and skills gaps in Africa are wide and persistent, and likely have a strong influence on women’s business decisions. 1. Differences in formal education. Average schooling levels of self- employed women in Africa vary substantially by country, ranging from under two years in Burkina Faso and Côte d’Ivoire to eight years in Ghana. Remarkably, regardless of the absolute level of the average, self-employed women have systematically completed fewer years of education than have self-employed men. Even in Ghana and Kenya, where self-employed women have completed on average the most years of education, this gender gap remains significant.184 As most countries have Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship now reached gender parity at the primary-education level, this gap will likely narrow over time as new cohorts join the labor market. 2. Differences in management skills. Beyond increased access to formal education, opportunities for developing management and entrepreneurial skills may be greater for men than for women.185 Young men may have Confidence / risk preferences V more opportunities to receive either formal training through specialized Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Female entrepreneurs in Africa have lower confidence in their abilities,191 providers or to be trained by a member of their network. Furthermore, which may make them less willing to compete192 (and win) – especially in women may also be less likely than men to enroll in training programs stereotypically male domains.193 in technical sectors with comparatively high average earnings such as mechanics, electronics, or construction.186 Women business owners in Africa frequently show less confidence than their male counterparts. For example, among entrepreneurs in Ghana, This report’s analysis finds that in Malawi and Mozambique, male women are 14% less likely than men to think they would make a good entrepreneurs often have greater technical skills than do female business leader.194 owners. In Malawi and Togo, the gender gap in financial literacy is significant. Data from Togo shows that more than 90% of business owners Women’s lack of confidence relative to men has varied and important have the confidence to manage a team or a business, but female owners consequences. Women are unlikely to decide to pursue opportunities in have less confidence in their managerial skills than males. In Nigeria, traditionally male – and traditionally more profitable – domains if they do Malawi, and Mozambique, however, women have confidence levels in not believe that they can succeed. Research shows that when individuals their management skills that are similar to those shown by men. believe that women are better at a stereotypically female task and men are better at a stereotypically male task, the gender gap in confidence The gap in skills may also be tied to a gap in training, with male-owned and competition entry is larger in the task that is stereotypically male.195 firms being more likely to participate in training programs than female- Furthermore, women who are responsible for others – as entrepreneurs owned firms (that is the case in the Democratic Republic of Congo, often are – are less confident in their abilities as compared to women who for instance). And these gaps also extend within firms: in Nigeria are only representing themselves. By contrast, men do not experience this and Uganda, male-owned firms are more likely than female-owned difference in confidence. This is the result of women setting a higher bar for enterprises to offer internal training courses. themselves when they represent the interests of others.196 Women’s lack of confidence could be tied to lower appetite for risk.197 The analysis carried out for this report does not show a clear pattern on this issue. Two datasets in Ghana show a self-reported gender gap in 86 87 the likelihood of taking risks among micro-entrepreneurs. In one of the datasets, the gender gap with regard to being a risk-taker is 23%. This difference remains large and significant after controlling for personal and household characteristics. However, the report finds no difference between men and women micro-entrepreneurs in Togo asked to choose between alternative businesses with different levels of risk (Table 4). In the context of a business plan competition in Nigeria, which studies indicate attract comparatively competitive women,198 female entrepreneurs say they would prefer a lottery ticket over a fixed amount of money more often than do male entrepreneurs. More analysis is needed on risk-taking behavior, including in the context of lab experiments in the region. In this regard, one recent laboratory experiment in Tanzania randomly assigned microfinance clients to different groups, including female-only, mixed-gender and male-only groups. Members of the female-only groups were more willing to take risks than were members of the male-only or mixed-gender groups.199 Table 4 Female and male entrepreneurs in Togo show broadly similar preferences for alternative businesses with various degrees of risk Option Profit in a bad Profit in a good Women Men Difference month (US$) month (US$) Business 1 30 30 8% 9% -1.1% Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship Business 2 27 57 10% 13% -3.2%* Business 3 24 72 16% 18% -1.6% Business 4 21 75 14% 12% 1.3% Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Business 5 18 90 13% 13% -0.3% Business 6 12 96 7% 5% 1.9% Business 7 6 94 12% 12% 0.2% Business 8 0 120 20% 17% 2.9% Source: Authors using IE database. Note: *** p<0.01, ** p<0.05, * p<0.1 Bottom line: Women’s lack of confidence can keep them from taking risks that could lead to higher returns. More experimental work on mechanisms by which this lack of confidence can be reduced and its implications for business performance may indicate promising future policy avenues. 88 89 Figure 23 Finance and assets VI While women’s access to formal financial services has improved, female entrepreneurs continue to control fewer assets than male entrepreneurs, affecting their capacity to invest in their business and Women typically receive loans of lower amounts than men Loans (amount) held by female-owned businesses as a proportion of those held by male-owned businesses (%) access loans of sufficient size. Access to finance is critical to the success of any business, and poses a Female problem for both men and women. The gender gap in obtaining loans in sub-Saharan Africa appears to be relatively modest. The microfinance Male industry’s focus on making services available to women has reduced 140% gender disparities in credit usage. A comprehensive study using data from 37 countries across Africa finds no evidence that women are disadvantaged in credit usage after controlling for firm and entrepreneur 120% characteristics.200 Data from the 2018 Global Financial Inclusion Database (Findex)201 shows that the share of women receiving a loan from any source 100% was five percentage points lower than that of men (43% of women and 48% of men). And although 6% of women and 7.9% of men borrowed from a financial institution in 2017, the gender gap in obtaining loans from financial 80% institutions is smaller in Africa than in any other region of the world. 60% However, the fact that women tend to have less assets and savings than men can significantly curb their financing opportunities. Financial 40% assets enable business owners to grow their asset base, fund productive investments, and smooth their household consumption. Where financial markets are imperfect, entrepreneurs’ ability to obtain finance is largely 20% determined by their ability to provide collateral rather than by the stream of revenues that their business can be expected to generate. Even where 0% laws do not restrict women’s access to and control over assets, providing collateral is harder for women due to their smaller asset base. DRC Malawi Mozambique South Africa Togo Uganda Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship This report’s analysis shows consistently large gender differences in the Source: Authors using IE Database size of outstanding loans for various target groups of entrepreneurs in Africa (Figure 23). In the Democratic Republic of Congo, loans obtained by female entrepreneurs are on average 43% of the value obtained by male business owners. In Malawi (micro), South Africa (SMEs), Togo (micro), and Uganda (micro), this proportion ranges from 38% to 74% – pointing to a Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa sizable gap in the volume of credit for both micro-entrepreneurs and SMEs. Additionally, when women do receive loans, they may not receive them on Only in Mozambique, where the sample is concentrated among larger the same terms as men. Getting loans of smaller amounts affects female firms for both women and men, do female-owned companies have higher entrepreneurs’ ability to use these loans productively.202 One study suggests aggregate loan amounts than male-owned enterprises. women pay higher average interest rates or receive loans with shorter maturities.203 However, analysis for this report in Benin, DRC, Ghana, Malawi, Mozambique, Nigeria, South Africa, Togo and Uganda did not find significant gender gaps in the following loan conditions when taking into account the size of the loan: interest paid, loan maturity, need to present a business plan, and the need to show credit history.204 Nevertheless, even after accounting for the size of the loans, the analysis suggests that women have a greater need than men to present collateral in Mozambique, Togo and Uganda, three of the five datasets containing this information.205 Finally, the Global Financial Inclusion database shows that about 38% of men and 27% of women in Africa had an account in a formal institution, while 24% of men and 18% of women had mobile money accounts.206,207 In Malawi, 68% of male and 47% of female business owners used personal savings for the initial investment in their business. However, these data ignore important differences. About 23% of women, compared to only 2% of men, had to rely on capital from their spouse to start their business.208 And it is clear that the gender gap in assets can translate into gender differences in productive outcomes. For example, a survey finds that in central Uganda, 30% of male-owned transport-related assets – which facilitate mobility and market access – compared to only 1% of women.209 90 91 In some countries, mobile money accounts might be helping to narrow the • Women’s networks include a higher share of family and kin gender gap in financial inclusion.210 In the eight countries where 20% or relationships.223 These “strong ties” – which can also be thought of more of adults have only a mobile money account, there is a statistically as “old” and long-standing connections – may be less valuable than significant gap between men and women in the likelihood of having an “weak ties” – or “new” and looser connections – in creating professional account. However, just two of these countries – Burkina Faso and Tanzania opportunities.224 This gender difference in network composition may stem – show a gender gap in having only a mobile money account. In Côte from norms limiting women’s ability to venture away from their home and d’Ivoire, for example, men are twice as likely as women to have a financial interact with strangers,225 and from women’s greater responsibilities in institution account – yet women are just as likely as men to have only a household production and childcare.226 mobile money account. • Gender-segregated networks can limit economic opportunities for female entrepreneurs. In a recruitment drive in Malawi, men’s strong In Africa, mobile phone ownership offers large opportunities among tendency to refer other men for jobs created a significant gender the unbanked – 54% of men have a mobile phone, as compared to 43% differential in access to opportunities.227 While that study focuses on of women. In several countries, including Mozambique and Senegal, access to formal employment, it suggests that men tend to benefit from unbanked women are about as likely as their male counterparts to own a network referrals at higher rates than women. mobile phone. And in some countries, such as Botswana and Zimbabwe, unbanked women are more likely than unbanked men to have a mobile A reliance on networks may be especially important in developing phone. countries due to the high prevalence of small firms and the associated need for sharing capital, the limited development of credit markets, and Bottom line: The gender divide in access to credit is not as strong as it once the insufficient formal enforcement of contracts.228 Experimental studies was, but with smaller asset ownership, women still struggle to get loans of conducted throughout Africa show that encouraging men and female the same size as men – a factor that likely fuels the capital investment gap. entrepreneurs to expand their networks can bring benefits for their businesses.229 Bottom line: With growing evidence on the importance of networks, understanding how the networks of female entrepreneurs vary from those Access to networks and information of men – and how those differences may impact their success – is vital. For VII more on this issue, see “deep dive” 2. Women do not have the same access as men to large and diverse social networks that can support the growth and competitiveness of their business. Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship A long-standing body of research demonstrates the value of good business connections.212 In the early stages of a firm’s development, business networks can influence the size of initial investments and the choice of  Household-level constraints activity.213 Throughout a firm’s life, the diversity of these networks can impact whether an entrepreneur has access to credit,214 learns about new information and market opportunities,215 and acquires the skills needed to Household allocation of productive resources VIII successfully operate their businesses. Peer support can have a significant Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa effect on female entrepreneurs’ ability to learn and apply new business Women often lack authority over the allocation of household assets skills.216 Networks can also provide important channels to customers and and may face more pressure to share resources – restricting both their suppliers, (informal) credit,217 and technology used in production,218 which willingness and ability to invest in their businesses. collectively allow small firms to enjoy economies of scale that would otherwise remain inaccessible. Women’s lack of control over the allocation of household resources may be a source of inefficiency if it means that assets are invested in male- Analysis for this report as well as other research suggests that women’s supported enterprises irrespective of managerial ability or the value of the business networks may differ from men’s in important ways: business opportunity.230 • Men’s and women’s networks are largely segregated by gender.219 The evidence on how African households allocate resources is limited, • Women tend to have network connections with the opposite gender especially when members of these households run several businesses. more frequently than the other way around.220 A study in Madagascar showed that while men have a business premium • Women may act strategically in pursuing business relationships from marriage, women do not – reflecting a husband’s greater ability with men to access the greater opportunities associated with men’s to marshal the labor contributions of household members for their networks. Research from Ethiopia and Uganda highlights the importance business.231 Research from Sri Lanka and Ethiopia demonstrates that of the support of mentors – most of whom are men – in assisting women women with greater decision-making power in the household and more in starting activities in male-dominated sectors. The lack of such cooperative husbands than those with lower scores are more likely to invest connections may limit women’s entry into non-traditional and possibly in their business and more able to grow their enterprises.232,233 However, more profitable activities.221 studies suggest that household behavior tends to follow a non-cooperative • Women’s networks typically command fewer resources than men’s. model. In Kenya, married couples were more likely to invest a monetarily Men are more likely to be connected with organizations occupying significant subsidy in livestock and household assets if received in a joint a central position within the business community, which fosters account than if received in an individual account. When the large subsidy connections with bigger firms.222 went to an individual account owned by either the husband or wife, either was more likely to invest in their own income-generating activities.234 92 93 Furthermore, research shows that female entrepreneurs struggle to direct capital to their business, a fact that can be either due to their own or others’ needs. A study in Ghana shows that giving female entrepreneurs in-kind grants is effective in increasing business profits, but providing them with cash grants is not. This may be because in-kind contributions are easier for women to keep in the business than cash. No such difference is found for male entrepreneurs, for whom both in-kind and cash grants prove effective.235 In-depth qualitative research among micro-entrepreneurs in urban Ghana shows that, even while striving for business success, some female entrepreneurs also try to achieve other non-business goals that may compete for available capital.236 A follow-up study using the same Ghana data, as well as other data from India and Sri Lanka,237 indicates that in households in which there is both a male- and female-owned enterprise, cash grants or microloans tend to be invested in the male enterprise, even when the woman is the cash grant/ loan recipient. When women are the sole household enterprise operator, these investments lead to positive increases in profits. The study concludes that female entrepreneurs are as capable as male entrepreneurs of making sound investment decisions, but they may not be able to control capital in the household. Households may decide to invest in male-owned enterprises because they yield higher returns. However, this difference in returns to capital between female and male-owned businesses may be driven by earlier biases in the allocation of resources rather than by innate differences in managerial ability. Additionally, various studies suggest that women may prefer – or may feel compelled – to spend money on domestic goods or their children238 – a bias that may help explain why in-kind grants are more effective for women-led businesses.239 Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship Inefficiency in intra-household allocation is compounded when female entrepreneurs are compelled to share resources from social connections outside of the home.240 The importance of these dynamics is backed up by an experiment in Uganda. When unconditional grants were distributed to microenterprise owners, the profits of firms owned by women in the treatment group who lived near their family were even lower than the profits of firms owned by similar women in the control group (i.e., those Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa who received no grants or other support).241 Bottom line: Household resource allocation is likely a major factor influencing women’s decisions regarding their businesses. Further research should include the identification of scalable mechanisms for involving couples, men and other male household members to help ease constraints related to the distribution of household resources. See “deep dive” 3. 94 95 Figure 24 Time constraints / Care IX Women spend more time than men on domestic chores, limiting the amount of time that they can dedicate to their business. Male entrepreneurs in Africa spend more time than female entrepreneurs on their businesses Gender gap in hours spent working on a business Research demonstrates that women dedicate substantially more time than men to household tasks242 – sometimes even when they contribute more than their partner to the household’s earnings.243 Women not only devote more time to household tasks, but they may also need to stay at home 18%*** at certain times of the day – even if those times are best for conducting 18%*** Benin 17%** business – or multitask between household and business responsibilities. 8%** 8%** Ghana (1) This report finds that in Uganda, women are three times more likely than 8%* men to take responsibility for household chores and caring for others 13%*** (children, sick and elderly) while operating their business, and 50% more 13%*** Ghana (2) likely than men to be engaged in multiple activities while running their 12%*** businesses. In Malawi and Togo, female entrepreneurs spend twice as 0% Malawi much time as their male peers taking care of children, the sick, or the 1% 6% elderly while simultaneously running their businesses. Moreover, in Togo -2% women spend three times more hours than men on childcare during -2% Mozambique business working hours.244 4% 11%* In agribusiness in the Democratic Republic of Congo,245 female plot 11%* Nigeria (2) 8% managers spend significantly more time than male plot managers on domestic and care work, while also spending significantly less time on 3%* South Africa 3%* productive activities, resting, and conducting social activities than male 3%* plot managers. In male-headed households, being a female plot manager 4%* is associated with spending even less time on productive activities. 3% Togo After accounting for differences in household composition and village 8%** and individual characteristics, women plot managers in the Democratic Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship 1% Uganda (1) Republic of Congo spend on average 20 minutes less per day than male plot -1% -7% managers on farm work and other income-generating activities, and an hour and 45 minutes more on domestic work.In the Democratic Republic -10% 0% 10% 20% 30% of Congo, 64% of plot managers think that having a child under five years of age lowers their productivity – with female plot managers six percentage points more likely to report this strongly adverse effect. This significant difference holds when controlling for household composition, as well as Unconditional gap ... controlling for age and tenure ... Core specification Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa for individual and household characteristics, and village characteristics. While this example is more linked to agriculture than entrepreneurship, it is Source: Authors using IE database. important to highlight given the limited granular data on these dynamics in Note: Ghana (1): Grants for micro-enterprises survey; Ghana (2): Tailoring survey; Nigeria (2): Business plan competition survey; Africa. Uganda (1): Kassida survey. Overall, both women and men entrepreneurs spend a lot of time working on their businesses, exceeding the standard of 40 hours a week. African entrepreneurs spend an average of between 50 and 67 hours per week on their businesses. However, women spend fewer hours than men at work: on average, men work 10% more hours per week than women. There is some variation across countries in Africa (Figure 24). The gender gaps in hours spent on the business tend to be smaller and not statistically significant in male-dominated sectors – perhaps because women in these sectors need to match men’s time commitment in order to compete. 96 97 Household and care responsibilities may limit the time women can dedicate to a business. The burden on women to complete household chores influences their choices in terms of what they can do with the time left, with these choices having implications for their economic decisions. Time is a limited resource – more time working means less leisure, and therefore implies a greater degree of “time poverty.”246 In Ghana, Malawi, Nigeria, and Uganda, being married is a factor that contributes to increasing the gender gap in the number of hours spent working in a business, suggesting that women with husbands and children face greater demands on their time for household tasks. Similarly, having children increases the gap in business performance in Mozambique and South Africa. Freeing women from some household responsibilities could help to expand their economic activities. There is emerging evidence, especially from outside of Africa, about the importance of childcare programs. However, this evidence mostly focuses on the impacts on children or on women’s participation in economic activities, rather than looking at the impacts on firm performance or women’s abilities to manage their businesses. A study in Mozambique finds that childcare services increased the caregiver’s employment rate by 26%.247 An impact evaluation in Kenya showed that mothers of children who received subsidized early childcare were on average 17% more likely to be employed. Women receiving this support were 1.3 percentage points more likely than in the control group to be running a business, an impact smaller than the one in wage employment.248 A quasi-experimental study in southern Togo indicates that Part 3 | Decisions and constraints: A blueprint for closing the gender gap in entrepreneurship enrolling children three to five years of age in preschool resulted in women being 37% more likely to work outside the home.249 Other studies have researched the effect of the price of childcare and women’s labor-market activity in low-income countries, concluding that the high costs of outside-home care have a negative effect on maternal employment.250 In Brazil, mothers’ employment rates increased from 36% Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa to 46% after the introduction of public-funded childcare services, with a significant increase in household incomes (although the cost of childcare services offset these higher incomes).251 In Argentina, a large pre-primary school-building program increased the likelihood of maternal employment by between 7 and 14 percentage points.252 Bottom line: Women’s time constraints related to domestic chores and care are likely a strong constraint on their business activities relative to men, but further rigorous evidence and analysis is needed on whether increasing childcare services and other types of interventions produce lasting effects on business performance. 98 99 04 Deep-dive analyses Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa From the initial list of nine underlying constraints, these issues were selected for deeper analysis for the following reasons: • Rigorous evidence of their importance in explaining gender gaps in business performance in Africa is not comprehensive. • Their measurement is challenging. • The mechanisms involved vis á vis business performance are unclear. • Highlighting these issues in this report can lead to further analytical work being done on them. • Understanding these constraints more thoroughly is critical in driving the next generation of program development. This section examines in greater detail the following underlying Given the limited data available in Africa on these underlying factors, most of the deep-dive analyses rely on surveys conducted in Ghana, Malawi, Togo, and/or Uganda. These countries were selected constraints affecting gender parity: for the deep dives in order to leverage existing impact-evaluation studies in each country, which can (i) social norms, (ii) networks and provide in-depth quality information that is hard to replicate from scratch in a new country. In the three studies, the initial impact-evaluation surveys were complemented with additional quantitative or information, and (iii) household qualitative surveys focused on the deep-dive issues. This makes the descriptive analysis completed in allocation of factors of resources. these deep dives quite unique in the literature on gender and enterprise development in Africa. 100 101 Deep dive 1 Deep dive 1: Figure 25 Do gender-biased social norms Gender-norm biases are common among entrepreneurs dampen business performance? Malawi Togo Agree receiving education more This deep-dive analysis draws on rich data collected for this report in recent surveys in Togo and 13% 22% important for boys than girls Malawi, as well as additional impact-evaluation data from Ghana and Malawi. It assesses the prevalence of gender-biased beliefs and norms and their potential contributions to the gender gap in business performance.253 Disagree men and women should have While this deep dive focuses on a small set of countries and is therefore not representative of all same opportunities 12% 12% of Africa, it is worth noting that in a ranking of countries by the proportion of female ownership represented in formal firms,254 Malawi ranks close to the middle, Ghana at the lower end of the top 10, and Togo ranks near the bottom, at tenth to last. The countries at the bottom of this list are Mauritania, Chad, and Sudan. Those leading this list are Lesotho, Botswana and Namibia. Agree a woman should work in a sector suitable for women 45% 43%  Main findings 1. Gender-biased beliefs are prevalent among surveyed entrepreneurs in Africa. Disagree a woman should work in most profitable sector, whatever it is 30% 26% 2. Both men and women have largely internalized gender-biased norms,255 with some important gender differences. Women are less likely than men to be biased toward the pursuit of business opportunities, but they are more likely to prioritize household needs. 3. Entrepreneurs with comparatively high levels of education are less likely to espouse gender-biased beliefs. Agree men make better managers than women 32% 42% 4. Holding gender biases is associated with lower investment levels and lower business performance, but this relationship is not robust in all settings. There is suggestive evidence in some settings that businesswomen holding more progressive views and those operating in male-dominated sectors face discrimination in their operations. Agree if a woman earns more than 51% 50% husband it’s going to cause problems Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa  Analysis Source: Authors using surveys for this report 1. Gender-biased beliefs are prevalent. • Both male and female entrepreneurs in this analysis have beliefs that • Mobility restrictions, which limit women’s opportunities outside of the can be generally classified as gender-biased. Half of the business home, are undergirded by gender norms that dictate socially accepted owners surveyed in Malawi and Togo believe that problems will arise if behaviors for women. Data from a 2014 survey in Malawi reveal that 32% a woman earns more than her spouse (Figure 25). This is also prevalent of entrepreneurs agree that a woman’s place is in the home and more outside of Africa; a recent study showed evidence of these attitudes than two-thirds agree that when a woman works outside the home, the in the United States, as well as their negative effects on women’s children suffer. employment and earnings potential.256 • Entrepreneurs in Togo and Malawi hold deeper gender biases regarding norms related to business abilities and performance than for norms Part 4 | Deep-dive analyses related to education and opportunities in general. While about 32% and 42% of entrepreneurs in Malawi and Togo respectively agree that men make better managers than women, only 12% of entrepreneurs in these settings disagree that men and women should have the same opportunities. And whereas more than 40% of respondents agree that women should work in sectors that are suitable for them and at least 25% think women should not work in the most profitable sector, only 13% of business owners in Malawi state that education is more important for boys than girls. 102 103 Deep dive 1 Deep dive 1 Box 7 2. Women are less biased than men regarding gender norms as they relate to business opportunities, but Using vignettes to assess business vs. household choices in Togo are more likely to prioritize household needs over the  Summary business. Vignettes allow researchers to tease out subtle differences • While differences in views on overall gender equality tend to be small in attitudes, perceptions, and norms. In a survey in Togo for in Togo and Malawi, men and women diverge markedly over questions this report, the team administered a set of carefully worded relating to men’s and women’s abilities to run a business, the appropriate vignettes to assess the entrepreneurs’ priorities for – and sectors for women’s work, and women’s mobility. biases around – business and family choices. • On average, women hold less powerful gender-biased beliefs than do men. In both Malawi and Togo, women are significantly less likely to agree that men make better managers than women. This difference  Methodology remains significant even after controlling for personal characteristics and sector of operations. Women are also more likely to agree that they Entrepreneurs listened to a vignette about an individual who owns a business and has just received a government should work in the most profitable sector, whatever that may be.257 grant. If the individual were to hire an employee with that money, she/he could double her/his profits. The individual had an alternative option of using the grant to pay for the children’s supplies in school. Respondents were asked two separate questions: 1) whether the individual should hire an employee (and not pay Figure 27 for the children’s supplies); 2) whether the individual should consult with the spouse before making this decision. A randomly selected half of the sample received the vignette question with a female name, and the other half Consistent gender differences around gender-biased norms received it with the individual having a male name. Malawi Togo  Key findings • For both male and female names, 48% thought that the individual should hire an employee, suggesting no bias at this level. Figure 26 Gender biases are prevalent in decision-making Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 100% Female *** name 81% 80% 70% Male 60% name agree if a woman earns more than husband it is going to 48% 48% cause problems 40% agree men make better managers than women 20% disagree a woman should work in most profitable sector, whatever it is 0% agree a woman should work in sector suitable for women Should hire an employee and Should ask not pay for children supplies her/his spouse agree receiving education more Part 4 | Deep-dive analyses important for boys than girls disagree men and women Source: Authors using survey for this report. should have same opportunities Note: *** p<0.01, ** p<0.05, * p<0.10. Tests of equality by gender of the entrepreneur represented in the vignette. Women Men Source: Authors using surveys for this report • Gender biases emerged around decision-making. About 81% of those who received the vignette with a female Note: *** p<0.01, ** p<0.05, * p<0.10 Tests of equality by gender. Unconditional gender differences reported. name replied that she should consult with her husband before making the decision. This was 11 percentage points higher than when respondents were asked whether the husband should first consult his wife. 104 105 Deep dive 1 Deep dive 1 • Women are also significantly less likely than men to support norms Figure 29 restricting women’s mobility. While only 14% of female entrepreneurs in Malawi (2014) agree that a woman’s place is in the home, 43% of male business owners think that this is the case. And three-quarters of male Women tend to prioritize household needs more strongly than men entrepreneurs agree that children suffer when a mother works outside the home, compared to 59% of women business owners. Female name Male name Figure 28 100% * ** 86% Gender differences relating to biases concerning women’s mobility 80% 77% 78% (Malawi) * 64% * 59% 60% 50% *** Agree a woman’s place is Women 47% in the home 38% 40% Men 20% 0% Should hire an employee and Should ask Should hire an employee and Should ask not pay for children supplies her/his spouse not pay for children supplies her/his spouse *** Agree when a mother works outside the home, the children suffer WOMEN MEN Source: Authors using surveys for this report Note: *** p<0.01, ** p<0.05, * p<0.10. Tests of equality by gender of the entrepreneur represented in the vignette, separately by gender of the respondent. Unconditional gender differences reported. *** Agree being housewife is as fulfilling as working for pay 0 10% 20% 30% 40% 50% 60% 70% 80% Source: Authors using surveys for this report Note: *** p<0.01, ** p<0.05, * p<0.10 Tests of equality by gender. Unconditional gender differences reported. Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa • However, women tend to prioritize family over business more than men 3. Entrepreneurs with comparatively high education do. The Malawi survey from 2014 asked male and female entrepreneurs levels are less likely to espouse gender-biased beliefs where they see themselves on a scale between preferring family or work: 27% of women and 19% of men see themselves closer to family than to • A measure of gender-biased beliefs is constructed for this report work.258 that aggregates six statements asked in the Togo and Malawi surveys (see Figure 25 for the specific statements). The report then analyzes • In the vignettes from Togo, both women and men hold biases in favor the relationship between personal characteristics and these beliefs. of their own gender. That is, they are both more likely to agree that It defines the gender-biased norm index as the proportion of the six entrepreneurs of their own gender should prioritize their business statements with which the individual agrees. The average of the index in (instead of school supplies) as compared to individuals of the opposite both Togo and Malawi is about 0.33; thus, on average, respondents hold gender (Figure 29). Nonetheless, women appear to prioritize household close to two of the six gender-biased beliefs.260 needs: 50% of women think the female name should hire an employee, while 59% of men agree the male name should hire an employee. Only • Figure 30 shows that comparatively more educated individuals – both 38% of women believe that the male name should hire an employee, Part 4 | Deep-dive analyses men and women – are less likely to be gender-biased. However, the when 47% of men think the female name should hire an employee. negative correlation is especially evident at higher levels of education, suggesting that primary education alone is not associated with weaker • Women are also more likely than men to agree that entrepreneurs should gender-biased beliefs. consult their spouse, independently of whether the entrepreneur is male or female.259 • Regression results confirm that education is a strong and statistically significant correlate of gender-biased beliefs for both men and women. Moreover, consistent with the previous section, women hold significantly fewer gender-biased beliefs than do men, as an overall average.261 106 107 Deep dive 1 Deep dive 1 Figure 30 Table 5 Gender bias of entrepreneurs declines as education increases Being male and having a low level of education are the main factors associated with holding gender-biased beliefs Malawi Togo MALAWI TOGO Dep. variable: gender- (1) (2) (3) (4) (5) (6) biased beliefs index ALL WOMEN MEN ALL WOMEN MEN -0.144*** -0.061** Woman (0.022) (0.026) -0.002* -0.001 -0.001 0.001 0.001 0.001 Owner’s age (0.001) (0.001) (0.001) (0.001) (0.001) (0.002) Years of education -0.016*** -0.008** -0.022*** -0.008*** -0.006* -0.008* completed (0.003) (0.004) (0.004) (0.003) (0.003) (0.004) -0.026 -0.054* 0.009 0.019 0.015 0.026 Married (0.029) (0.031) (0.059) (0.024) (0.031) (0.045) Source: Authors using surveys for this report Female Male 0.004 0.008 0.001 0.001 -0.001 0.003 Household size (0.003) (0.005) (0.004) (0.004) (0.006) (0.006) -0.002 -0.011 0.001 0.015 0.024 0.005 Number of Young children (0.012) (0.012) (0.018) (0.012) (0.018) (0.017) 0.531*** 0.315*** 0.528*** 0.341*** 0.292*** 0.338*** Constant (0.081) (0.103) (0.122) (0.059) (0.077) (0.087) Sector dummies Y Y Y Y Y Y Observations 484 195 289 480 233 247 R-squared 0.186 0.133 0.150 0.047 0.046 0.067 Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 4. Gender-biased beliefs are associated with lower investment and performance levels. These relationships are partly explained by other factors such as education levels. • In Malawi and Togo, gender-biased beliefs are negatively associated with firm-level capital investment (Figure 31). Holding more gender- biased beliefs is associated with a 70% to 80% difference in capital investment in Malawi and Togo. In Malawi, this relationship is reduced Part 4 | Deep-dive analyses and statistically insignificant after controlling for the entrepreneur’s characteristics (such as years of schooling) and the sector of operation, but remains very large for women.262 It may also be the case that norms reduce investment by pushing women into specific low-investment sectors of operation. In Togo, gender-biased beliefs are associated with lower levels of capital investment even after accounting for the sector and the entrepreneur’s characteristics. No similar pattern emerges in either country when examining the number of hours worked, which is not correlated with gender norms. 108 109 Deep dive 1 Deep dive 1 • As with investment levels, the inverse relationship between gender bias Figure 31 and profits remains large in Malawi. Holding more gender-biased beliefs is associated with profit levels that are on average 52% lower, with this effect rising to 68% lower profits among female entrepreneurs in Malawi. Gender-biased beliefs associated with lower levels of capital The correlation between the two factors is not statistically significant but investment remains greater than 30% when accounting for individual characteristics and the sector of operations. The importance of individual characteristics in mediating the association between norms, investment -71%*** Unconditional and profit levels may be due to the role of norms in influencing (and being influenced by) key life choices – such as educational attainment, -24% Controlling for sector and individual characteristics marriage, and the number of children. -77%* Women - unconditional MALAWI • For female entrepreneurs in Togo, on the other hand, the relationship between gender biases and profits is positive. The correlation is -50% Women - Controlling for sector and individual characteristics statistically insignificant when one accounts for other characteristics, but is still large.263 Further analysis points to occupational segregation -68%*** Men - unconditional as the driver of this result. The positive and significant relationship between bias and women’s profits is concentrated within sectors where -25% Men - Controlling for sector and individual characteristics women are not highly represented. These findings indicate that female entrepreneurs in Togo who hold comparatively progressive views on gender equality and operate outside “traditional” female sectors may be marginalized or face social discrimination. This is supported by the fact that they show lower profits than those with more biased beliefs, even -81%*** Unconditional after accounting for individual characteristics, sector, and firm inputs and characteristics. -73%*** Controlling for sector and individual characteristics -84%*** Women - unconditional TOGO -73%** Women - Controlling for sector and individual characteristics 5. Conclusion +77%*** Men - unconditional -67%*** • A large share of both male and female entrepreneurs in Malawi and Togo Men - Controlling for sector and individual characteristics hold gender-biased beliefs. Men are more likely to hold biases regarding how and in what sector women should run a business. Women, meanwhile, are more likely to value family over business, which may influence their entrepreneurial decisions. Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Source: Authors using surveys for this report Note: *** p<0.01, ** p<0.05, * p<0.10. Individual characteristics include gender, age, years of education, marital status, household • Educational attainment is associated with a reduced level of bias, but a size, and number of young children. Conditional regressions also include sector dummy variables. Investment measured using relatively high level of education is needed before these biases become inverse hyperbolic sine.269 less common. • Holding more gender biases is associated with lower levels of capital investment in Malawi and Togo and lower profits in Malawi. In addition, norms may be so entrenched in life choices that it may be the case that factors helping explain the relationship between norms and business performance are also influenced by these biases. • There is limited causal evidence with regard to how strongly social norms impact business performance. A research agenda is needed to test the mechanisms and importance of social norms as constraints to business development. In particular, rigorous impact evaluations of interventions seeking to address targeted gender biases – or at Part 4 | Deep-dive analyses least circumvent them – can provide crucial insights into the relative importance of this factor. • This research agenda should build on existing evidence in parallel areas of study. Solutions for tackling embedded social norms in the context of enterprise development can be drawn from a number of promising areas of research on norms, including large-scale institutional changes,264 changes in regulations,265 addressing mobility issues,266 overcoming sector sex-segregation issues,267 and community and couple dialogues.268 110 111 Deep dive 2 Deep dive 2: Figure 32 Do gender differences in Women more consistently report being helped by a person of opposite sex entrepreneurial networks fuel the Women Men gender performance gap? This deep-dive analysis assesses gender differences in African entrepreneurs’ business networks Share of respondents who report and their potential contributions to the gender gap in business performance. It draws on a range of being helped by a person of the detailed surveys collected for impact-evaluation studies (Ghana, Malawi, Togo, and Uganda), as well as opposite gender other than their *** two surveys conducted in Togo and Malawi in preparation for this report. spouse in the past 12 months Share of respondents who report being helped by a person of the  Main findings opposite gender other than their *** spouse when they started their 1. Female entrepreneur’s business networks are mostly comprised of other women. business 2. Men have larger business networks than women. 0% 20% 40% 60% 0% 20% 40% 60% 3. Business networks with “strong ties,” and family members in particular, are part of the process of business creation and development for both male and female entrepreneurs. Women are more MALAWI TOGO likely to use this type of networks. 4. Women rely on their networks when starting a business and for financial support, but men more intensively use their business networks to share information, equipment, and supplies. 5. The relationship between networks and business performance is likely not straightforward, and dependent on aspects such as the depth of the relationships and their influence on various areas of Source: Authors using surveys for this report. Note: *** p<0.01, ** p<0.05, * p<0.10. business development.  Analysis 1. Female entrepreneur’s business networks are 2. Men have larger business networks than women. Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa mostly comprised of other women. • In Malawi,271 female entrepreneurs have 18% fewer people in their • Consistent with earlier evidence,270 the analysis for this report suggests networks than do male entrepreneurs. A similarly sized gap (23%) that African entrepreneurs’ business networks are strongly segregated emerges when considering networks within the entrepreneur’s sector. by gender; that is, men interact primarily with other men, and women This difference between the sizes of within-sector networks is 19%, with other women. For example, in Uganda, women account for nearly and is statistically significant after accounting for firm and owner 80% of women’s networks, but less than 10% of men’s. characteristics. • Women’s networks appear to be more gender-diverse than men’s. In • In Uganda, male entrepreneurs’ networks tend to be twice as large as Uganda, women are more likely than men to list at least one person of those of female business owners. And men reported an average of 3.2 the opposite gender in their networks. In Togo and Malawi, only a small members in their core business network versus 2.9 for women. Despite minority of men and women indicate having received support from the differences in network size, women and men in Uganda appear to people of the opposite gender (other than their spouse), either when they be equally satisfied with their networks: 72% of women and 77% of men started their business or in the past 12 months. In Malawi, women are agree that networks are very important, and 77% of all entrepreneurs (no more likely to report having benefited from this support (Figure 32). gender differences) state that the current level of information sharing is Part 4 | Deep-dive analyses optimal. • Among entrepreneurs who report knowing a specialist, women are more likely to point to a male specialist than men to a female one. The gap is particularly large in Malawi, where only 2% of the men named a female specialist, while almost 40% of women named a male one. 112 113 Deep dive 2 Deep dive 2 3. Business networks with “strong ties,” and family Figure 34 members in particular, are part of the process of business creation and development for both male and Married women more likely to receive support from husbands than female entrepreneurs. Women are more likely to use vice versa this type of networks. Women Men • In Uganda, family and friends provide support to business owners. Family accounts for approximately 17% of women and men’s networks, while friends account for roughly 65%. The large majority of any given Share of respondents who report entrepreneur’s business connections in Uganda are within his or her being helped by their spouse in the *** *** business or residential area. past 12 months • In Togo and Malawi, most entrepreneurs receive help from relatives Share of respondents who report when starting their businesses. Women are more likely than men to being helped by their spouse when receive help from relatives at the start, and report less support from they started their business *** *** non-relatives. • Men and women also appear to rely on different categories of relatives. 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Men are more likely to receive assistance from their parents, but also from other relatives including siblings. Women, on the other hand, often MALAWI TOGO receive assistance from their spouse. These patterns are consistent in both countries, and remain similar not only at a business’ start, but also with regard to subsequently provided assistance (Figures 33 and 34). Source: Authors using surveys for this report. Note: *** p<0.01, ** p<0.05, * p<0.10. • In Togo, women are more likely than men to hire employees based on 272 social connections rather than skills or degree (17% vs 4%). Figure 33 4. Women are more likely to use networks when starting their business and for financial support, but Men and women have different composition of networks men use their business networks more intensively to share information, equipment and supplies. Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Women Men • In Togo and Malawi, most entrepreneurs indicated that they had access to networks when launching their business, and women were more Share of respondents who report likely to have received help at that stage. Female entrepreneurs in Togo *** *** being helped by a non relative are also more likely than male ones to search for advice if they launch a new product or service. This is in line with findings from South Africa, where female-owned enterprises that had discussed partnerships Share of respondents who report with other firms are significantly more likely than male-owned to be being helped by other family *** a larger firm.273 However, the importance of networks for operating a members (siblings, son/daughter, business may decline over time. In both Togo and Malawi, more than cousin, uncle/aunt) 60% of entrepreneurs indicated that they did not receive support in the preceding 12 months. Share of respondents who report *** • Women are more likely than men to benefit from financial support being helped by their parents from their networks. The surveys in Togo and Malawi for this report Part 4 | Deep-dive analyses consistently show that women more frequently receive financial assistance from their networks, both when they launch their business and in the previous 12 months (Figure 35). This difference remains 0% 20% 40% 60% 0% 20% 40% 60% significant even when controlling for sector, marital status, and level of education. However, in Uganda, men appear more likely than women to MALAWI TOGO cooperate financially with their network: 58% of men and 48% of women report this type of cooperation (Figure 36). Source: Authors using surveys for this report. Note: *** p<0.01, ** p<0.05, * p<0.10. 114 115 Deep dive 2 Deep dive 2 • Men are more likely to share equipment, inputs and information. More Figure 36 men than women report having been taught a skill in Malawi, and having shared equipment when they started their business in Togo. Men in Togo are more likely to know a ‘specialist’, meaning “somebody with Men in Uganda more likely to share information on equipment superior management or technical knowledge relevant to the activity.” This seems to be driven by the sector of operations, with male-owned firms typically in sectors seen as “more technical.” In Uganda, over 90% Women Men of the entrepreneurs share information on techniques and skills. Men appear more likely than women to share information on equipment, reflecting both differences in the sector of activity and in the nature of 100% their interactions (Figure 36). 80% • Women receive different types of support by gender: they are more likely *** to obtain capital (as a loan or a gift) from men, and are more likely to 60% be taught a skill by other women. This is true across surveys in Ghana, Togo, and Malawi (Figure 37). 40% 20% 0% Shares info on Shares info on Shares inputs Shares Financial techniques/skills equipment and equipment clients cooperation Source: Authors using IE database. Note: *** p<0.01, ** p<0.05, * p<0.10. Figure 35 Female and male entrepreneurs receive different types of assistance Women Men Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Share of respondents who report having been taught a skill (at any ** point) Share of respondents who report having received financial assistance (at any point) *** ** 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% MALAWI TOGO Part 4 | Deep-dive analyses Source: Authors using surveys for this report. Note: *** p<0.01, ** p<0.05, * p<0.10. 116 117 Deep dive 2 Deep dive 2 Figure 37 • There are several potential explanations for why the relationship between networks and business performance is not straightforward: Women receive relatively more financial support from men and more ›› More than the size of the networks, the quality of these relationships technical support from other women may be more important for business development. ›› The analysis for the report excludes failed businesses, where the lack of networking support may have been most important. TOGO 2016 MALAWI 2017 GHANA 2016 ›› Network effects may be captured through other dimensions 80% considered in the analysis. In Malawi,276 gender differences in network 70% size lose their significance after controlling for sector and owner characteristics, suggesting strong associations between network size 60% and these other dimensions. 50% ›› Women – particularly those with large and diverse networks – may 40% be prevented from reaping the benefits of their networks by other constraints. These include gendered social norms and the decision- 30% making process within their household. 20% ›› Entrepreneurs – both male and female – may make additional investment in their networks to compensate for an initially low 10% level of financial or human capital. This is known in the networks 0% literature as the “network compensation hypothesis,” and implies Have been Have received Have been Have received Have been Have received that the data could show a negative correlation between networking taught a skill financial assistance taught a skill financial assistance taught a skill financial assistance and performance even when networking actually helps business development.277 Source: Authors using IE database and surveys for this report. ›› Having an extensive network may be a mixed blessing, with additional Note: *** p<0.01, ** p<0.05, * p<0.10. by/from a woman by/from a man costs imposed on entrepreneurs. Previous research278 shows that family pressure can have adverse effects on female-owned businesses. Women who receive more support from their network may also face more demands on their own resources from the members of this network. These two effects may cancel each other out, so that women with particularly supportive networks may end up not performing 5. The relationship between networks and business better than women receiving less assistance. performance is not straightforward, and is dependent on aspects such as the depth of relationships 6. Conclusion and their influence on various areas of business development. • The analysis for this report indicates that entrepreneurs’ networks in four different countries in Africa are strongly segregated by gender, and Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa • The evidence from rigorous evaluations suggest networks can play a women’s networks are different from men’s in important ways. particularly important role in reducing transaction costs, by fostering trust, information sharing, and supporting alternate enforcement • Men’s networks are larger, and evidence suggests that they are more mechanisms. An impact evaluation in Ghana found that increasing likely to provide opportunities for sharing equipment. Women often rely access to a new business network had a positive effect on business on their spouse, in particular for financial support. performance.274 Still, women were less likely than men to have interacted with members of their network. An experiment in Ethiopia, Tanzania, and • Additional experiments creating networking opportunities for different Zambia found a positive effect with regard to the diffusion of business categories of entrepreneurs would be helpful in identifying the type practices for the managers of small firms assigned to collaborate with of businesses that stand to benefit the most from diversifying their large firms.275 connections. The following questions seem important at this juncture: • This report analyzed the correlation between the characteristics of ›› Is networking more important in the early or later stages of business entrepreneurs’ networks and the performance of their businesses. development? The analysis first considers the relationship between network size and ›› To what extent do other gendered constraints dampen the benefits of business performance. This report does not find a significant correlation expanding one’s networks? ›› Why do businesses networks remain so strongly segregated by Part 4 | Deep-dive analyses between network size and profits. In Togo and Malawi, entrepreneurs who report having received help when they started their business do gender? not appear to be performing any better than those who indicated having ›› Are there opportunities to loosen the networks’ “ties” across gender received no such help. In Uganda, entrepreneurs belonging to networks and family presence? in which equipment was shared do not appear to be performing better ›› What is the effect of network diversification on entrepreneurs’ choice than others. In Malawi and Togo, female entrepreneurs who have been of activity? taught a skill by someone in their network appear to have higher profit and capital investment levels. For men on the other hand, no such relationship was observed. 118 119 Deep dive 3 Deep dive 3: • Husbands in Ghana seem generally supportive of their spouse’s business. Perhaps because of the importance of women’s businesses for meeting household needs, only 10% of female entrepreneurs report that Do intra-household relationships their husbands are unsupportive of their decisions to run businesses. Qualitative research also finds a widely shared expectation that women affect the strategic choices of female will operate businesses and receive support from their husbands to make this happen, especially at the time of business start-up. Recent research entrepreneurs? in Ethiopia suggests that while some husbands constrain women’s business activity, many husbands provide various forms of support that are important for women’s entrepreneurship activities.280 This section draws on a survey of micro-entrepreneurs in urban Ghana, along with complementary qualitative research examining the same population. These data permit an in-depth analysis of how women’s businesses are affected by relationships within their households. The findings are specific to urban Ghana, but suggest ways that intra-household dynamics might be influential in women’s 2. Spouses have incomplete information about each business decisions elsewhere. other’s earnings. Undoubtedly, the burdens of meeting household needs place pressure on both men’s and women’s • Previous studies have documented a lack of cooperation between businesses, especially in low-income households. The data suggest, however, that female spouses in maximizing household income.281 In the Ghana survey, entrepreneurs may face additional challenges because: when asked to guess their spouse’s income, only about 32% of female 1. How much and when they use business income for household needs is often determined by factors entrepreneurs and 21% of their husbands guessed a figure that outside their control; and was within 20% of the actual figure. In the qualitative study, female 2. Additional visible business income can mean reduced support from the spouse for meeting entrepreneurs indicated that they did not share information about their household needs, which can increase the financial burden on women. income with their husbands, and their husbands did not share it with them. Relatedly, pooling income was also unexpected and viewed as uncommon. • Husbands of female entrepreneurs are more likely to overestimate their  Main findings spouse’s earnings. Figure 38 shows the frequency of income over- and underestimates by spouses when both members of the couple are 1. Women’s businesses are important for meeting household needs. entrepreneurs. Nearly 40% of husbands of female micro-entrepreneurs 2. Spouses have incomplete information about each other’s earnings. overestimate their wife’s income, while 40% underestimate the amount. 3. Women generally maintain control of their business income, but that does not always mean they Women in the sample have somewhat more accurate information about have flexibility in determining how to spend it. their spouse’s earnings, but some of the husbands in this sample are 4. Women and men have incentives to hide their income due to pressure to share it and to increase salaried, which means their income is easier to guess. contributions to household needs. 5. Women’s independence in business management is associated with higher profits, but it is still unknown whether this is a causal relationship. Figure 38 6. Households manage their finances in a range of different ways, and future research and programs Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa should explore links between intra-household relationships and the success of women’s businesses. Most female and male entrepreneurs in Ghana misestimate their spouse’s income Gender gap in hours spent working on a business  Analysis 80% Wife’s estimate of husband income 1. Women’s businesses are important for meeting 60% Husband’s household needs. 42% estimate of 39%*** 40% 40% wife income • More than three-quarters of women micro-entrepreneurs in urban 32%*** Ghana agree that their household would have a hard time paying for 26% Part 4 | Deep-dive analyses 21% food or school-related expenses without the income earned from their 20% business. Surveys in other countries confirm the importance of women’s businesses to households.279 This runs counter to the popular narrative, documented in qualitative fieldwork in Ghana, that the man is the 0% primary breadwinner and that it is his responsibility to pay for housing, Accurately estimate her/ Over-estimate her/his Under-estimate her/his school fees, food, and medical expenses. his spouse income spouse income spouse income Source: Authors using IE database. Note: *** p<0.01, ** p<0.05, * p<0.1 for gender differences on estimates. 120 121 Deep dive 3 Deep dive 3 3. Women generally maintain control of their income, but that does not always mean they have flexibility in determining how to spend it. • Only 8% of female entrepreneurs say that they do not have control over how their own income is spent. • Direct economic coercion is relatively uncommon as well: less than 10% of female entrepreneurs in Ghana say their husbands compel them to give money in the previous 3 months, and nearly all of these women report that they usually control their income. About 11% anticipate that their husband would take some of their money if they received a grant to expand the household’s economic opportunities. • Despite their relative degree of control over their own income, female entrepreneurs have less say than their husbands over how household needs are met. Drawing on 16 longitudinal surveys from 12 countries (including eight surveys in sub-Saharan Africa), a recent study finds that small firms operated by women are more likely than male-owned firms to close because of household shocks (such as illness or other family needs). In fact, when explaining the reason for business closure, women report illness or family needs (34%) almost as frequently as they cite making a loss (36%). In contrast, only 12% of men mentioned family reasons as explanation for firm closure.282 • Qualitative research in Ghana suggests women receive money from their husbands to meet household needs, but that the money is often inconsistent and in some cases inadequate. Quantitative data indicates that 81% of women receive regular transfers from their husbands for household maintenance, called “chop money.” Corresponding estimates from Malawi (84%) and Togo (91%) are similarly high. In Ghana, on average, the amount of chop money given to female entrepreneurs every month is roughly equal to the average level of profits for women’s businesses, suggesting that on average it doubles women’s business income.283 • In Ghana, 91% of husbands decide unilaterally how much to give to Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa their wife for household needs, without consultation between the spouses. This uncollaborative financial-management arrangement can leave women faced with hard tradeoffs between meeting business or household needs. The requirement that they be ready to make up for unpredictable shortfalls in their spouse’s income or household contributions also means that entrepreneurs highly value saving for emergencies and insist on retaining access to liquid savings, even when they want to make a business investment. These findings are in line with a previous study in Ghana that documented that giving women in-kind grants resulted in increased business profits, but that giving cash grants – which can be more easily diverted for household priorities – did not have the same effect.284 4. Women have fears about showing how much they Part 4 | Deep-dive analyses make. • As documented elsewhere in Africa,285 entrepreneurs have reasons not to disclose their income levels. In a lab experiment in Tanzania, women were less inclined than men to share income information with their spouse.286 They seem to face family pressure to share income; indeed, about half of the female entrepreneurs in Ghana agree that husbands or other family members end up requesting money whenever they have it on hand. More than three-quarters agree that, in general, people who do 122 123 Deep dive 3 Deep dive 3 well are likely to receive requests from family and friends to help out with Box 8 expenses. Consequently, over two-thirds of women agree that machines and equipment held by their businesses are a good way of saving money, so that others do not take it. Men face similar pressures to share, and are Vignette on allocation of resources in household equally unlikely to be transparent about their income. The survey in Ghana asked respondents to identify with one of the following three women in Ghana: Rejoice, • Women in Ghana reported a need to hide income from their spouse, Grace, or Lydia. in particular, because of the fear that their husbands would reduce their contributions to the households if they knew that their wives had money.287 In the quantitative data, around 65%-70% of women say their Rejoice Grace Lydia husband does not know how much they have in their savings. 100% 80% 60% 5. Women’s independence in business management 47% is associated with higher profits, but more research 40% 33% is needed to understand whether higher profits are a 20% 20% result of that independence. 0% • Consistent with previous evidence,288 limited decision-making power is associated with fewer hours of work and lower profits among women. Rejoice is a Grace is a businesswoman. Lydia is a businesswoman. Female entrepreneurs in Ghana have lower profits on average if they businesswoman. Before Unlike Rejoice, other She is different from Grace need to ask their spouse permission to start a business (43% of the she can decide how much household members pay and Rejoice because she female entrepreneurs surveyed are in this category) or to purchase of her income she can for the same household decides how much of her equipment (28%); if they cannot make decisions about how to spend their spend on inventory and expenses each month. income to spend on inventory own income (8%) or household income (27%); or if they cannot make supplies for her business, Since she has the same and supplies for her business their own decisions about borrowing money for the household (50%). A she first has to wait to responsibilities each month, after discussing household lack of independent decision-making power over household income is see how much other she can decide on her own needs and income with other also associated with lower levels of capital investment in the woman’s household members how much of her income household members. They sit will contribute to meet to spend on inventory and together and determine how business. household needs. [20% of supplies for her business. much each of them should the sample] She does not have to wait to contribute to household needs • Perhaps more surprisingly, female entrepreneurs also do better on see what other household based on what the needs are average when their spouse does not provide certain types of support for members will contribute that month, what each of them their business. Those who report that their husband does not help when to meet household needs. earned, and each of their Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa they have a difficult month (7%) have higher profits, and those whose [47% of the sample] priorities for their businesses. husbands do not provide money for making household purchases (19%) [33% of the sample] also have higher profits. In addition, female entrepreneurs who think that their husband would decrease his contribution to the household if he knew that his wife had money have higher average profits (11%) than those who do not. This relationship could be due to women with higher profits needing less support from their husbands, or it could be due to women’s greater independence, leading to more investment on their own. • Related findings can be seen in the results of a vignette survey question that asked female entrepreneurs to identify with one of three illustrative Female entrepreneurs who identify with Grace have higher profits descriptions of women’s intra-household financial management. Nearly than those who identify with Lydia or Rejoice. Grace is different half of the female entrepreneurs identified with a story that depicted a from Rejoice and Lydia in two important ways: she is relatively woman (Grace) whose income is kept separate from her husband’s and independent from her husband and she has predictability in her who has a distinct set of responsibilities within the household. One-third financial responsibilities. It is unknown whether the conditions of identified with a story about a woman (Lydia) who has a cooperative Part 4 | Deep-dive analyses independence and predictability lead to higher profits or whether relationship with her husband and who sits with her husband to those with higher profits are able to generate those favorable determine the best way to spend and invest the money available. A final conditions. 20% identified with a story of a woman (Rejoice) who has little control – she waits to see what her husband contributes to the household and then Those who identify with Lydia – a more cooperative model of household financial management – have lower profits than others. makes her decisions about consumption and investment. Given that these are female entrepreneurs who collaborate with their husbands to make consumption and investment decisions, it is possible that they are making strategic choices to maximize household income, which may not result in investment in the woman’s business. 124 125 Deep dive 3 6. Conclusion • Households manage their finances in a range of different ways, and future research should explore links between intra-household relationships and the success of women’s businesses. • With different businesses in the household, these firms and their owners face different needs and constraints. Larger or more formal businesses may require higher capital investment to remain competitive, while smaller or informal firms may require day-to-day working capital to maintain operations. This starting point will affect decisions in the household in the allocation of resources. • Business decisions by female entrepreneurs are affected also by how they and their husband manage all sources of income and how they distribute expenditure responsibilities. In Ghana, it is common for husbands and wives to maintain separate income streams and to be responsible for different expenses. Income transparency is not expected and not practiced. Most married women are given housekeeping allowances by their husbands and are expected to meet daily household needs using the allowance as well as their own income when necessary. This can mean that when and how much women use business income for household needs is often determined by factors outside their control. There is a need for research to examine whether these patterns are common in other contexts. • In-depth qualitative research in Ghana shows that women micro- entrepreneurs strategically manage their income to meet both business and household needs. While striving for business success, women also pursue strategies to ensure continued support from their husband, to shore up their ability to meet daily household needs, and to plan for long- term security, all of which can lead to the de-prioritization of business investment.289 Some women fear that additional visible business income will mean reduced support from their spouse for meeting household needs. • Not all households exhibit the same level of cooperation, however, and differences may matter for women’s business decisions. In Ghana, survey Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa responses indicate that there is variation in how households manage their finances. The data also suggest that women’s independence in business management is associated with greater profits, but it is unclear whether this is a causal relationship in either direction. Recent studies in other contexts have documented the importance of variation in levels of cooperation for household investments.290 • Future research should examine the extent to which different financial- management practices within households can be encouraged, thus leading to business investments by female entrepreneurs. Additional research is also needed on how behavior in households where entrepreneurs are not married or living with a partner differs from households where entrepreneurs are married. • More work is needed to identify possible policy responses either to think Part 4 | Deep-dive analyses differently about the role of women’s businesses in the household, or to support female entrepreneurs in the simultaneous achievement of their business and non-business goals. 126 127 05 Path forward: How policy and decision makers can act to eliminate the gender gap Despite high levels of entrepreneurship among African women, their businesses lag behind those owned by men. This report illustrated this performance Across the region and globally, policies and understanding291 why some interventions help Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa gap and examined the differences programs supporting businesswomen abound. A female entrepreneurs and others do not. The few of them have been rigorously evaluated. This available evidence is classified as follows: in strategic decisions and section: underlying factors behind it. • Credible: The program or policy is supported • Provides an overview of existing programs by strong evidence from impact evaluations supported by governments, development and inferential research. It addresses a Gender-neutral policies that partners, private companies, and NGOs gender-specific constraint, and there is seek to create a more conducive to address gender inequality in enterprise more than one rigorous study in Africa with a development in Africa; positive impact on female entrepreneurs. business environment and support • Presents policy recommendations grounded • Emerging: A small body of rigorous evidence entrepreneurs will not be sufficient in impact evidence to address the gender- is available, but further testing is required. specific constraints to enterprise development; The program or policy addresses a gender- to address these constraints – • Identifies common features associated with specific constraint, but there is only one or no and may instead widen existing the successful design and implementation of impact evaluations from Africa. At least one gaps. But the good news is that programs to promote female entrepreneurship rigorous study in another region or a study in Africa; in the region without gender-disaggregated with targeted strategies, the • Proposes areas to improve knowledge of what results suggests a potential positive impact. international community, national works to support gender and entrepreneurship • Not promising: Policymakers should not and makes recommendations on the role of pursue these policies and programs in the governments, NGOs, and other governments and development partners in this absence of new positive evidence. There is stakeholders can help to address area. at least one impact evaluation in Africa with the specific challenges that female gender-disaggregated analysis, and the This chapter builds on the previous analysis to impacts on the business performance of entrepreneurs face and unleash investigate how various interventions address female-owned firms are at best mixed. their productive potential. gender-specific constraints, which is critical to 128 129 The chapter provides practical guidelines for Box 9 policy design and implementation. Actual project examples are provided as case studies. The chapter also outlines what needs to be Other policy areas that support While this report focuses on female-owned firms, several related policy areas also include the promotion of female entrepreneurship and are critical for Africa’s shared growth and poverty reduction. studied in future impact evaluations, providing a female entrepreneurship road map for future analytical work. The chapter also identifies – as much as it can – success factors, potential risks, and mitigating measures. While evidence is typically based on a set of countries, and generalizing to others is not necessarily straightforward, local adaptation will be important as a means of ensuring sound implementation of the most promising ideas. This includes aspects that can improve the targeting, uptake, or effectiveness of policies among female-owned firms, even for a. Active labor-market b. Graduation programs and cash transfers c. Support to women programs also benefiting male entrepreneurs. programs for young for women in ultra-poor households farmers Combining interventions that address multiple women reinforcing constraints to female entrepreneurs The constraints driving the performance Graduation programs for the very poor are becoming Women represent almost half may also be critical for progressively reducing gap for female entrepreneurs are similar to increasingly popular. These programs usually target of the agricultural labor force in the gender gaps in business performance. those faced by women – especially young women in the poorest households, offer a combination Africa, a critical sector for the women – seeking access to employment, of training, cash or in-kind asset transfer, and technical continent’s growth and poverty Moreover, while the evidence presented on including self-employment. These include assistance, and may also include savings mechanisms, reduction. However, there is a the most-promising interventions is typically gaps in skills and capital relative to health information, and life skills training. Mounting significant gap in the productivity based on the results for the average female- men, and the underlying social norms global evidence on the impact of these “wrap-around” of male and female farmers, owned businesses, there are often women about women’s roles. A recent review programs points to them as a cost-effective way to as reported in a recent World of experimental or quasi-experimental increase household assets, consumption, and food Bank/ONE Campaign study.301 who can benefit more than others from each evidence of policy interventions to increase security.295 The best known of these studies analyzed The gap is driven by unequal credible or emerging solution. In particular, the employment of young women in such programs in six countries and found significant access to and use of inputs such the presence of an adequate infrastructure and Africa292,293 finds mixed evidence on skills positive impacts on household assets, consumption, as land, labor, seeds, fertilizer, business environment that ensures targeted programs, including vocational and business income and revenues, among other factors.296 and equipment. In addition, female entrepreneurs have appropriate access skills. Fewer studies have evaluated the women farmers are constrained Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap to markets and inputs are critical aspects impact of constraints related to caregiving Graduation programs that involve providing women by lower education levels and to assess before introducing specific firm- responsibilities, occupational choice, with productive assets such as a goat or a cow can be limited access to markets. level interventions. Being careful in ensuring mobility and safety. The most promising expensive, but evidence shows that the benefits typically The report identified 10 policy replications of successful interventions are results are those that simultaneously tackle outweigh the costs. In some instances, large cash grants areas in which addressing key multiple constraints, including providing have also worked well to increase the income of very constraints could narrow this taking place among similar target groups technical and vocational education and poor households. For example, young women297 in very gap. Policy recommendations of firms/business owners can, in principle, Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa training (TVET), combined with “girl-friendly” poor households in northern Uganda received relatively include strengthening land increase the chances of success. aspects to address limited networks, care large cash grants (equivalent to 42 times the median rights; addressing labor obligations, and capital constraints. Among monthly income).298 The program doubled the monthly constraints by enabling women these, a randomized controlled trial of the earnings of women engaging in small retail activities by to hire more paid labor; Empowerment of Adolescent Girls and spurring changes in the women’s occupational choices. increasing women’s access to Young Women (EPAG) project in Liberia But despite the positive economic outcomes, there was mechanization and providing showed impressive results, increasing no evidence of increased independence or community childcare centers; increasing employment by 47% and earnings by about status among women, or of reductions in gender-based the use of high-quality seed and 80% (or USD 32 per month).294 Participants violence. fertilizers; tailoring extension were asked to select between a wage and an services to women’s needs; entrepreneurship track, and a major share Cash transfers have become the bedrock of social facilitating access to markets by of the gain in earnings was obtained in the protection programs in Africa and globally, and have women; and improving female latter. The project combined socio-emotional been proven to increase household consumption and farmers’ education levels. and employment skills training, and income, as well as to reduce poverty.299 Cash transfers included features to address young women’s also lead to improved employment and education constraints such as monetary incentives outcomes for women and girls and reduce the incidence for attendance, small mentorship groups, of their physical abuse. Impacts are also strong for savings accounts, and childcare. men. Cash transfers do not necessarily lead to greater impact on other measures of women’s economic empowerment, including on time allocation within the household. The impacts seem to depend on household dynamics, which are hard to change.300 130 131 1. Overview Table 6 of existing Overview of typical interventions to support female entrepreneurs programs in Africa Programs to support female Intervention Examples entrepreneurs are becoming increasingly Legal and regulatory reforms to address legal World Bank Group investment climate programs in Côte d’Ivoire and the Democratic Republic of Congo popular in Africa as governments, discrimination limiting women’s ability to conduct provided support to revise family laws and other legal texts to give all women equal opportunities to own and development partners, NGOs and businesses. inherit property, and to conduct businesses without husband’s authorization. the private sector recognize their contribution to economic development African Development Bank’s African Women in Business (AWIB) Initiative identifies business environment and job creation.302 constraints for female entrepreneurs to access credit in Tanzania, Uganda, Kenya, Ethiopia and Cameroon. Training and advice/mentoring programs to provide World Bank Women X Program in Nigeria offers growth-oriented female entrepreneurs a package that Nevertheless, few programs are designed support to female entrepreneurs in improving their includes business training (in-class and online), mentoring, access to networks and linkages to financial to address the specific constraints business management skills. institutions. faced by female-owned businesses. These programs are typically built out Goldman Sachs’ 10,000 Women Project is a corporate initiative to train growth-oriented female entrepreneurs of comparable interventions on private in 56 countries working with 100 academic, non-profit and bank partners to offer a six-month long business sector development in the region. management training program as well as networking, mentoring, and access to finance. The program in Rwanda also included a business plan competition. Moreover, only a fraction of programs supporting female entrepreneurs include Credit lines and technical assistance to financial IFC and Goldman Sachs Foundation’s Women Entrepreneurs Opportunity Fund provides lines of credit to and rigorous impact evaluations, making it institutions to enhance access to finance for female technical assistance to financial institutions in 26 developing countries to reach female-owned businesses. difficult to assess their effectiveness. entrepreneurs. World Bank’s Ethiopia Women Entrepreneurship Development Project supports female-owned enterprises in Ethiopia through a dedicated credit, coupled with technical assistance to participating financial institutions to Table 6 provides an overview of the develop tailored products for women, and entrepreneurship training to female entrepreneurs. typical interventions being implemented in various programs in Afica, including Women Entrepreneurs Finance Initiative (We-Fi) is a multi-donor facility to provide financial support and some relevant examples as an technical support to promote female entrepreneurs and unlock their access to finance and markets in illustration. The table does not aim developing countries (see Box 13 for more). Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap to be comprehensive or provide an Mentoring programs and networking platforms Cherie Blair Foundation for Women’s Mentoring Women in Business matches female entrepreneurs in endorsement of these initiatives. connecting women business owners with successful developing countries with male and female entrepreneurs for a 12-month mentoring program using an online However, the recommendations that female entrepreneurs in the region or abroad to build platform. follow in the next section can be their confidence, develop strategies to grow their compared with the type of interventions businesses, and expand their networks. Power Africa and USAID’s Women in African Power is a network to promote the participation and advance of that are actually being pursued at scale. women in Africa’s energy sector. Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa African Development Bank’s “50 million women speak” is a regional digital platform seeking to offer female entrepreneurs and networking opportunities and information on financial and non-financial services. Private sector or NGO-led interventions to Coca-Cola 5by20 initiative offers women business owners, working with Coca cola globally, skills building strengthen the integration of female-owned firms courses, financial services, mentoring and networking opportunities, working in collaboration with partners, in their supply chains. including IFC. WeConnect International identifies and trains female-owned enterprises and connects them with international buyers. Business plan competitions provide relatively large World Bank Kenya Youth Employment and Opportunities project includes a business plan competition cash grants and advice to startups, including led by targeting young entrepreneurs, and a focus on encouraging women participation. A similar World Bank women. program in Guinea-Bissau included a feature that 50% of participants in all stages of the competition (business training, business plan design support, and grants’ winners) had to be women. Business incubator and accelerator programs Standard Chartered Women in Technology Incubator Kenya Program offers training, mentorship and advice support growth-oriented firms with a comprehensive to female-owned startups. technical assistance package. USAID and Feed the Future Accelerating Women Entrepreneurs Award provides a package of technical assistance to the winners of a competition among female-owned or led agribusinesses in sub-Saharan Africa. Entrepreneurship summits and awards to give Africa Women Innovation and Entrepreneurship Forum (AWIEF), includes an annual conference bringing visibility to women as entrepreneurs, discuss together regional female entrepreneurs, an exhibition, and an awards ceremony recognizing women as common issues that affect them, and strengthen entrepreneurs’ contribution to sustainable development. their networks. Source: Authors and websites of the initiatives mentioned.303 132 133 2. What works and why: Evidence on interventions to support women-led businesses Drawing on the best available evidence from rigorous impact evaluations, this section assesses the effectiveness of – and prioritizes – those policies seeking to address constraints to women-led businesses. The policies are categorized into one of six policy areas and by the main constraint they try to address. However, these policy areas should not be considered in isolation. As explained in this report, a growing body of evidence supports the idea that the performance of female-owned businesses is affected by the interplay of several gender-specific constraints. Increasingly, programs for female entrepreneurs recognize this and combine interventions targeting multiple constraints. Table 7 summarizes the main conclusions from this review. Based on the credible evidence of positive impact, the following policy interventions are Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap recommended for scaling up to support female entrepreneurs in Africa: 1. Training programs that apply lessons from psychology to encourage women to act with an entrepreneurial mindset; 2. Supporting women with secure savings mechanisms; and, Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 3. Injecting large cash grants for female-owned businesses in the context of business plan competitions. In addition, the following set of policies and interventions, based on emerging supportive evidence, offer good potential and would benefit from further assessment of impact: 1. Removing legal constraints to gender equality; 2. Strengthening land tenure rights for women; 3. Expanding women’s linkages to new business networks; 4. Offering women-friendly training designs, including peer support; 5. Providing in-kind grants to female-owned firms; 6. Introducing financial innovations that reduce collateral requirements, including psychometric scoring; 7. Facilitating access to childcare services; 8. Engaging men to provide a more supportive environment for female entrepreneurs; and, 9. Incentivizing women to cross over to male-dominated sectors by sharing information on expected returns in those sectors, and through early exposure in the form of apprenticeships and male role models. 134 135 Table 1 LEGEND CREDIBLE EVIDENCE OF POSITIVE IMPACT ON BUSINESS OUTCOMES What works to support female-owned firms in Africa? EMERGING EVIDENCE OF IMPACT ON BUSINESS OUTCOMES EVIDENCE OF NO/LOW IMPACT ON BUSINESS OUTCOMES (NOT PROMISING) Key findings from rigorous impact evaluations POLICY AREA CONSTRAINT ADDRESSED TYPOLOGY OF FIRMS MAIN CONCLUSIONS All firms Removing legal gender biases and gaps in the implementation of laws increase women’s agency Legal discrimination and intra-household bargaining power 1. Removing regulatory and institutional Legal discrimination All firms Strengthening land rights for women increases their time and effort in entrepreneurship constraints for female entrepreneurs Easing constraints to formalization by itself is not sufficient to help female-led micro-enterprises Informality Micro-enterprises grow Providing traditional managerial training alone does not typically improve the business Skills Micro-enterprises performance of small female-owned firms Skills; confidence/risk preferences; Micro-enterprises and Training addressing socio-emotional skills and gender-specific content leads to high levels of social norms small-business owners impact on business performance 2. Improving skills and Micro-enterprises and Expanding firms’ access to new networks may, in the right settings, have positive impacts on Networks and information networks small-business owners business performance Providing mentoring on top of traditional business training has limited additional value to micro- Skills; networks and information Micro-enterprises entrepreneurs Skills; confidence/risk preferences; Micro-enterprises Complementing delivery of training programs with direct peer support may be promising social norms Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap Finance and assets Micro-enterprises Microcredit has only limited effects on business outcomes for women Finance and assets; allocation of factors Providing in-kind grants can lead to higher profits for more successful women micro- Micro-enterprises of production entrepreneurs Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 3. Improving access to Finance and assets; confidence/risk Start-up or existing Large cash grants for growth-oriented firms selected through a business plan competition can capital and assets preferences; skills businesses help overcome capital constraints for women Finance and assets; allocation of factors Providing women with access to secure mechanisms for savings – including mobile savings – can Micro-enterprises of production increase business investment Alternative credit scoring technologies using psychometric tests offer the promise of easing Finance and assets Micro-enterprises women’s access to larger business loans Time constraints/care All Providing childcare can increase female participation in the workforce 4. Easing household constraints GBV; time constraints/care; allocation Micro-enterprises Engaging men can potentially foster a more supportive environment for female entrepreneurs of factors of production 5. Addressing social Providing information on earnings in traditionally male-dominated sectors and early exposure norms regarding Skills; networks and information; Young entrepreneurs through apprenticeships and male role models can encourage female entrepreneurs to enter women’s occupational confidence/risk preference these sectors decisions 6. Facilitating access to Training does not eliminate harassment by guards at border crossings, but can make female Social norms; GBV Micro-enterprises markets traders aware of ways to minimize harassment 136 137  Policy area 1: with higher returns. An impact evaluation of In some countries, customary laws prevail in reforms to Kenya’s Succession Law removing matters of marriage, property, and inheritance, Removing regulatory, administrative, and institutional constraints for inequality in inheritance rights among male and and are explicitly exempted from constitutional female entrepreneurs female siblings similarly found the reform was principles of non-discrimination.313 Botswana is associated with greater bargaining power for one such example, but there, women’s groups married women.311 successfully challenged the constitutionality of customary rights favoring male heirs.314 However, changing formal laws will not be sufficient to address legal gender inequalities Status of evidence: Emerging. No specific for entrepreneurs, given prevailing social norms evidence of impact associated with removing and weak enforcement capacity. For example, legal gender inequalities on the business an analysis of subnational business regulations performance of female entrepreneurs, in Nigeria found that women often send a man but evidence of positive gender outcomes to deal with administrative requirements for including labor-force participation rates and businesses, even if no formal legal constraints women’s bargaining power from targeted exist.312 Customary laws are also very important. policy reforms in Africa. A good investment climate matters for all firms. Removing legal gender biases and Over the past 10 years, African governments have significantly improved the regulatory gaps in implementation of laws environment for business.304 Policymakers should increase women’s agency and consider three important gender dimensions intra-household bargaining power. when approaching investment climate reforms. First, tackling inequalities in the legal status Policymakers should seek to remove legal Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap and property rights accorded to men and women barriers restricting women’s property rights, could strengthen female entrepreneurs’ control occupational choices, and legal capacity to over assets. Second, some gender-neutral Box 10 enter contracts. Legal barriers, which are more regulations could actually widen inequalities. prevalent for married women, can restrict For example, a regulation that disproportionally women’s rights to own, purchase, manage, negatively impacts the small-business sector Family law reform in the Democratic Republic of Congo Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa or dispose of land and other property that may be especially burdensome for women, since can be used as collateral for loans, which female entrepreneurs tend to be concentrated reduces women’s bargaining power within the Before the reform, the Family Family Code was adopted by law gives women co-ownership among small firms. Third, gender disparities Code in the Democratic the parliament in June 2016. of household assets, the household.307 in the ability to influence public policy debates, Republic of Congo prohibited Now, married women can sign husband continues to hold given the varying degrees of women’s political married women from signing contracts, register land or the legal status as the head of African countries have followed a positive global representation and participation in public life, a contract, registering land or enterprises, go to court, and household with the authority to trend in eliminating legal gender disparities.308 can translate into comparatively less favorable their company, going to court, open a bank account without administer the joint property. In recent years, Togo, Côte d’Ivoire and the or opening a bank account needing their husbands’ And the law – including the policies for women business owners.305 Democratic Republic of Congo have joined the without their husbands’ permission. The minimum new Family Code amendments list of countries that have removed provisions permission. marriage age for women was – are not always properly Applying a gender lens to changes in policies denying married women head-of-household also increased. implemented. Frequently, and regulations, including their enforcement, benefits (Box 10). Few studies rigorously Reforming the Family Code women, particularly widows, can help identify how constraints to female required a coalition of The government of the are illegally denied pensions evaluate the impact of these reforms, partly entrepreneurs in the three areas could be government champions and Democratic Republic of and inheritance rights. due to the difficulty of isolating the reform’s addressed. Attention to strengthening property private sector, civil society, Congo has also signed various effect from other factors. An evaluation309 of the rights seems particularly important given and development partners. international and regional changes in Ethiopia’s family law to require both women’s challenges in access and control The ministries of Gender and agreements for the protection spouses’ consent to administer common marital Justice, with support from the of women and young girls, of assets.306 The following overview assesses property, eliminate the husband’s prerogative World Bank, proposed revisions such as the Convention on the evidence on the effectiveness of removing to deny permission to work outside the home, to the Family Code in 2013, the Elimination of All Forms regulatory, administrative, and institutional and increase the minimum age for marriage, removing most of the marital of Discrimination Against constraints for female entrepreneurs, including: found an increase in women’s non-home work authorizations. Advocacy efforts Women (CEDAW). However, 1. Removing legal gender inequalities and their continued until the revised challenges remain. While the across all age groups.310 The study supports implementation gaps; the idea that providing married women more 2. Strengthening land rights; and, bargaining power in the household allows Source: World Bank Feature Story. What does it mean to be a woman entrepreneur in the Democratic Republic of Congo? January 3. Formalization. them to make decisions to engage in activities 10, 2017. 138 139 Strengthening land rights for The scaled-up Rwanda land-titling program and Available evidence on the benefits for formalized What’s next: Expanding a pilot land-titling program in Ghana also found firms suggests limited impacts. In Malawi,329 women increases their time and increased off-farm business activity among providing only assistance in registering the our understanding of what effort in entrepreneurship. women and men.323 However, evidence of impact business had no impact for either men or women works to reduce legal and of land formalization on access to finance is on bank-account usage, savings, credit, business limited. The impact evaluation in Rwanda finds performance or other outcomes of interest. institutional barriers Secured land rights can provide women with no effects on credit or land markets, due to However, the combination of formalization greater control over assets for business ventures, Future research should seek to expand the limited development of land registries at that assistance and the bank information session while increasing their bargaining power within still-limited understanding of how changes stage.324 resulted in significantly large impacts on having the household. It can increase their financial in family, property and inheritance laws a business bank account, financial practices, capital by allowing them greater access to credit affect sector choice, investment decisions, Status of evidence: Emerging. No specific savings, and use of complementary financial using land as collateral, and providing income and access to capital by women. It should evidence of impact on female-owned products including insurance. They also had from land sales and rentals.315 Land is the main study as well how reviews of bankruptcy businesses, but positive evidence on gender positive impacts of 28% and 20% respectively productive asset for African households, but land and inheritance laws reduce risks of assets outcomes of land formalization, including on the sales and profits of female-owned tenure rights remain skewed against women, getting expropriated. It should explore how investment in land and increase in off-farm enterprises.330 who make up a minority of landholders in the different policy reforms – like registries work. region.316 of movable collateral, credit information Status of evidence: Not promising. Limited systems including on microfinance, and evidence that policies to support business A growing body of impact evaluations of land regulations underpinning micro- and small Easing constraints to formalization formalization, taken alone, have positive formalization programs in Africa finds that, when to medium enterprise (SME) finance – impact on the performance of female- accompanied with the right policy framework, is not by itself sufficient to help owned enterprises. Emerging evidence of impact women’s opportunities to get loans. these programs contribute to positive gender outcomes, including investments by women on female-led micro-enterprises grow. positive impact when combining business Even where formal laws do not formalization with complementary their land.317 However, these studies focus mostly discriminate against women, prevailing interventions such as the provision of on agricultural land with limited evidence on the For informal female-owned firms, formalization gender biases may lead to implementation business bank accounts. impact on entrepreneurship in urban settings. could in theory ease constraints in access to gaps that disadvantage female credit, networks, government contracts, or entrepreneurs331 – for example, through Past land-titling programs in Africa did not harassment from tax officials. Yet in practice, inconsistent court rulings. More work is always benefit women, and therefore their there is no evidence that spurring formalization needed to identify the impact of these design needs to include features to increase by simplifying business registration has direct gaps on women and how they can be women’s secured land tenure.318 For example, a effects on those outcomes. addressed. Understanding the effectiveness gender-sensitive land regularization program in of complementary interventions, such as advocacy, information campaigns, provision Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap Rwanda offered co-ownership to married men Global evidence suggests limited uptake of and women, resulting in a cost-effective way to easier business registration among informal of legal services to female entrepreneurs, strengthen women’s land tenure rights.319 In an firms.325 Policymakers have experimented with and nudges that enhance the application of urban land regularization program in Tanzania, more hands-on ways to encourage formalization. legal reforms supporting gender equality, price incentives were effective in encouraging A program offering cost-free assistance326 to are areas of further study. In addition, demand for co-titling.320 Similarly, adding register informal firms in Malawi substantially interventions to strengthen business Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa incentives in a program in Uganda – information increased registration for both male and female women’s participation in public-private nudges and conditional subsidies – increased entrepreneurs, although few firms chose to dialogue and other business to government the take-up of joint titling by households by register for taxes.327 More modest increases (B2G) feedback mechanisms can also be 25%-50%.321 Beyond co-titling, programs should in registration from a program in Benin that areas of future research. ensure non-married women also benefit. In combined business and tax registration suggest Benin, a participatory land-demarcation program that the perceived higher costs of formalization, Finally, new programs could shed light on and certification program helped female-headed such as paying more taxes, factor into the how gender-biased social norms influence households retain control over their land. It also decision to register. Moreover, formalization rates the implementation of legal reforms. contributed to changing perceptions toward among women were lower in the Benin study.328 Establishing causality in this context is inheritance rights for wives and daughters.322 In both cases, registration rates for men and challenging since several factors may women were higher when additional incentives contribute to changes in gender norms were offered, such as a bank-information session and the change process tends to be non- and business bank account in Malawi, and linear.332 banking services, training, and tax-mediation services in Benin. 140 141  Policy area 2: of the impact evaluations have a relatively small Training addressing socio- sample size, resulting in insufficient statistical Developing skills and networks power to uncover gender differences or any emotional skills and gender- impact overall when they might occur.339 specific content leads to high levels of impact on business performance. Furthermore, the type of training and target audience could help explain the limited impact of traditional business training. First, many The importance of socio-emotional skills – programs focus on micro-enterprises, which including self-confidence, leadership, creativity, may ex ante have limited possibilities for risk propensity, motivation, resilience, and growth, and whose owners may require other self-efficacy – on business outcomes is well- types of skills.340 Second, the training may established.347 There is emerging evidence not address other constraints hindering firm from impact evaluations of the importance growth, such as social norms shaping women’s of strengthening these skills for female entrepreneurial mindsets or the household entrepreneurs in Africa. allocation of resources.341 Third, the training may not sufficiently impact business practices An entrepreneurship program in South Africa in a way that improves firm performance. A focused on women had positive impact on recent review of these programs342 finds a strong participants’ profits and sales six months correlation between business practices and firm after the training ended, with participants also performance across seven countries, including showing improved motivation and confidence.348 The analysis presented in Part 3 of this report Providing traditional managerial Ghana, Kenya, and Nigeria.343 This suggests A study in Togo finds that personal initiative revealed evidence of gender differences in that training of sufficient quality and duration training – seeking to foster self-starting, future- a variety of business, technical, and socio- training does not typically improve could have a large enough impact on business oriented and persistent behavior – has positive emotional skills among entrepreneurs – as the business performance of small practices to improve performance,344 assuming and significant effects on the sales and profits well as gender gaps in access to training – to there are no other binding constraints, and of men- and women-led micro-enterprises, female-owned firms. that drop-out rates do not increase with longer while managerial training alone did not have an develop those skills. Moreover, differences in the size, influence, and roles of men’s and women’s training durations.345 impact (Box 11).349 Women who received personal business networks can potentially contribute to Despite the popularity of business training initiative training saw their profits increase by differences in business outcomes. programs, global evidence of their impact on Household dynamics could also contribute to 40%, compared to a 5% increase among those firm performance, and specifically on that of different business-training impact for women. who underwent a traditional business-training The following overview assesses the evidence female-owned firms, is at best mixed. Business In a lab experiment in the context of a program program. The gains stemmed from increased Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap on the effectiveness of entrepreneurship training aims to lead to improved business offering business training for micro-enterprises business investment, expanded innovation, and interventions to impart skills and expand practices, and in turn better business outcomes. in Tanzania, women were less inclined than men the proactive pursuit of financing sources. A business networks, including: (a) traditional Yet reviews of impact evaluations of business- to share income information with their spouse. previous pilot experiment with small-business managerial training; (b) development of other training programs find that while they can affect This suggests women may have less control owners in Uganda also found a similarly positive skills that may be particularly relevant for the business practices of the beneficiaries, their over their income and therefore benefit less sales impact associated with personal initiative women; (c) networking with other firms; (d) impact on firms’ growth, survival rates, and from the gains associated from greater business training.350 Similarly, a study in Uganda examining Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa mentorship directly linked to a training program; profits remains limited for women.334 knowledge.346 very recent high school graduates (not focused and (e) peer support on top of a training. on existing firms like this report) found that Some limitations make it difficult to assess the Status of evidence: Not promising. Mixed providing hard entrepreneurial skills (finance, This section also emphasizes the need to better effectiveness of business-training programs evidence as to whether traditional business- accounting and marketing) was successful at understand what types of training and additional on female entrepreneurs. First, such programs training programs have an impact for creating businesses within the target population. support may be most effective for different types vary widely in content, duration, and delivery,335 female entrepreneurs, and little evidence of However, only an alternative soft-skills course of female entrepreneurs.333 and this may partly explain why studies find impact on micro-entrepreneurs. However, on communication, persuasion, negotiation and wide variation with regard to impact. A review generalizing conclusions is difficult due leadership was able to lead both to firm creation of entrepreneurship training services in Ghana, to the wide variety of approaches and the and impacts on profits.351 Moreover, recent Kenya, and Mozambique found a variety of small sample sizes of most training impact research in Mozambique demonstrated that programs targeting different audiences and evaluations. providing a video to increase the aspirations of objectives, but these programs were not market vendors (45% of them women) led to very generally customized to their beneficiaries’ large impacts on business performance.352 needs and did not pay enough attention to the development of business acumen Furthermore, combining business training and entrepreneurial mindsets (considered with socio-emotional skills training focusing important elements for success).336 Business- on female entrepreneurs is also effective. training programs typically cover general Participants in an innovative training program business practices such as record-keeping and emphasizing self-esteem and entrepreneurship marketing.337 Less common are programs that initiative in Ethiopia – DOT ReachUp!353 – had 30% combine general business management skills higher profits than the control group. The training with sector-specific technical training.338 There also increased the motivation and confidence of is also rarely any gender-specific-content in participants.354 business-training programs. Additionally, some 142 143 An impact evaluation of a five-day training Box 11 program targeting female micro-entrepreneurs in rural markets in Kenya using the International Labour Organization’s (ILO) Gender and Togo Personal Initiative Training Entrepreneurship Together – the Get-Ahead program355 also shows significant impact. This program takes a gender perspective to building Personal Initiative (PI) Training, 11% increase in average profits that base and have stronger a type of psychological relative to the control group.359 impacts for those with lower entrepreneurship skills by complementing basic entrepreneurship training, levels of human capital. The management skills with topics such as cultural teaches a proactive, self- PI training had even more analysis of the PI training barriers facing business women and the division starting approach to of an impact for female program in Togo suggests that of household and business activities. The study entrepreneurship. It focuses entrepreneurs, who usually neither effect is first-order reveals increases in firm profits, survival, and on future-oriented, persistent see little or no improvement for women: the training is growth, and these effects were enhanced three behavior and developing “an after traditional business successful for those with all years after the training, suggesting that the entrepreneurial mindset.” training. Women who received levels of schooling, and does effect of the training consolidated over time. Participants learn to look PI training saw their profits not show significant differences These impacts were due to improved business for ways to differentiate increase by 40%, compared with a range of different themselves from other to a 5% increase for those human-capital measures. practices including marketing, record-keeping businesses, as well as to who had traditional business The results suggest that the and inventory management. Given the sustained anticipate problems and training. Additionally, because PI training can be effective increase in profits, the training, at a cost of about overcome setbacks. This of the increase in their profits, for women with a wide range $200 per participant, is credibly cost-effective. fosters better planning skills participants in the training of existing human-capital In addition, this work did not find evidence that for opportunities, and better program recouped the cost of levels.360 the intervention negatively impacted non-trained long-term preparation. the training (about $750 per competing businesses in the same neighborhood. person) within a year, thus Building on these results, The markets as a whole appear to have grown The government and the demonstrating the cost- the team collaborated with in terms of the number of customers and sales World Bank teamed up with effectiveness of the training. the Mexican Institute of psychologist Dr. Michael Frese, Entrepreneurs (INADEM) and volumes.356 to implement the PI training PI training led to more CREA (a Mexican NGO that in the context of a randomized than just a boost in profits. supports female entrepreneurs) In Ethiopia, an intensive 12-day classroom controlled trial with small Entrepreneurs who took PI to design a program that training program for female entrepreneurs, businesses in Lomé, Togo. The training introduced more merged PI training with spread over six months focusing on leadership, study compared the impact innovative products into their traditional business-literacy networking, mentoring, business development of the psychology-based PI line of business than those who training. The training is and communications, led to impacts on profits training to the IFC’s Business participated in the Business now being customized Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap on average of 144% three years after the Edge training program. Edge training. They also and implemented in other intervention, including through the development A control group received borrowed more and employed countries including Ethiopia, no training. Researchers more workers. Madagascar, Mauritania, of new business activities and business conducted four follow-up Mozambique, Jamaica, and practices.357 The participants were all women in surveys over the two years Moreover, there are reasons Nicaragua. Additional research the agribusiness sector, who had to be owners or following the training program. why different types of is underway to evaluate the managers of established businesses with at least Entrepreneurs who took the PI managerial training might be impacts of an adapted PI Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa one non-family employee. Moreover, participants training saw their profits rise more effective for those who training for the agriculture had to secure referral letters from another well- on average by 30% relative to already begin with a higher sector and to adapt and scale established business person. They were also the control group. Business existing base of human capital, this training to other sectors expected to nominate six to eight other women Edge participants, in contrast, but also reasons why they may and country contexts. who they would like to mentor after the business had a statistically insignificant compensate for those without training. The training cost roughly $2,500 per participant, but the three-year impacts more than cover the cost during that period. Finally, a program focused on adapting the Kaizen approach – continuous improvement through a process-oriented, commonsense, and cost-oriented approach to productivity improvement – to the garment industry in Tanzania (over 80% female-owned) had large impacts on profits three years after the intervention.358 Status of evidence: Credible. Positive evidence exists on the impacts of psychosocial skills and gender-specific content on business outcomes for female entrepreneurs in Africa. 144 145 In the right settings, expanding months were still benefiting from that business Providing mentoring on top of What’s next: Expanding our relationship. Moreover, the changes among the firms’ access to new networks may group engaging in active networking primarily traditional business training has understanding of what works have positive impacts on business related to access to market-specific information limited additional value to micro- with regard to improving performance. on topics such as high-demand locations, low- cost suppliers, or high-profit products. There entrepreneurs. skills and networks As discussed in “deep dive” 2, networks can were no effects on the more general business In theory, mentoring support could expand Despite the popularity of skills- support the sharing of information, equipment, skills covered in the classroom. The research business knowledge learned in classroom development programs for female workers, finance, and customer knowledge and also showed that the networking impacts were training programs. However, evidence from entrepreneurs in the region, evidence of relationships. In particular, developing “new” concentrated among business owners who mentoring programs for micro-entrepreneurs their effectiveness is mixed. Some new and looser connections can be most fruitful more frequently went to the market to purchase in Africa does not support this assumption. A programs are promising with regard to in identifying such opportunities. The gender inventory.364 program in Kenya providing individualized and addressing gender-related constraints. differences in network composition may be group sessions by mentors to complement a There are currently opportunities to further limiting women’s opportunities for growing their Similarly, a six-month program365 that asked business-training program found no additional test policy interventions and rigorously businesses. prominent businesswomen to nominate seven benefits associated with mentoring, while the assess their impact, including by: mentees, and provided the businesswomen with cost per participant of the mentoring program It is difficult to build and test a program that the tools to be effective mentors also showed no was double that of the training.368 This is • Testing different mechanisms and allows firms to benefit from the full set of more than short-term impact.366 The mentees consistent with evidence from outside Africa, sequences when combining skills and opportunities inherent in networking, including reported having on average 7.41 mentoring which has found no lasting impact of mentoring financial training components. through costs, investments and sales. In Ghana, meetings. One month after networking meetings for female entrepreneurs as compared to • Evaluating psychology-based training entrepreneurs were randomly incentivized to ended, there was a 24% increase in profits for training alone.369 under different settings, for example, in participate in a joint activity requiring daily the mentees’ main businesses, and an increase the context of conservative social norms. collaboration.361 This intervention – considered in record-keeping. About two-and-a-half years Status of evidence: Not promising. Evidence • Studying different ways to incentivize comparable with being given a business network later, there was no impact on any business of limited impact of mentorship programs for female entrepreneurs to join business – had a positive effect on capital investment and outcomes, including practices, profits, or the female entrepreneurs in Africa. networks, including male networks, firm profits, and led to the diffusion of business number of business activities performed. The and assessing such programs’ impact practices among entrepreneurs assigned to the intervention was intended to leverage social on self-confidence and the availability same team. Researchers found that women – networks to disseminate business skills, but did of information for business decision- around 60% of the sample – were less likely than not have lasting impacts.367 Complementing delivery of training making. men to have interacted with members of their assigned team, but the study did not present Outside the region, an important recent study programs with direct peer support • Comparing the impact and cost- effectiveness of in-class delivery of gender-disaggregated effects on business showed that organizing business associations for may be more promising. managerial training versus, or in Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap performance given the small sample size. the owners and managers of comparatively new combination with, personalized business Chinese firms increased sales by 8%. It also had Policy practitioners should consider designing advice and coaching. An experiment in Ethiopia, Tanzania, and Zambia, positive impacts on profits, business partners, skills-development programs in ways that • Testing different modalities of in-class in the context of a business plan competition access to finance, and management skills. The account for the barriers women face when skills-development programs, for found a positive effect on the diffusion of effects persisted one year after the conclusion growing their businesses. In-class learning instance by asking participants to bring business practices for the managers of firms of the meetings. Managers shared business- and the process of applying the lessons from Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa a friend, male mentor, or husband to the assigned to collaborate with other peers.362 Less relevant information, particularly when they the training depend on the setting in which training sessions. than 20% of participants were women and no were not competitors, showing that the meetings women operate. To encourage participation and • Preparing and evaluating different sets of gender-disaggregated results were presented. facilitated learning from peers. Managers created provide room for peer support, a program in training programs for different kinds of The study found positive information-diffusion more business partnerships in regular meetings India provided female entrepreneurs with the entrepreneurs and business managers, effects with regard to information about VAT than in one-time meetings, showing that the opportunity to come with a friend to a two-day including by size, sector, stage of registration and having a bank account. Diffusion meetings improved supplier-client matching. The training program. The promising results show development, and market orientation appears to be a combination of “diffusion of study also suggests that the topic of discussions increased household income and business (e.g., local vs. export). innovation” and simple imitation. No effects on can matter. activity, and effects were particularly strong for • Evaluating integrated support to growth- business performance were discussed in the women from socially conservative groups.370 This oriented startups through business study. Status of evidence: Emerging. Access to new suggests that peer support can help ease the incubators and accelerators. networks shown to have positive impacts constraints that social norms and confidence A program in Kenya that paired female in specific settings in the region, but no may impose on female entrepreneurs. entrepreneurs with more experienced firms in gender-disaggregated impact analysis has the community led to an average increase of 20% yet demonstrated lasting effects on business Status of evidence: Emerging. There is early in the entrepreneurs’ profits.363 All participants performance after networks are dissolved. evidence from outside of Africa on the added received a cash grant; some were enrolled in value of peer support in training programs. a business course; and some were connected to an experienced mentor. The networking with the more experienced firms had positive impacts for as long as the relationship lasted. The effect faded as the matches dissolved, suggesting the importance of a continuous networking relationship for improving business performance. Those that kept networking after 12 146 147 148 Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 149  Policy area 3: while the impact for new firms was largely Providing in-kind grants may improve the effect insignificant.381 For more experienced firms, of credit interventions. In Ghana, assistance Increasing access to finance microfinance complemented informal sources of in buying inventory or machinery led to large credit instead of replacing them, as in the case of profit impacts for female-owned firms, but only the new firms. for those that were initially more profitable.387 In-kind grants may offer a self-commitment Because the main gap between men and device because they cannot be as easily used for women is in the size of loans obtained and not household expenditures as cash grants.388 in access to credit, microcredit may become an important tool if it can help women bridge the Additionally, neither in-kind nor small cash gap in the amount of credit available to them. grants appear to be effective for low-productivity, This is not typically the case in most programs. A subsistence-type firms. Targeting more possible conclusion from the evolving debates on successful female entrepreneurs seems to pay microfinance is that, while not a panacea, it can off, at least in the case of in-kind support. be a mechanism to smooth household liquidity constraints, and when well-targeted, a possible Status of evidence: Emerging. One study source of capital for micro-entrepreneurs.382 showing that in-kind grants have a positive impact on profits for comparatively profitable Credit terms may also affect the impact of female-owned businesses run by micro- microcredit, which typically carry high interest entrepreneurs in Africa. The analysis presented in Part 2 and 3 on Microcredit has only rates and no grace period. An experiment in explaining factors underlying the gender gap in India that relaxed the immediate repayment limited effects on business firm performance pointed to gender differences obligation associated with microfinance showed Large cash grants for growth- in firm capital stock (such as inventory, outcomes for women. lasting impacts on investment rates and profit equipment, and other property) as a key levels among micro-enterprises receiving the oriented firms selected through contributor to Africa’s gender performance gap. Microcredit services are one way of alleviating grace period.383 Among those firms, the most a business plan competition can While the gender gap between male and female the capital constraints faced by women. Women business owners in Africa associated with simply represent 54% of microcredit clients in sub- risk-averse and those with high short-term help women overcome capital liquidity constraints benefited the most. Overall, taking out a loan is almost negligible, a deeper Saharan Africa.373 A first round of six microfinance this suggests that the way microcredit is offered constraints. examination of the impact-evaluation data used experiments (including in rural Ethiopia) shows can affect its impact, although identifying what in the report reveals that the gap in the amount mixed impacts on business investment, size, and features may be more positive for women micro- Providing large cash grants as part of a business of accessed credit is far more pronounced. profits. Eight of the 10-point estimates on these enterprises requires further study. plan competition can help address the capital Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap Gender differences in savings can also contribute measures are positive, with two of the positive constraints of growth-oriented firms, including to gender differences in capital investment. ones reaching statistical significance. The results Status of evidence: Not promising. Limited those owned by women. in Ethiopia case could not confirm any impact evidence that microfinance has a positive A growing number of approaches for increasing on business outcomes.374 The studies found little impact for female entrepreneurs in Africa, A business plan competition389 in Nigeria the capital available to female entrepreneurs evidence of an increase in household incomes.375 although there is emerging global evidence provided cash grants averaging $50,000 to have been tested within and outside the region that it can work under certain conditions for 1,200 existing or new entrepreneurs, who were Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa with varied effects. These typically focus on However, these studies have been criticized for certain subsets of entrepreneurs. competitively selected and received training micro-enterprises, and are often combined their lack of statistical power to detect impacts. and coaching. The program had larger positive with additional entrepreneurship support such According to another recent study,376 many effects on the likelihood of operating a firm as training. As explained earlier in this report, outcome effects are large on average, but with Providing in-kind grants instead for women than men, suggesting that women the context, endowments, and household-level high variance. Combined with low take-up rates are indeed capital constrained. It had similarly factors drive the way female entrepreneurs invest of the microcredit interventions offered, this of small cash grants can lead to large and positive effects on employment, sales, capital and labor in their businesses. Women means that the studies do not have sufficient higher profits for successful women and profit levels for men- and female-owned may have less decision-making power over the power to detect results. Pooling the data from firms. The path to positive impact seems to be use of funds, may be less willing to compete the six studies together improves this power micro-entrepreneurs. through the additional access to capital and the and take risks, or may face more household for most outcomes, with researchers377 finding hiring of labor triggered by the grant, while the responsibilities than men – with important effects of 29% on profits. However, this review As in the case of microfinance, small cash grants program had little effect on mentoring, networks, implications for their capital-investment has also been criticized for its limitations,378 may not lead to higher profits and employment or entrepreneurial attitudes.390 Moreover, the decisions. and other reviews of existing impact evaluations growth for low-income businesswomen. An business plan competition turned out to be a globally have found mixed results associated with experiment in Tanzania found that the provision useful approach to selecting entrepreneurs Research on the impact of business-development microcredit on firm creation and survival, and of small cash grants had no effect on business capable of growing their businesses beyond programs attempts to tease out the household little evidence that it contributes to sustained performance for male or female micro- microenterprise status. However, just one-fifth and enterprise effects separately, as these may increases in sales and profits.379,380 entrepreneurs.384 A study in Ghana found no of applicants in the first round were women, differ. The effects of programs addressing capital effects for female entrepreneurs, but positive suggesting that extra efforts are required to constraints for female entrepreneurs may vary if The diversity of firm types can complicate efforts impacts for men.385 Outside the region, an increase the number of women applicants.391 other members of the household (usually men) to assess the average impact of microcredit. example from Sri Lanka showed that male micro- also own businesses.371 This section assesses For example, a recent study in India found entrepreneurs invested and earned returns on A second study to assess the impact of winning the evidence from those interventions which that microcredit produces substantial and small and large grants, while women invested grants in business plan competitions was run have addressed the capital and credit constraints lasting benefits on several metrics of size and only the large grants, and earned no return, on in Ethiopia, Tanzania, and Zambia.392 These specific to female entrepreneurs.372 performance within already existing firms, average, on those investments.386 competitions provided grants of $1,000 to 150 151 emerging entrepreneurs, 22% of whom were Providing women with access to programs to target women more efficiently, Alternative credit-scoring women. The study found large impacts on the and provide them with greater control over likelihood of being an entrepreneur, as well as secure mechanisms for savings, household expenditures.400 Digital payments technologies using psychometric on business performance (sales and profits) including bank accounts and can provide recipients with greater privacy and tests may ease women’s access to and access to banking services.393 No gender mobile savings technology, can control over wages earned, which can in turn larger business loans. differences in impact were identified. provide an increased incentive to engage in increase their business investment. wage employment.401 They can also help women In environments where women are less likely Additional impact evaluations of business plan smooth unexpected income shocks by providing to own fixed assets than men such as houses competitions are underway in Africa. The results Unequal bargaining power within the access to money or support from a broader social and land that can serve as loan collateral, one will provide further evidence of the impact of this household and domestic expenditure needs network. solution may be to offer collateral-free (or approach, including how results are influenced can affect women’s ability to finance their reduced collateral) loans underpinned by solid by the size of different grants and the selection business activities. Therefore, providing female A study in Kenya finds that in areas where mobile credit-risk assessment mechanisms. Using process. entrepreneurs with mechanisms to set aside payment provider M-PESA expanded relatively financial technology (known as fintech), financial money for their business can help keep it strongly, female-headed households experienced institutions can access previously untapped A key challenge with business plan competitions separate from household demands. greater increases in consumption than did male- data that produces insights into customers and is ensuring that the selection process is open headed households. The rise in consumption markets. Data can be derived from sources and based on the merits of business plans, and For example, providing female market vendors in came hand-in-hand with an increase in savings such as psychometric tests, mobile phones, not captured by well-connected businesses. Kenya with access to savings accounts enabled by female-headed households and coincided with social media, web browsers, utility payments, To overcome this risk, the program is usually large increases in business investment (over a shift in women’s occupations from subsistence and point-of-sale trans¬action devices. Using designed with high visibility, and the scoring 45%) and consumption (37%), while no impact farming to business and retail occupations this data, it is possible to better understand a of proposals is typically done through an was found in providing such accounts to male (with 185,000 women moving from farming into borrower’s cash flows, character traits, and anonymized process and/or by independent and motorbike drivers.395 Adding access to business business occupations).402 networks to better calculate the risk associated reputable evaluators. bank accounts to support formalization led to with current customers, and to expand financial significant increases in women’s use of business Research in Tanzania assessed the six-month services’ reach to new and previously unbanked One of the most recognized strengths of the bank accounts and insurance, and also enabled effect from promoting the registration of a borrowers. business plan competitions is the selection more women to separate household and mobile savings account among women micro- of a proportionately small number of winners business money.396 This led to large impacts on entrepreneurs in Tanzania, both with and without A pilot project in Ethiopia is evaluating the after receiving a large number of applications. sales and profits for female entrepreneurs.397 business training. The study showed that women implementation of an alternative credit scoring However, the importance of that selection save substantially more through the mobile technology to enhance a financial institution’s process has yet to be formally evaluated. On the other hand, increased access to financial account than those that did not receive the ability to lend to female entrepreneurs. In Moreover, after the initial hurdles of selecting the services does not always translate into greater encouragement, and that the business training the absence of collateral, and with limited most promising applications, choosing the high- use by women. In Kenya, researchers found bolstered this effect. Women also obtained information available on the creditworthiness Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap growth entrepreneurs (and those that will see the that providing free ATM cards, which reduced more small-size loans through the mobile of women borrowers, psychometric testing is a highest impact from the grant) has proven to be withdrawal fees and increased account account. The business training further led to an promising solution. While it is widely known that difficult.394 accessibility, increased overall account use.398 increase in advanced business practices. The a borrowers’ character relates to their likelihood However, men significantly increased their study found no significant evidence that these to repay a loan, financial institutions have found Status of evidence: Credible. Studies in Africa usage of the accounts, whereas women reduced impacts translated into greater investment, it challenging to quantify these traits. Years of show that large grants made in the context account usage. sales, and profits in the period studied, but there Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa research suggest that someone’s personality of business plan competitions have a positive was increased business expansion through the can be broken down into measurable traits impact on the employment, sales and profits Research in rural settings also suggest that creation of profitable secondary businesses, as such as locus of control, fluid intelligence, of male- and female-owned firms. is difficult for poor households to start using well as improvements in women’s empowerment im¬pulsiveness, confidence, ability to delay bank accounts. A study comprising Uganda and subjective well-being.403 gratification and conscientious¬ness, and that and Malawi399 offered bank accounts to mostly these traits can be used to predict credit risk unbanked individuals. The sample included Status of evidence: Credible. Rigorous through a psychometric test. 70% women. About 46% of the households in evidence from more than one study in Africa Uganda and 37% in Malawi had a business. The shows that providing female-owned firms with Financial technologies dependent on mobile rate of account opening was 54% in Uganda and savings mechanisms has a positive impact phones or internet access are less viable in 69% in Malawi. But a much smaller percentage on business investment and performance. a market like Ethiopia, where only 16% of the actually used the accounts; within two years, 17% The interventions need to be well-targeted to population use the internet and 51 out of 100 of households had made at least five deposits ensure appropriate uptake from the relevant people have mobile phone subscriptions.404 in Uganda, and just 10% had done so in Malawi. target group. For this reason, psychometrics – lit¬erally the One important question discussed in the study “measurement of the mind” – has emerged as is whether the population was simply too poor to a promising option for creating a better picture save, or whether the bank accounts were simply of Ethiopian borrowers. Unlike other fintech not well tailored to their needs or habits. The data solutions, psychometrics could create study suggests that both factors were important. data on borrowers that did not exist before. The psychometric test has been adapted to Technology can play an important role in include more visual and interactive exercises for overcoming these risks. The use of mobile members of the population in Ethiopia with low money is spreading in Africa. Growing evidence literacy levels and limited familiarity with digital backs the benefits of digital payments for technology. women. In particular, mobile payments allow 152 153 In this pilot project, applicants can complete the What’s next: Expanding our Other policy interventions to ease capital psychometric test in just 45 minutes. Applicants constraints that could be tested in the African scoring above a certain level on the test can use understanding of what works context include: their score as a form of collateral to help assure to increase access to capital the lender that they will be able and willing to • Assessing the effectiveness of equity and credit investments on female-owned firms, pay back the loan. Preliminary results suggest the psychometric test is a reliable indication compared with credit. This could include Future research should compare the of whether an entrepreneur will repay a loan. studying mechanisms such as venture capital household effects of programs to ease Customers who scored at a high threshold on the or angel investors. capital constraints to the individual test were seven times more likely to repay their • Assessing the impacts of matching grants on enterprise effects, as well as to the loans compared to lower performing customers. business performance.411 Matching grants are intra-household dynamics that lead to used to help expand business-development capital allocation decisions (Box 12). Similarly, an evaluation in Peru using a services, upgrade technology or other Understanding differences in bargaining regression-discontinuity design and credit- equipment, or introduce innovations. power between spouses can help bureau data found that the psychometric test • Offering crowdfunding as a potential source of practitioners design effective interventions increased small and medium-size enterprise capital for SMEs. to support women-led businesses. More loan use by up to 59 percentage points for • Financing with movable collateral, such as research is necessary to understand the applicants without a credit history, without leasing or micro-leasing, factoring, trade impact of combining interventions to ease leading to worse repayment behavior.405 finance. capital constraints, such as cash grants or • Targeting female entrepreneurs with a savings mechanisms, with design features Status of evidence: Emerging. The use of business insurance product.412 Using these (e.g., increasing control and privacy) to psychometric tests as an alternative to financial instruments to overcome lack of increase the likelihood that capital will be collateral has produced promising results. investment due to risks of fires, robberies or invested in female-owned business. other incidents, as well as macroeconomic shocks impacting likelihood of failure. • Testing different approaches to incentivize savings, including through mobile money platforms and commitment devices. • Leveraging information and communication technologies (ICT) to develop innovative solutions to address credit constraints (e.g., Box 12 data-driven lending for microcredit, non- Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap collateralized loans, mobile financial services ranging from savings accounts to more Intra-household dynamics and interventions to address capital complex financial products related to farmer constraints insurance). • Combining savings and grants with different kinds of training and business advice. Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa To effectively address capital constraints for women, to hide income. But for those couples not hiding • Assessing mechanisms of reducing corruption programs need to account for how women and men income, the results go in opposite direction: A large in lending practices, as well as measure its cooperate within a household, as well as potential positive impact was seen for men who received the importance. biases in the allocation of household resources. In an loan, paired with a negative impact for women who experiment in Kenya, married couples were randomly received the loan and grant in addition to training. This assigned different levels of subsidies flowing into research points to the need to explore approaches to individual and joint savings accounts.406 When the supporting female entrepreneurs that consider the higher subsidy was assigned to a joint account, the power relations within the household.409 couples were more likely to invest in livestock and household assets. They were more likely to invest in Moreover, household and enterprise effects may their income-generating activities when the higher differ and should be estimated to fully understand the subsidies were assigned to their individual accounts. impact of a policy intervention. One possible factor This suggests that control over resources matters for explaining the limited impact of microfinance (or their allocation. grants) for female-owned enterprises is the presence of other enterprises in the household, with the capital In a related phenomenon, women may prefer to being allocated to the highest-return – often male- hide income from their spouse when they have led – enterprise. A study reviewing household and comparatively less control over how it will be used.407 enterprise effects on microfinance clients in India A recent study assesses the gender-differentiated and cash-grant recipients in Sri Lanka and Ghana outcomes of several entrepreneurship support suggests that new capital infusions (through a grant or programs in Uganda. These programs provided loan) may be invested to maximize household income either a loan, a grant, or a loan or grant in addition to rather than business activities. When the woman is business-skills training. The study used a behavioral the single entrepreneur in the household, the capital game to determine whether spouses hide their income shock leads to positive returns comparable to men’s. from each other.408 At the aggregate household and However, in multiple-enterprise households, women- individual level, no significant effects on economic led enterprises experience no or little benefit.410 outcomes were found for men and women who choose 154 155  Policy area 4: b. Engaging men can foster a Status of evidence: Emerging. No direct evidence yet exists showing that interventions Easing household constraints more supportive environment for encouraging men to take on a more equitable women as entrepreneurs. division of household labor and to provide a more supportive environment for female Engaging spouses and other male household entrepreneurship have a positive impact on members has the potential to shift the household the performance of female-owned firms, but division of labor, and to change norms and indirect evidence supports further testing. behaviors419 regarding household decision- making and the role of women as caregivers and business owners. Interventions that seek to a. Providing childcare can engage men are increasingly common, especially increase female participation in in the areas of reproductive health and gender- based violence, although rigorous evidence of the workforce. their impact remains limited.420 Women as entrepreneurs spend a Offering childcare may be an effective way to ease disproportionate share of their time on time constraints faced by female entrepreneurs. A recent program from Rwanda that engaged household work and childcare relative to men. As mentioned in Part 3, several studies, including young men and couples in group discussions The time constraints imposed by these activities in Kenya, Mozambique and Togo, have looked on gender-equitable behaviors has shown can spill over into other productive work. Indeed, at childcare from the perspective of female promising results. The program offered a the analysis in Part 3 revealed that women labor-force participation rates, and have found structured curriculum of 15 sessions. About business owners spend 3% to 17% less time significant positive impacts.413 half of the sessions offered a men-only space to per week than men working on their business discuss sensitive issues regarding sexual health – even after accounting for individual, firm, and However, there are currently no published and gender-based violence, topics that men may sector characteristics. In addition, the analysis studies that specifically assess the impact of be reluctant to talk about with their partners. What’s next: Expanding our These included open discussions among men illustrated how women’s lack of control over childcare on business performance. about their role in the household as fathers understanding of what works the allocation of household resources may be a source of inefficiency. An impact evaluation in Kenya showed that and partners, as well as on their relationship to ease household constraints mothers of children who received subsidized with their own parents. In the couple sessions, To relax the household constraints faced by early childcare were on average 1.3 percentage inter-personal communication, decision-making, This is an incipient area, but an important female entrepreneurs, some programs seek to points more likely than in the control group to be household roles and violence were discussed. one to address gender constraints that expand access to care services for children and running a business.414 Outside Africa, a study in are deeply rooted in norms about men Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap older family members. Others aim at engaging China found that a one-percentage-point increase An evaluation of this program found a significant and women’s roles and responsibilities in other household members to foster a more in access to affordable childcare increased the and meaningful impact across a range of health society. There are opportunities for further supportive environment for women business female entrepreneurship rate by 0.47 percentage and behavioral outcomes. Women reported a policy testing of different household- owners. points, with the entrepreneurship rate increasing lower incidence of physical and sexual violence level approaches to support female more in provinces with greater increases in from their intimate partners. Both women entrepreneurs, such as: access to childcare.415 and men reported a lower incidence of violent Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa punishment directed against children. Men and • Understanding if improvements in However, the high costs of childcare services women were more likely to share household childcare access produce lasting effects can diminish the potentially positive effects on decision-making on finances and fertility. on business performance. women’s economic activity.416 More broadly, Finally, men shared more of the household • Studying different mechanisms for childcare policies need to be designed jointly chores, and spent almost an hour more per day involving couples, men and other male with other policies affecting female labor doing them. While this study did not measure household members with the aim participation, such as education, and must take impacts on business performance, the results of easing constraints related to the cultural norms into consideration.417 Several highlight the potential for such interventions to distribution of household decision- other factors affect the take-up of childcare transform gender roles within households and making, caregiving and household services, including parents’ employment, the have a positive impact on a range of development responsibilities, and intimate partner availability of other caregivers, affordability, outcomes.421 violence. location and opening hours, and trust in the • Understanding if programs that involve service provider.418 Careful policy design and Joint couples training can also be used to engaging men can complement governance of childcare provision is also critical promote greater equality in the allocation of interventions aimed at promoting for a favorable impact on children’s development household resources. A pilot project in Ghana proactive behavior. and long-term prospects. is testing the relative effectiveness of providing • Comparing approaches designed to small cash grants that are conditional on both provide access to savings and capital, Status of evidence: Emerging. No studies the beneficiary and her partner attending a focusing on household collaboration (e.g., from Africa demonstrate the impact of training program on the allocation of resources training of couples), with those focusing care services on business performance. within the household.422 The training program on female entrepreneurs’ control and Global evidence, including from Africa, show uses interactive exercises focusing on how privacy. increases in women’s labor-force participation women-run businesses can be as profitable as • Testing the importance of more holistic rates associated with greater provision of those run by men, the importance of supporting in-household approaches to overcoming childcare services. these businesses, and how this can benefit the gender norms and improving business entire household. opportunities for women. 156 157  Policy area 5: to opt for male-dominated fields. Most women – including the provision of safe employment Addressing social norms regarding women’s occupational choices who enrolled in these male-dominated trades environments for women – should be identified.433 eventually dropped out from the program, which suggests that information alone may not be Status of evidence: Emerging. Limited sufficient.429 evidence exists as to whether interventions designed to encourage women to cross However, in Republic of Congo, a vocational over to non-traditional sectors have positive training program on different trades (male and impact; indirect evidence supports further female-dominated) is providing information to testing. applicants on returns. The program produced two videos: one simply describing the different trades and available training programs (version A), and one including the same descriptions but also information on median earnings in each trade (version B). Applicants are randomly invited to watch either version A or B before submitting As presented in Part 3, an entrepreneur’s sector Providing information on earnings an application for a specific trade. This will of operations partly explains gender differences in business performance. Recent studies in the in traditionally male-dominated help clarify whether having better and more accessible information on potential earnings region found that female entrepreneurs who sectors, along with early exposure encourages women to opt for typically male- operate in male-dominated sectors perform through apprenticeships and male dominated trades. much better than those in traditionally female- dominated sectors, and as well as men in these role models, can encourage women Programs to encourage crossovers should higher-return activities.423 to enter these sectors. consider exposing women to the sector through apprenticeships and by including male mentors, An initial step to increase participation of women A set of policies are being tested across thus providing an encouraging environment for in traditionally non-female sectors would Africa aimed at encouraging more women women to enter non-traditional sectors. The be to remove legal restrictions to women’s “crossovers” to enter the higher-value sectors studies in Uganda and Ethiopia found that early occupational choices. In Africa, 62% of countries traditionally dominated by men, including exposure to a male role model was a key factor impose restrictions on women’s access to providing information on earnings opportunities, in women’s decision to enter a male-dominated certain professions.424 However, legal reforms by mentorship and apprenticeship programs, and sector. Ugandan women with a male role model Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap themselves are not sufficient to encourage more training sessions focusing on sector-specific are 12% to 22% more likely to be a crossover.430 In women to join sectors previously reserved for business and technical skills. Ethiopia, crossover firms were more likely to have What’s next: Expanding men. started their business based on an opportunity our understanding of what The analysis thus far suggests that the One area of intervention includes providing provided by their husband. In addition, qualitative works in addressing social information on the earnings opportunities in research has found that a father’s occupation can differences in participation in higher-return male-concentrated sectors. Accurate information influence the decision to enter a male-dominated norms regarding women’s Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa sectors are not driven by endowments such as can address misperceptions of earnings in sector by providing access to capital or start-up occupational choices education or access to finance, but instead by traditionally female sectors and help women, funds.431 psychosocial factors – particularly the influence especially young women, make more informed There is substantial scope to test and of male role models and exposure to the sector decisions when choosing their sector of activity. Emerging evidence suggests that the provision evaluate new approaches to helping women by family and friends.425 Time preferences shaped Female entrepreneurs in traditionally female- of technical training, even without explicitly switch into more profitable sectors in by social norms may also contribute to sector dominated sectors incorrectly believe they make targeting occupational choices, can shift different settings. Potential ways to expose selection.426 the same or more than their counterparts in women’s internalized norms regarding their women to these sectors include: male-dominated sectors. In Uganda, 80% of appropriate sectors of employment. A recent non-crossovers who make less than crossovers study in Nigeria found that graduates of an ICT • Providing encouragement from mentors, think they make the same or more as the women training program were 23% more likely than family and friends to nudge women to in male-dominated sectors.427 In Ethiopia, women a comparison group to be employed in that make the switch. in traditionally female sectors incorrectly believe sector two years later, and that among those • Leveraging apprenticeships and similar that their profits are the same or higher than women initially more biased against professional on-the-job learning opportunities to their female counterparts in crossover sectors, attributes for women, the likelihood of switching familiarize women with a sector, while when in fact they are lower for 64% of firms.428 to the ICT sector was three times higher than building their skills to operate in it. for non-biased women.432 However, absent more • Connecting women with male-dominated In a vocational training program in Kenya, women deliberate policies, providing technical training networks and opportunities (often as exposed to information on expected earnings is unlikely to have a substantial impact on part of local content development), even and a video of female mechanics were almost occupational choice for women. for the supply of products and services nine percentage points more likely to express in sectors in which women are already a preference for taking vocational training in Potential risks related to an increase in present. a male-dominated trade, and five percentage harassment and gender-based violence should • Exploring mechanisms to help women points more likely to enroll in one. Younger be assessed when encouraging women to enter deal with social backlash resulting from and more educated women were more likely non-traditional sectors, and mitigation measures their entry into male-dominated sectors. 158 159  Policy area 6: Facilitating access to markets for female entrepreneurs Across Africa, policymakers are implementing Training does not eliminate Status of evidence: Not promising. There programs to connect businesses with domestic, is no evidence, based on a first study, that regional and global markets. These programs harassment at border crossings, training of border guards and small traders need to pay attention to gender-specific but can make female traders aware by itself reduces the level of harassment by constraints to make sure female-owned firms of ways to minimize harassment. border guards. However, it is too early to benefit from them. To date, they have not typically rule out training as an effective policy option, been oriented toward the constraints outlined in Small informal traders, mostly women, handle particularly if it were to be coupled with Part 3 of this report. a significant volume of trade in Africa, yet fall improvements in the institutional environment outside official trade statistics. A survey of small (e.g., making sure border officials are paid on There is limited, albeit growing, recognition of the women traders in the DRC-Rwanda border found time, providing mechanisms to address legal role that African women play in trade, as small that 80% of respondents have been asked for a complaints). informal traders, producers of tradable goods bribe, and more than half have been subject to Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap or service providers, or leaders of exporting harassment. Lack of information among women firms.434 The specific challenges that each of traders about their rights and responsibilities these groups face when accessing markets emerged as a key constraint. need to be understood and addressed. For example, informal women traders often lack An impact evaluation conducted along the knowledge about border-crossing procedures DRC-Rwanda border measured the effects of Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa and are subjected to requests for bribes and a training program for (mostly women) traders harassment.435 There is little rigorous evidence of regarding border procedures, customs tariffs, gender-related outcomes associated with policy and their rights and personal safety during interventions promoting access to domestic or crossings. Border officials also received training foreign markets, whether in Africa or globally. on procedures and tariffs, governance issues and This is an area in which further study is acutely gender issues. Traders who received the training needed. responded strategically, by crossing the border earlier than the control group – 23 minutes earlier on average – thus minimizing encounters with guards and other border controls (official or unofficial). Rates of harassment dropped by 29% among participants, and the share of traders reporting paying nothing at the border increased by 5%, despite the lack of impact on the total official and unofficial charges paid. There was no impact on overall profits.436 The results suggest that in a context of ambiguous rules, the traders provided with information opted to minimize risk by adjusting their behavior. Therefore, changing the behavior of border guards may require different or additional interventions. 160 161 What’s next: Expanding our profit increases between 15% and 25% by reduce transaction costs and increase SME understanding of what works securing export orders. The study finds strong participation, including by female-owned evidence of “learning by exporting” among firms. Some countries have introduced more in addressing women’s access treatment firms.439 proactive policies to include female-owned to markets firms (or other disadvantaged groups) such • Improving networking capital for female as: (i) establishing mandatory goals or A range of policy interventions are being entrepreneurs through online platforms. targets, (ii) requiring firms awarded contracts tried to facilitate access to markets, for Increasing opportunities for female-owned above a certain threshold to include female- male- and female-owned businesses firms to increase their networking capital440 owned firms as subcontractors; (iii) providing in Africa and elsewhere. Unfortunately, can in theory help them access new business preferences for female-owned firms (e.g., there is limited to no gender- opportunities. Policy interventions seeking a price advantage); and (iv) setting aside differentiated evidence that might enable to support women exporters should not rely contracts for female-owned firms.446, 447 recommendations about what works. on traditional male-dominated networks, Rigorous evaluations would help clarify the Below are potential approaches that could as women may not be connected to these potential impact of these interventions. expand female entrepreneurs’ market- networks and thus may miss out on program participation opportunities, but which benefits.441 require further study. Using internet and mobile phone platforms • Reducing the degree to which small- offers an opportunity for SMEs, including those scale female cross-border traders are owned by women, to exchange information hassled. Some proposed interventions to and potentially gain new clients.442 The rapid test based on quantitative and qualitative expansion of broadband access, expected to research with small traders include:437 reach 80% of the African population by 2020,443 ›› Creating a charter listing the rights along with the fast growth in smart-phone use, and obligations of traders, and paves the way for the development of such displaying it at a visible place at platforms. There has not, however, been a customs offices. rigorous impact evaluation of these platforms ›› Increasing safety measures at the to date. border including cameras, help lines, etc. • Facilitating female-owned firms’ access to ›› Improving access to information. government procurement contracts. ›› Introducing a simple trade regime Access to government procurement represents Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap (e.g., the simplified trade regime an important potential market for SMEs. of the Common Market for Eastern Governments frequently use the public and Southern Africa (COMESA) for procurement framework as a way to achieve shipments under $1,000). socioeconomic objectives, including the promotion of female entrepreneurs. Yet these Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa • Export promotion programs targeting policies have costs due to limits on competition female-owned firms. These programs, and resources needed to minimize fraud. To provided by trade and investment determine the impact and cost-effectiveness support institutions, NGOs, or of these policies, rigorous policy evaluations business associations, can include are needed. A study in Brazil shows that information on market requirements winning at least one government contract and regulations, assistance in meeting in a given quarter increases firm growth export requirements, and networking by 2.2 percentage points over that quarter. assistance and links to foreign buyers. No gender-disaggregated results were Impact evaluations from the Middle East presented.444 and North African region have identified positive, though not always persistent Governments can apply several specific effects associated with these policy measures to help reduce challenges for interventions. No gender-differentiated female-owned firms in accessing public results have been reported. A study of contracts, similar to those aiming to facilitate a public export promotion program in the access of smaller firms to public Tunisia finds the program led to faster procurement processes. These initiatives export growth and a diversification of include streamlining tender documents products. Yet the program’s impacts and procurement processes, pre-qualifying dissipated after three years, suggesting female-owned firms, avoiding contract that firms may need additional support bundling, allowing time for tender preparation, to stay competitive in export markets.438 providing feedback to potential suppliers, Another study in Egypt finds that rug and enforcing rules requiring the prompt producers offered the opportunity to payment of suppliers.445 Another potential export to the United States experienced intervention is eProcurement, which can help 162 163 3. Designing effective programs to support female entrepreneurs Table 8 Checklist for assessing potential country- or community-level gender-specific constraints Understanding the landscape of female-owned firms and the constraints they face is the first step in the design of any business program or intervention (Table 8). Moreover, effective program design calls for context- specific solutions, as environments for female entrepreneurs and the Business typology and †† number and size of female-owned businesses types of businesses they run vary within and across countries, sectors, and performance †† sector disaggregation socioeconomic backgrounds.448 As more and more countries embark upon †† formal vs. informal status of female-owned businesses solutions aimed at overcoming the gender gaps in business performance †† wage-employment opportunities available for women (see, for example, Box 13), setting the bar high in terms of program design is critical for achieving success. Business operating †† legal discrimination Relying on a combination of quantitative and qualitative methods can help context †† political voice provide a full picture of female entrepreneurs and the constraints they face. †† sexual and gender-based violence Useful sources of data include enterprise surveys, household surveys, and †† participation in social and business networks †† interactions with individuals of the opposite sex surveys from impact evaluations. Reports like the World Bank’s Women, †† perceptions about women’s role in household/caregiving activities Business, and the Law, or country-specific investment climate or value- †† perceptions about sector choice chain studies can provide further insights into legal constraints and other †† mobility constraints issues affecting female-owned firms. Gender-disaggregated diagnostics including those that can be integrated in IFC’s Country Private Sector Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap Diagnostic (CPSD) can be important instruments to design effective sector- Education and skills †† literacy levels oriented programs. †† education attainment at secondary and tertiary levels †† access and use of ICTs Conducting in-depth qualitative research can shed light on important †† availability of business and technical skills training programs, and questions at design. Consultations with women business associations and women’s participation in such programs individual business owners can also help identify underlying constraints Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa and hone program design. When gathering information on sensitive topics, such as domestic violence or psychological well-being, a mixed-methods Finance and assets †† ownership of land and property approach can be most effective.449 †† ownership of other tangible assets †† access to savings mechanisms †† access to finance Confidence/Risk-taking †† willingness to compete †† risk appetite Household-level †† control of household assets constraints †† patterns of household collaboration †† time use preferences †† availability and affordability of childcare Source: Adapted from World Bank – WLSME (2013). 164 165 Box 13  Getting targeting right Women Entrepreneurs Finance Initiative (We-Fi) In Africa, many entrepreneurs – both women and men – are driven to start their own businesses by limited opportunities for wage employment. Thus, The Women Entrepreneurs to encourage and facilitate and mentoring programs. many tend to be subsistence-oriented and comparatively uninterested in Finance Initiative (We-Fi) is a greater use of and engagement Finally, We-Fi supports growing their businesses.450 Growth-oriented entrepreneurs are most likely program launched in 2018 with with female-owned SMEs; piloting innovative ideas in to benefit from enterprise development programs but identifying such an initial funding allocation collect high-quality financial areas related to promoting individuals remains a challenge. Evidence from recent impact evaluations of US$ 340 million to support data on female-owned the expansion of the number programs and activities that businesses and impact of women as entrepreneurs offer some useful lessons on targeting. eliminate barriers that women evaluations; and leverage donor in male-dominated sectors, face in starting and growing funds, ultimately channeling circumventing norms regarding The effectiveness of enterprise development programs for women – such successful SMEs in a variety these directly or indirectly to willingness to compete, as those providing skills, training or credit – can vary depending on the of sectors, and strengthen the women-led SMEs. increasing the take-up of characteristics of the beneficiary firms, including size and the education enabling environment for such financial technology, and or experience levels of managers (Box 14), as well as the surrounding firms. In the public-sector window, funding rigorous impact environment and infrastructure available to them. In India, an impact We-Fi seeks to help evaluations. evaluation of microcredit lending identified growth entrepreneurs as those We-Fi supports complementary governments improve the who had started their businesses prior to the availability of the microcredit, approaches through two business environment for The initial allocation of funding windows: public and private female entrepreneurs by in Africa includes programs and found that these firms were most likely to benefit from the credit.451 sector windows. The We-Fi identifying and addressing led by the Islamic Development Therefore, programs seeking to support high-growth entrepreneurs’ seeks to provide dedicated binding regulatory and legal Bank in Nigeria and Mali, skills development or access to credit may achieve the most benefits by resources to foster innovation constraints, creating market and those implemented by considering targeting larger and more experienced businesses. However, and new approaches to opportunities and developing the World Bank Group in training programs must clearly demonstrate their value if they want to removing the constraints faced innovative programs to support Côte D’Ivoire, Mozambique, entice the busy owners of such firms to participate. by women as entrepreneurs, the growth of female-led Nigeria, Senegal, Tanzania, while also helping elevate the businesses that align with and Zambia. The World Bank Reports from community members can be a reliable way of identifying issues so as to drive action by country commitments to program will also include the men and female entrepreneurs most suited for participation in such governments and the private gender equality. The We- global programs on collecting sector. Fi supports proposals that gender-disaggregated data programs, as demonstrated by a project distributing cash grants in can improve the business on formal enterprises, impact India. Community members are typically good at assessing the micro- In the private-sector window, and legal environment evaluations, and tourism- entrepreneur’s abilities and potential return on capital, although the We-Fi is looking to support for women businesses in sector opportunities and reports lost accuracy when participants were aware they could influence Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap financial services for women such areas as collateral analysis. the allocation of cash grants and present friends and family members in customers and women- registry, land registration, a favorable light.452 Some ways to mitigate misreporting include small led SMEs through financial business registration and monetary incentives, public vs. private reporting, and cross-checking intermediaries and non- complementary targeted reports to identify those who might give favorable reports to each other. bank financial institutions; interventions (e.g., help with implement performance-based setting up business bank incentives and risk sharing for accounts); women’s voices in Business plan competitions are becoming popular in the region following Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa banks and non-bank financial- public-private dialogues; and the success of the Nigeria You Win! Program. These have proven a useful service providers, including in access to financial services. way to identify high-growth micro-enterprises.453 Applicants respond the form of blended finance; We-Fi also supports technical to a widely publicized call for business ideas and then participate in a support angel, seed, venture assistance and grants to help competitive process. The programs typically provide a small amount of and growth capital-type open markets to female-owned training to applicants to help them prepare a business proposal, which is facilities; provide financial businesses, through advisory then scored by business experts. Expert panels are a common mechanism tools, incentives, and technical projects to increase women’s used to select the winning firms. A recent paper compares the predictive assistance to increase engagement in export and power of expert opinions with entrepreneurs’ surveys to gather applicant corporate sourcing from and trading activities, value-chain to female entrepreneurs and analysis and market linkages, information in the context of a business plan competition in Ghana.454 It women-led SMEs; provide online platforms, supplier finds that ability – measured as a combination of education, cognitive skills, advisory services and technical development programs, and financial literacy – and to a lesser extent, managerial practices, were assistance to financial business incubators, innovation good predictors of enterprise growth among participants. Experts’ opinions intermediaries, business centers, and capacity- added predictive value on top of the surveys. However, identifying the networks, funds, corporates, building through training fastest potential growers remains tricky. In the business plan competition and accelerators/incubators and technical assistance in Nigeria, among those participants reaching the semi-final stage, neither the business-plan scores (provided by experts), nor the entrepreneur or firm characteristics predicted those firms most likely to grow most quickly or benefit most from the program.455 Source: Adapted from we-fi.org and supporting documents. 166 167 Box 14 • Lowering “the stakes” of the competition: Women may set a higher bar for themselves, compared to men, when the stakes, such as responsibility for someone else’s earnings, are higher.463 One possible response is Targeting may matter in skills development having competitions where the stakes, such as the size of the prize, are not as salient in the awareness campaign. • Feedback on their ability to compete: Women may tend to underestimate Effective targeting identifies level. A group of firms were In South Africa, an impact their relative ability and, in turn, their chances of winning. This perception and attracts those firm owners offered basic business training, evaluation finds that training may lead women to stay out of the competition. Therefore, evaluating and who will benefit most from while another group received on marketing and finance providing feedback on their relative ability could address misperceptions and most effectively apply the training plus individual among men and women the content of the training.456 specific consulting sessions small-business owners leads and encourage more women to compete. Comparatively experienced and coaching. In line with to significant improvements in • Support from peers: This report described that successful female business owners often other studies, the basic profits one year after training entrepreneurs in Africa typically receive strong support from their closed already have the technical business training did not delivery. The higher profits networks, especially their families. Building support from peers, such as and entrepreneurial skills result in improved business were produced by improved allowing participants to bring a relative or friend to the program activities, necessary to grow their performance. The enhanced business practices in line with may increase women’s willingness to enter a competition. business.457 However, more support was found to improve the type of training received. • Incentive from role models: With the goal of encouraging women to experienced business owners business practices, but only More established formal see themselves as being able to replicate the success of competitive are also more likely to decline the comparatively experienced businesses were found to entrepreneurs, programs can bring in local or regional role models to the participation in training participants improved their benefit more from the finance programs, perhaps due to the earnings. The impact was training, while those with less interventions or awareness campaigns. higher opportunity costs of larger with every additional previous exposure to different training.458 Designers of such year of experience. This finding business contexts and markets programs should thus consider suggests that experience may reaped higher profits from the how the training content and increase the ability to translate marketing training. Influenced selection mechanisms will the training program’s lessons by the choices taken, the attract these firms. into improved business impacts were larger for the practices, or that more more motivated, formal, larger, Considering time Female entrepreneurs may face several barriers when taking up training, A study of a program targeting experienced entrepreneurs may and experienced businesses.460 and mobility due to their caregiving obligations, social norms or safety concerns female entrepreneurs in not be constrained by other constraints when restricting their movement, lack of resources, or lower literacy levels. A Tanzania highlights how barriers hindering business designing training study in Kenya finds that observable entrepreneur characteristics predicted firm impacts can vary by the growth.459 differences in attendance rates in a business-training program for female programs entrepreneur’s experience small-business owners.464 Women over 35 years old and unmarried women were more likely to attend, suggesting that household responsibilities may constrain attendance rates. Previous participation in a training program Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap and living in a large household or close to the training location were also associated with higher attendance rates. On the other hand, owners of  Building gender-sensitive aspects in program firms with relatively high profits were less likely to attend, possibly due to design time demands associated with the business, or to lower perceived benefits from the training. Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Gender-specific constraints, like those discussed earlier in this report, may prevent female entrepreneurs from participating in business plan The schedules, durations, and locations of training programs should competitions, training, and other enterprise development programs. therefore be tailored to the specific needs of targeted groups of female Women may decide not to participate in a business-plan competition if they entrepreneurs. Approaches to be considered and evaluated depending on are competing against men. Or they may drop out of a training program if the context include: the schedule interferes with their household obligations or requires them to travel to a distant location. Understanding these constraints and identifying • Reduced schedules (e.g., part-time instead of full-time sessions), so ways to address them during program design will increase the likelihood women can continue to tend to their businesses during the training. of program effectiveness. Below is a discussion of the approaches being • A non-sequential curriculum that allows trainees to complete a subset of tested to address some of these constraints. More rigorous evidence is sessions when their schedule permits. needed to know what solutions work best. • Conducting the training in a nearby location and/or offering transportation. • Providing childcare during training. • Offering food (including food to take home). Addressing Even after participating in a training program, women may still be more • Adjusting training content to participants’ literacy levels. women’s lower reluctant than men to enter a competition.461 By acknowledging this • Being sensitive to cultural norms (e.g., female instead of male coaches). willingness to gap, practitioners could explore different ways of encouraging women to compete compete, such as: Finally, leveraging technologies such as mobile phones for training delivery and business advice can potentially help women overcome some time • All-female competitions: Women may be more likely to enter a and mobility constraints, although more research is needed to assess the contest if they are only competing against other women. The Kenya effectiveness of such mechanisms.465 lab experiment described in Box 6 found that women were more likely to enter a competition when only facing other women.462 Business plan competitions, for example, could reserve one round for women participants. Women may also prefer participating in women-only training sessions or mentoring groups. 168 169 4. Expanding our understanding of what works Evidence on what drives the existing performance To design successful policy interventions, it is gap between men and female-owned businesses important to identify all the binding constraints in Africa – and on how to support female that businesswomen face, and design solutions entrepreneurs – is growing, yet remains to address them jointly. This requires impact limited. Closing the performance gap requires evaluations that “horse-race” different a combination of clear policy actions at scale, interventions to address similar constraints – better targeting of beneficiaries, and rigorous to gauge the effectiveness of different delivery research and impact evaluations to expand our methods – as well as evaluations combining knowledge of what works. This section outlines complementary interventions that address joint opportunities to move this important agenda binding constraints in cases where research forward, including suggestions on the roles that indicates they exist. Moreover, studies assessing governments and other stakeholders can play. the impact of different policy options targeting the constraints of female entrepreneurs need Developing effective policies to help African to consider the specific channels through which female-owned businesses thrive requires a these constraints affect business profits. For continued push to expand the understanding instance, it would be useful to know whether of existing gender gaps and the underlying the priority of business-training programs constraints behind them. Researchers should should be to provide technical knowledge on use an array of methods, including qualitative specific business practices or to build women’s and inferential research, to learn about the confidence in launching new ventures. constraints different groups of women face while operating their businesses across Africa. Innovative policies could be tested to address the underlying constraints female-owned enterprises Conducting this research ex ante and in face in Africa. Table 9 lists potential policy combination with the preparation of impact options to overcome specific constraints, and evaluations will lead to the design of better thus contribute toward learning about whether Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap solutions to be tested. For example, this research these constraints are binding. is critical to understand which constraints need to be jointly targeted for an intervention to result Within the long list of ideas presented in Table 9, in higher profits for female entrepreneurs. one promising strategy is to first learn about the Therefore, it is possible to gain additional insights effectiveness of the interventions that have some about the strength of different constraints in emerging evidence, including securing land Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa the context of testing different intervention tenure rights, alternatives to collateral, exposure approaches. to male-dominated sectors, and childcare services. When a policy intervention succeeds at promoting the growth of female-owned To increase the external validity of these businesses, it confirms that the intervention studies, it would be important to replicate effectively addresses a binding constraint for promising policy interventions in different these firms. However, the well-estimated contexts. Additionally, longer follow-up periods absence of impact from a policy intervention on will help confirm if results hold over time. the performance of female-owned businesses Finally, research design should consider how can be due to one of three reasons: statistical power will be achieved as low take-up rates and/or high attrition rates are common 1. The constraint targeted is not binding, so challenges in many programs supporting female addressing it has no impact on business entrepreneurs. outcomes; 2. The constraint is binding but the intervention tested is not an effective way of addressing it due to design or implementation issues; or 3. The constraint targeted is important, the intervention tested is an effective way of addressing this constraint, but other constraints bind at the same time. In the latter case, an intervention relaxing only one constraint will not affect business outcomes. 170 171 Table 9 Potential policy interventions to study effectiveness for female-owned businesses in Africa 1. Regulations & 2. Access to skills 3. Access to capital 4. Household 5. Sector 6. Access to Design institutions and networks segregation markets Legal barriers to Entrepreneurship Alternatives to Different models Information to Trade facilitation Mechanisms property rights, control training focused collateral (fintech for delivering enter male- and related to identify of assets/household on building self- innovations) childcare dominated interventions for high-growth decision-making, confidence services sectors traders entrepreneurs business operations Leasing Classroom delivery Engaging men Exposure to Export Mechanisms to Securing land tenure compared to, or in Equity vs finance vs to create more male-dominated promotion address issues rights addition to coaching grants supportive sectors through programs of women’s and/or business advice environment internships and confidence and Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap Matching grant training Outsourcing willingness to Business to Business incubators/ programs Increasing models to compete government feedback accelerators for couples’ Support from procure services loops / public-private women Factoring / trade collaboration male role Mechanisms to Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa dialogue (PPD) finance vs. approaches models Online deliver training Worker skills to strengthen marketplaces Crowdfunding women’s control Women in and privacy leadership roles Government Angel investment on income/ in firms procurement spending programs Digital savings decisions for SMEs (preferential Mental accounting margins, payment Helping firms periods) graduate (e.g., from microfinance Linkages institutions to banks) programs connecting SMEs with large firms/SEZs 172 173 Steps for African policymakers Steps for development partners, corporations, and civil society 1 2 3 1 2 Support concrete Ensure private- Scale up policies that policy actions that sector development have shown credible demonstrate strong strategies and policies results. Provide funding to test innovative Use findings from this report to commitment to female include a gender focus approaches and research that inform the design of programs. entrepreneurs like that addressing the contribute to closing the gender eliminating existing specific challenges entrepreneurship gap while legal barriers and faced by female expanding global knowledge of fostering women’s entrepreneurs. what works. participation in public life, including promoting female role models. 3 4 4 5 6 Continue advocacy efforts by For larger firms, consider the Support the testing Invest in promoting Involve men in policy highlighting the business and gender dimensions in their supply Part 5 | Path forward: How policy and decision makers can act to eliminate the gender gap and evaluation of the systematic advocacy and in the economic case for removing the strategies and opportunities to promising approaches collection of gender- implementation of obstacles constraining female integrate female-owned firms. – and share the differentiated data solutions – given that entrepreneurs’ growth. findings widely. that captures the men are husbands, but Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa performance, are also more likely endowments, and to also be bankers, preferences of female- inspectors, trainers, owned firms. and policymakers, it is important to engage men at multiple levels in efforts to provide better opportunities for female entrepreneurs. 174 175 Appendix 1 The gender gap “D” is expressed as the mean outcome difference: Technical appendix on the (2) D = E (YMF ) – E (YFF ). decomposition method Replacing Equation (1) into Equation (2) and taking the expectations, the gap is: (3) D = E (X’)ßMF + E (εMF ) – | X E (X’ )ßFF – E (εFF ) = E (X’ | X )ßMF – E (X’ )ßFF’ and Equation (3) could be rewritten as: (4) D = ß 0 ,MF + ∑ k= K 1 E ( X k,MF ) ß k,MF – ß 0 ,FF – ∑ k= K 1 E ( X k,FF ) ß k,FF where ß 0 ,g is the intercept of each gender model. Subsequently, the analysis can estimate Equation (1) using a pooled sample of owners controlling for owner gender, i.e. including a dummy variable This appendix provides a detailed technical overview of the decomposition that identifies female-male owners. ß * is the vector of coefficients from this method used in this report. This technical appendix is based on O’Sullivan regression. The inclusion of the gender dummy is meant to avoid a possible et al. (2014) and Fortin, Lemieux, and Firpo (2011). distortion of the decomposition results due to the residual group difference reflected in coefficients ß. Hence, rearranging Equation (4) and adding and Decomposition methods have been widely used in economics to analyze the subtracting the intercept coefficient of the pooled regression ß 0* and the contribution of different factors to increase or decrease an outcome gap terms E ( X g,k ) ß k * , one obtains: (e.g., productivity gap, gender wage gap, union wage gap, etc.). The report includes the Oaxaca-Blinder (OB) regression-based mean decomposition. The OB decomposition is the most extensive method employed in applied economics over the past three decades. (5) D = ∑ k= K 1 [ E ( X k,MF ) –E ( X k,FF )] ß k * + Component 1: Endowment Effect The method seeks to decompose differences in mean outcomes across two groups. Despite being very simple, this method requires a strong set of assumptions. First, it follows a partial equilibrium approach, K K where observed outcomes of one group are used to construct various ( ß 0 ,MF – ß )+ ∑ E ( X k,MF ) ( ß k,MF – ß ) + ( ß – ß 0 ,FF ) + * 0 * k * 0 ∑ E(X k,FF ) (ßk * – ß k,FF * ) k= 1 k= 1 counterfactual scenarios of the other group. Second, estimations used as inputs in OB decomposition are based on correlations, and hence Male Structural Advantage Female Structural Advantage cannot be interpreted as estimates of underlying causal parameters, as Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Component 2:Structure Effect noted by Fortin, Lemieux, and Firpo (2011). Hence, expressions used in this report such as “the main factors that contribute to the gender gap in productivity are” should, therefore, be viewed in this light. Nevertheless, the decomposition methods allow an examination of the relative quantitative where ß 0 ,FF , ß 0 ,MF , ß 0 * , ß k,FF , ß k,MF , ß k * (k=1,…,K) are the intercept and slope importance of factors in explaining an observed outcome gap, suggesting coefficients of each covariate included in the regressions of male, female which factors drive the gender gap for further analysis and policy and pooled owners’ samples. interventions. Third, the analysis assumes linear effects, leaving aside non- linear effects. Equation (5) shows the aggregate decomposition. The first component is the endowment effect, or the explained part of the gender gap by differences This work explores productivity gaps across firms with a female owner (FF) in the levels of variables between both groups. The second component and firms with a male owner (MF). The background of this report relies on is the structure effect, or the unexplained part of the gender gap that is an OB regression-based mean decomposition. It considers as a productivity driven by deviation of each group’s return from the corresponding “average” outcome measure (Y) the log of profits, sales, value-added, investment, return. The first term of the structure effect represents the male structure labor hours, and other firms’ relevant outcomes. Once a productivity advantage, which is the portion of the gender gap accounted by deviations measure is picked, the following equation is estimated: of male returns from average returns. The second term of the structure effect represents the female structure advantage, which is the portion of (1) Yg = X’ßg + εg’ the gender gap driven by deviations of female returns from average returns. Appendix & Endnotes where g indicates the gender of the owner, X’ is a matrix (nxK+1) with K In practice, the OB decomposition is very simple to estimate. For example, observable owner and firm characteristics; ß is the associated vector of the endowment effect is the sum of all differences between male and intercept and slope coefficients; and εg the error term under the assumption female owners’ covariates means, valued at the corresponding average that E (εFF ) = E (εMF ) = 0. return. Thus, the analysis needs to compute sample means and estimate 176 177 the average return from the pooled model. In the case of the structural effect, besides means calculation, one needs to add the Ordinary Least Squares (OLS) estimated coefficients (intercepts and slopes) of all models. The coefficients are obtained estimating Equation (1) by OLS for (i) female owners, (ii) male owners, and (iii) both groups. In the first case, this analysis obtains the female-owner model coefficients (ß FF ); in the second, the analysis obtains the coefficients of the male-owner model (ß MF ); and in the third, this work gets the coefficients from the pooled model over both groups (ß *). The report’s analysis estimates two specifications of Equation (1): i) a “core” specification and ii) an “expanded” specification. The “core” specification is common to all countries and it includes the following controls: Firm characteristics: Number of employees (measured in hours of work), capital stock (inventory + equipment + property/land value), business practices, has a bank account, separates business from household money, formalization (firm is registered, pays the city assembly, pays taxes), innovation, firm has a loan, firm has at least one additional worker, sector. Owner characteristics: Age, tenure (when there is no information on tenure, firm age is used as a proxy), marital status, household size, number of children per adult, education. The “expanded” specification includes a set of additional covariates to the “core specification” to explain a greater share of the gender gap. The added variables are not all the same among the countries. As mentioned above, coefficients are interpreted as correlations and not as causal effects. Fortin, Lemieux, and Firpo (2011) present a detailed explanation of the assumptions required to identify the population parameters of interest. They impose two identification restrictions: (i) overlapping support and (ii) ignorability. Overlapping support implies that no single value of X = x or ε = e exists to fully identify female firms’ productivity. Ignorability refers to the random assignment of female owners in firms conditional on observables attributes. These assumptions as Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa additive linearity and zero conditional mean (i.e., ε is independent of X) are requirements to identify the individual contribution of each covariate in detailed decomposition. It should be noted that even the last both assumptions may not hold, the aggregate decomposition would remain valid as long as overlapping support and ignorability assumptions are accomplished. Appendix & Endnotes 178 179 Appendix 2 Description of datasets used in analysis Name Country Year Number of observations Modules Source Men Womena,b,c Total Impact Evaluation Surveys Entreprenant Initiative Survey Benin 2015 366 606 972 Owner Information. Firm information. Information about finances and credits. Capital, income, Benhassine, McKenzie, Pouliquen, and expenses and profit of the firm. Firm status. Taxes. Features and Practices of the firm. Santini (2018) Grants for Microenterprises Survey Ghana 2009 337 503 840 Introduction. Personal and business information. Balance sheet information. Income statement Fafchamps, McKenzie, Quinn, and information. Informality. Household roster and expenditures. Woodruff (2014) Tailoring Enterprises Survey Ghana 2008 65 88 153 Business overview. Staff. Current assets. Finance. Income. Karlan, Knight, and Udry (2015) Female Enterprise Survey Kenya 2013 0 3.426 3.426 Personal and business information. Family background and childhood. Finance and loans Mckenzie and Puerto (2017) information. Assets, income, expenses, and profit. Competitive environment. Business skills and practices. Empowerment. Business Registration Impact Evaluation Survey Malawi 2011 1.807 1.195 3.002 Basic information about the business. Premises and equipment. Finance. Business registration. Campos, Goldstein, and McKenzie (2015) Profits, revenues and expenditures. Financial literacy and preferences. Workers. Training and networks. Matching Grant Scheme for Business Mozambique 2012 831 364 1.195 Owner information. Activity. Installations and equipment. Financing. Bank account and other Campos and Montalvao (2015) Performance Survey financial services. Evaluation of practices and competencies. Profits, revenues and expenses. Business registration. Workers. Growth and Employment Survey Nigeria 2016 3.047 1.254 4.301 Business information. Profits, sales and expenditures. Finance. Anderson-McDonald, Buba, and McKenzie (2017) Business Plan Competition Survey Nigeria 2012 2.248 459 2.707 Demographic characteristics and socioeconomic background. Personal information. Employment Mckenzie (2015) Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa in your firm. Financing your business. Business financials. Business practices. Online Marketplace Survey South Africa 2012 1.771 846 2.617 Owner information. Profits and revenues. Capital. Labor. Business practices. Bossuroy, Campos, Coville, Goldstein, Roberts, and Sequeira (2012) Virtual Business Incubator Tanzania 2010 0 821 821 Basic information about the business. Employment status. Time use. Finance. Profits, revenues Bardasi, Gassier, Goldstein, and Holla and expenditures. Skills. Harassment. (2017) Training for Informal Firms Survey Togo 2012 711 789 1.500 Contact information. Information about the start of the business and the business sector. Labor. Campos, Frese, Goldstein, Iacovone, Financial. Activities related to the firm. Personality and attitudes. Education and work experience. Johnson, McKenzie, and Mensmann (2017) Kassida Survey Uganda 2011 408 325 733 Enterprise indentifying information. Employees. Premises and equipment. Profits, revenues and Campos, Goldstein, Pimhidzai, Stein, Zia expenditures. Finance. Regulation, standards and security. Training. (2013) Loans, Grants, and Training Impact Evaluation Uganda 2012 130 232 362 Owner information. Profits and revenues. Capital. Labor. Finances. Business practices.e Fiala (2013, 2015) Survey Census/household data Enterprise Census (Manufacturing) Ethiopia 2011 1.091 161 1.252 General information. Sales. Costs. Investments. Appendix & Endnotes Enterprise Census Ghana 2003 1.965 955 2.920 Organization. Industrial activity. Accounting records. Persons engaged. National Household Survey DRC 2012 1.728 2.734 4.462 Sociodemographic characteristics. UPI filters. Firm characteristics. Labor. Profits and sales. Expenditures and expenses. Equipment, investment and financing. Problems and prospects. 180 181 Name Country Year Number of observations Modules Source Men Womena,b,c Total Enterprise Surveys Formal firms with more than 5 employees Angola 2010 136 207 343 General information. Sales and supplies. Finance. Labor. Investment Climate Constraints. World Bank Benin 2009 85 48 133 Botswana 2010 108 150 258 Burkina Faso 2009 214 104 318 Burundi 2014 97 60 157 Cameroon 2009 200 149 349 Cape Verde 2009 74 74 148 Central African Republic 2011 72 76 148 Chad 2009 113 33 146 Congo 2009 116 13 129 Côte d'Ivoire 2009 343 169 512 DRC 2013 438 89 527 Eritrea 2009 72 68 140 Ethiopia 2011 399 227 626 Gabon 2009 150 14 164 Ghana 2013 501 211 712 Kenya 2013 428 350 778 Lesotho 2009 103 17 120 Liberia 2009 81 15 96 Madagascar 2013 285 186 471 Malawi 2014 355 152 507 Mali 2010 86 111 197 Mauritania 2014 125 25 150 Mauritius 2009 185 187 372 Namibia 2014 309 262 571 Niger 2009 46 31 77 Nigeria 2014 2.123 469 2.592 Rwanda 2011 121 97 218 Senegal 2014 488 105 593 Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Sierra Leone 2009 140 8 148 South Sudan 2014 585 149 734 Sudan 2014 601 54 655 Tanzania 2013 602 197 799 Togo 2009 109 29 138 Uganda 2013 508 239 747 Zambia 2013 435 281 716 Formal firms with less than 5 employees Burkina Faso 2009 65 19 84 General information. Sales and supplies. Finance. Labor. Investment Climate Constraints. World Bank Cameroon 2009 74 42 116 Cape Verde 2009 57 46 103 Côte d'Ivoire 2009 66 26 92 DRC 2013 305 106 411 Ethiopia 2011 96 48 144 Appendix & Endnotes Kenya 2013 196 161 357 Madagascar 2009 67 45 112 Mauritius 2009 39 36 75 Rwanda 2011 81 64 145 Togo 2009 93 24 117 182 183 Name Country Year Number of observations Modules Source Men Womena,b,c Total Informal firms Angola 2010 71 43 114 General information. Sales and supplies. Finance. Labor. Investment Climate Constraints. World Bank Botswana 2010 52 46 98 Burkina Faso 2009 93 27 120 Cameroon 2009 76 46 122 Cape Verde 2009 47 82 129 Côte d'Ivoire 2009 82 45 127 DRC 2013 364 116 480 Ghana 2013 251 472 723 Kenya 2013 316 214 530 Madagascar 2009 52 74 126 Mali 2010 92 27 119 Mauritius 2009 81 45 126 Rwanda 2011 145 95 240 Household Surveys International Income Distribution Database (I2D2) Angola 2008 22.059 23.339 45.398 Labor World Bank Burundi 1998 15.656 16.918 32.574 Burkina Faso 2009 27.571 29.584 57.155 Botswana 2009 12.965 14.246 27.211 Central African Republic 2003 14.897 15.725 30.622 Côte d'Ivoire 2008 30.388 29.311 59.699 Congo 2005 12.827 13.772 26.599 Comoros 2004 8.936 9.021 17.957 Cabo Verde 2007 16.311 17.445 33.756 Ethiopia 2012 33.846 37.127 70.973 Gabon 2005 18.375 19.228 37.603 Ghana 2012 28.460 29.518 57.978 Guinea 2002 27.336 29.426 56.762 Gambia 2010 18.468 20.054 38.522 Guinea Bissau 1993 13.217 14.045 27.262 Kenya 2005 32.918 33.807 66.725 Liberia 2007 9.936 10.098 20.034 Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Lesotho 2010 10.155 10.836 20.991 Madagascar 2010 29.394 29.980 59.374 Mali 2003 20.805 20.675 41.480 Mozambique 2008 24.368 26.809 51.177 Mauritania 2008 35.330 39.791 75.121 Mauritius 2012 19.946 20.562 40.508 Malawi 2010 27.555 28.848 56.403 Namibia 2003 24.443 22.093 46.536 Niger 2011 12.405 12.720 25.125 Nigeria 2012 29.416 30.134 59.550 Rwanda 2010 32.490 35.908 68.398 Senegal 2011 79.627 87.992 167.619 Sierra Leone 2011 18.337 19.130 37.467 São Tomé and Principe 2010 6.961 7.104 14.065 Swaziland 2009 6.717 7.428 14.145 Appendix & Endnotes Chad 2003 19.440 20.105 39.545 a IE datasets: Owner gender. bCensus data: For Ethiopia, it considers women if the owners of the firm are all females. Togo 2011 14.521 15.260 29.781 For Ghana, it considers the owner gender. cEnterprise Surveys: number of firms that have at least one female owner. c I2D2: gender of the individual surveyed. dMain groups of variables used it (South Africa questionnaire has not Tanzania 2011 32.059 33.862 65.921 modules or sections). eMain groups of variables used it. Uganda 2010 8.779 9.137 17.916 DRC 2005 35.978 36.707 72.685 Zambia 2010 49.961 52.240 102.201 184 185 References Banerjee, Abhijt, Esther Duflo, Nathanael Goldberg, Inducing Informal Firms to Formalize Make Sense? Dean Karlan, Robert Osei, William Pariente, Jeremy Experimental Evidence from Benin.” Journal of Shapiro, Bram Thuysbaert, and Christopher Udry. Public Economics. 157, pages 1-14. (2015a). “A multifaceted program causes lasting progress for the very poor: Evidence from six Berge, Lars Ivar Oppedal, Kjetil Bjorvatn, and Bertil countries.” Science. 348 (6236): 1260799. Tungodden. (2012). “Human and financial capital for micro enterprise development: Short-term Banerjee, Abhijit, Emily Breza, Esther Duflo, and and long-term evidence from a field experiment in Cynthia Kinnan. (2015b). “Do Credit Tanzania.” Dartmouth University. Adoho, Franck, Shubha Chakravarty, Dala Korkoyah, Anderson, Stephen, Rajesh Chandy, and Bilal Zia. Constraints Limit Entrepreneurship? Heterogeneity Mattias Lundberg, and Afia Tasneem. (2014). “The (2016). “Pathways to Profits: Identifying Separate in the Returns to Microfinance.” Buffett Institute Berge, Lars Ivar Oppedal, Kjetil Bjorvatn, Armando impact of an adolescent girls employment program: Channels of Firm Growth through Business Global Poverty Research Lab Working Paper. 17-104. Pires, and Bertil Tungodden. (2015). “Competitive the EPAG project in Liberia.” World Bank Policy Training.” World Bank Policy Research Working in the lab, successful in the field?” Journal of Research Working Paper (6832). Paper (7774). Baland, Jean-Marie, Isabelle Bonjean, Catherine Economic Behavior & Organization Volume 118, Part Guirkinger, and Roberta Ziparo. (2016). “The C, pages 303-317. African Development Bank. (2015). “African Women: Angelucci, Manuela and Robert Garlick (2016). economic consequences of mutual help in extended An Agenda for Action - Africa Gender Equality Index “Heterogeneity in the Efficiency of Intrahousehold families.” Journal of Development Economics. vol. Berge, Lars Ivar Oppedal, Kartika Juniwaty, and 2015.” Resource Allocation: Empirical Evidence and 123 (C). pages 38-56. Linda Sekei. (2016). “Gender composition and group Implications for Investment in Children.” Draft dynamics: Evidence from a laboratory experiment Agyei-Holmes, Andrew, Niklas Buehren, Markus paper. Bardasi, Elena, Shwetlena Sabarwal, and Katherine with microfinance clients”. Journal of Economic Goldstein, Robert Osei, Isaac Osei-Akoto, and Terrell. (2011). “How Do Female Entrepreneurs Behavior & Organization Volume 131, Part A, Pages Christopher Udry. (2018). “The Effects of Land Title Arias, Omar, David Evans, and Indhira Santos (2017). Perform? Evidence from Three Developing 1-20. Registration on Tenure Security, Access to Credit, “The Skills Balancing Act in Sub-Saharan Africa”. Regions.” Small Business Economics 37.4. Investment and Production: Evidence from Ghana”. World Bank. Bellemare, Marc and Casey Wichman. (2018). Draft paper. Bardasi, Elena, Marine Gassier, Markus Goldstein, “Elasticities and the Inverse Hyperbolic Sine Arraiz, Irani, Miriam Bruhn, Claudia Ruiz Ortega, and Alaka Holla. (2017). “The profits of wisdom: Transformation”. Manuscript. Aker, Jenny, Rachid Boumnijel, Amanda McClelland, and Rodolfo Stucchi. (2017). “Are psychometric tools The impacts of a business support program in and Niall Tierney. (2016). “Payment Mechanisms a viable screening method for small and medium- Tanzania.” World Bank Policy Research Working Berlinski, Samuel and Sebastian Galiani. (2007). and Antipoverty Programs: Evidence from a size enterprise lending? evidence from Peru”. World Paper (8279). ”The effect of a large expansion of preprimary Mobile Money Cash Transfer Experiment in Niger.” Bank Policy Research Working Paper (8276). school facilities on preschool attendance and Economic Development and Cultural Change 65, 1, Barros, Ricardo Paes de, Pedro Olinto, Trine Lunde maternal employment,” in Labour Economics, 14, pages 1-37. Ashraf, Nava. (2009). “Spousal Control and Intra– and Mirela Carvalho. (2011). “The Impact of Access pages 665-680. Household Decision Making: An Experimental Study to Free Childcare on Women’s Labor Market Ali, Daniel Ayalew, Matthew Collin, Klaus Deininger, in the Philippines.” American Economic Review 99, Outcomes: Evidence from a Randomized Trial in Bernhardt, Arielle, Erica Field, Rohini Pande, and Stefan Dercon, Justin Sandefur, and Andrew Zeitlin. no. 4, pages 1245–1277. Low-income Neighborhoods of Rio de Janeiro.” Natalia Rigol. (2017). “Household Matters: Revisiting (2014). “The Price of Empowerment: Experimental the Returns to Capital among Female Micro- Evidence on Land Titling in Tanzania.” World Bank Ashraf, Nava, Natalie Bau, Nathan Nunn, and Bastian, Gautam, Iacopo Bianchi, Markus Goldstein, entrepreneurs.” NBER Working Paper (23358). Policy Research Working Paper (6908). Alessandra Voena (forthcoming). “Bride Price and and Joao Montalvao. (2018a). “Short-Term Impacts Female Education.” Journal of Political Economy. of Improved Access to Mobile Savings, with and Bertrand, Marianne, Erzo Luttmer, and Sendhil Ali, Daniel Ayalew, Klaus Deininger, and Markus without Business Training: Experimental Evidence Mullainathan. (2000). “Network Effects and Welfare Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Goldstein. (2014). “Environmental and Gender Aterido, Reyes, Thorsten Beck and Leonardo from Tanzania. CGD Working Paper 478. Cultures.” The Quarterly Journal of Economics, 115 Impacts of Land Tenure Regularization in Iacovone. (2011). “Gender and finance in Sub- (3), pages 1019-105. Africa: Pilot evidence from Rwanda.” Journal Saharan Africa: Are women disadvantaged?” World Bastian, Gautam, Mehrab Bakhtiar, Eliana Carranza of Development Economics. Volume 110, pages Bank Policy Research Working Paper (5571). and Markus Goldstein. (2018b). “Training Mentors? Bertrand, Marianne, Emir Kamenica, and 262–275. Experimental Evidence from Female Owned Jessica Pan (2015). “Gender Identity and Relative Atkin, David, Amit Khandelwal, and Adam Osman. Microenterprises in Ethiopia.” Draft paper. Income within Households.” Quarterly Journal of Alibhai, Salman, Niklas Buehren, and Sreelakshmi (2014). “Exporting and Firm Performance: Evidence Economics, 130 (2), pages 571-614. Papineni. (2015). “Female Entrepreneurs Who from a Randomized Trial.” CEPR Discussion Paper Batista, Cátia and Julia Seither. (2018). “Reference Succeed in Male-Dominated Sectors in Ethiopia.” (10276). Points and Entrepreneurship: Evidence from a Beyer, Sylvia. (1990). “Gender Differences in the World Bank Africa Gender Innovation Lab. Policy Field Experiment on Aspirations, Goal Setting, and Accuracy of Self-Evaluations of Performance”. Brief Issue 12. Audretsch, David and Maryann Feldman. (2004). Business Skills”. CSAE 2018 working papers. Journal of Personality and Social Psychology, 59 (5), “Knowledge Spillovers and the Geography of 960-970. Alibhai, Salman, Niklas Buehren, and Sreelakshmi Innovation.” in Henderson and Thisse eds., Beaman, Lori, Niall Keleher, and Jeremy Magruder. Papineni. (2016). “From Learning to Earning: An Handbook of Urban and Regional Economics, vol. 4. (2017). “Do job networks disadvantage women? Bicchieri, Cristina. (2006). “The Grammar of Society: Impact Evaluation of the Digital Opportunity Trust Evidence from a recruitment experiment in Malawi.” The Nature and Dynamics of Social Norms.” (DOT) Entrepreneurship Training.” World Bank Balafoutas, Loukas and Matthias Sutter. (2010). Journal of Labor Economics, vol. 36, no. 1. Cambridge University Press. Africa Gender Innovation Lab. Policy Brief Issue 17. “Gender, competition and the efficiency of policy interventions.” Working Papers in Economics and Beckmann, Daniela and Lukas Menkhoff. (2008). Blackden, Mark and Quentin Wodon. (2006). Appendix & Endnotes Amanatullah, Emily, and Michael Morris. (2010). Statistics, No. 2010-12. “Will women be women?: Analyzing the gender “Gender, Time Use, and Poverty in Sub-Saharan “Negotiating Gender Roles: Gender Differences in difference among financial experts.” Discussion Africa.” World Bank Research Working Paper (73). Assertive Negotiating Are Mediated by Women’s Banerjee, Abhijit, Dean Karlan, and Jonathan papers // School of Economics and Management of Fear of Backlash and Attenuated When Negotiating Zinman. (2015). “Six Randomized Evaluations of the Hanover Leibniz University, No. 391. Blair-Loy, Mary. (2003). “Competing Devotions: on Behalf of Others”. Journal of Personality and Microcredit: Introduction and Further Steps.” Career and Family Among Women Executives”. Social Psychology, 98(2), 256-267. American Economic Journal: Applied Economics. Benhassine, Najy, David McKenzie, Victor Harvard University Press, Cambridge, 7(1), pages 1-21. Pouliquen, and Massimiliano Santini. (2018). “Does Massachusetts. 186 187 Blattman, Christopher; Eric Green, Jeannie Annan, Bruhn, Miriam and David McKenzie. (2013). “Entry Campos, Francisco, Markus Goldstein, Laura and the Unrealized Promise of Technological Catch- and Julian Jamison. (2013). “Building Women’s Regulation and Formalization of Micro enterprises McGorman, Ana Maria Munoz Boudet, and Obert Up.” Washington, DC: World Bank. Economic and Social Empowerment through in Developing Countries.” World Bank Policy Pimhidzai. (2015). “Breaking the metal ceiling: Enterprise: An Experimental Assessment of the Research Working Paper (6507). Female entrepreneurs who succeed in male- Clark, Shelley, Caroline Kabiru, Soni Laszlo, and Women’s Income Generating Support Program in dominated sectors.” World Bank Policy Research Stella Muthuri. (2017). “Can Subsidized Early Child Uganda”. Logica study series; no. 1. World Bank, Buehren, Niklas, Markus Goldstein, and João Working Paper (7503). Care Promote Women’s Employment?: Evidence Washington, DC. Montalvão. (2016). “Girls’ Empowerment, Sibling from a Slum Settlement in Africa.” Research Rivalry, and Competitiveness: Evidence from a Campos, Francisco, Michael Frese, Markus Contribution Paper. GWP-2017-05. Boltz, Marie, Karine Marazyan and Paola Villar. randomized control trial and a lab experiment in Goldstein, Hillary Johnson, Leonardo Iacovone, (2017). “Income Hiding and Informal Redistribution: Uganda.” World Bank Policy Research Working David McKenzie, and Mona Mensmann. (2017). Costa, Rita and Bob Rijkers. (2012). “Gender A Lab-in-the-Field Experiment in Senegal.” Working Paper (7699). “Teaching personal initiative beats traditional and Rural Non-Farm Entrepreneurship.” World Papers 20170004, Université Paris 1 Panthéon training in boosting small business in West Africa.” Development 40. Sorbonne, UMR Développement et Sociétés. Bulte, Erwin, Robert Lensink, and Nhung Vu. (2017). Science 357 (6357), pages 1287-1290 “Do Gender and Business Trainings Affect Business Correll, Shelley. (2001). “Gender and the Career Botha, Melodi, Gideon Nieman, and Jurie Outcomes? Experimental Evidence from Vietnam.” Campos, Francisco, Michael Frese, Markus Choice Process: The Role of Biased Self- van Vuuren. (2006). “Evaluating the Women Management Science 63 (9), pages 2773-3145. Goldstein, Hillary Johnson, Leonardo Iacovone, Assessments.” American Journal of Sociology 106 Entrepreneurship Training Programme.” David McKenzie, and Mona Mensmann. (2018). (6), pages 1691–1730. International Indigenous Journal of Buvinic, Mayra and Rebecca Furst-Nichols. (2014). “Is Personal Initiative Training a Substitute or Entrepreneurship, Advancement, Strategy and “Promoting Women’s Economic Empowerment: Complement to the Existing Human Capital of Correll, Shelley. (2004). “Constraints into Education, 2, pages 1–16. What Works?” World Bank Policy Research Working Women? Results from a Randomized Trial in Togo.” Preferences: Gender, Status, and Emerging Career Paper (7087). AEA Papers and Proceedings 2018, 108, pages Aspirations.” American Sociological Review 69 (1), Bossuroy, Thomas, Francisco Campos, Aidan 256–261 pages 93–113. Coville, Markus Goldstein, Gareth Roberts, and Buvinic, Mayra and Megan O’Donnell. (2016). Sandra Sequeira. (2013). “Shape Up and Ship Out? “Revisiting What Works: Women, Economic Campos, Francisco, Markus Goldstein, and David Crant, Michael. (1995). “The proactive personality Gender Constraints to Growth and Exporting in Empowerment and Smart Design.” Center for McKenzie. (2018). “How should the government scale and objective job performance among real South Africa,” in: Women and Trade in Africa: Global Development. bring small firms into the formal system? estate agents.” Journal of Applied Psychology, 80, Realizing the Potential, pages 129-165. Experimental evidence from Malawi.” World Bank pages 532-37. Cadot, Olivier, Ana Fernandes, Julien Gourdon, and Policy Research Working Paper (8601). Bowles, Hannah. (2012). “Psychological Perspectives Aaditya Mattoo. (2015). “Are the benefits of export Croke, Kevin, Markus Goldstein, Edouard Mensah, on Gender in Negotiation.” HKS Faculty Research support durable? Evidence from Tunisia.” Journal of Castilla, Carolina and Thomas Walker. (2013). and Michael O’Sullivan. (2018). “ Up Before Dawn: Working Paper Series RWP12-046: John F. Kennedy International Economics. Volume 97, pages 310-324. “Is Ignorance Bliss? The Effect of Asymmetric Cross-border Trade Experiment in Eastern School of Government, Harvard University. Information between Spouses on Intra-household Democratic Republic of Congo.” Draft paper. Cai, Jing and Adam Szeidl. (2018). “Interfirm Allocations.” American Economic Review, 103 (3), Brenton, Paul, Elisa Gamberoni, Catherine Sear, Relationships and Business Performance.” The pages 263-68. Croke, Kevin, Markus Goldstein and Alaka Holla. Maria Elena Garcia Mora, Sabrina Roshan, Louis Quarterly Journal of Economics, 133 (3), pages (2017). “Can job training decrease women’s self- Njie Ndumbe, Susan Ityavyar, John Baffes, Laura 1229–1282. Cech, Erin, Brian Rubineau, Susan Silbey, and defeating biases? experimental evidence from Maratou-Kolias, Antoine Coste, Nora Dihel, Michelle Caroll Seron. (2011). “Professional Role Confidence Nigeria.” World Bank Policy Research Working Christian, Jose Guilherme Reis, Thomas Bossuroy, Campbell, Jacquelyn. (2002). “Health Consequences and Gendered Persistence in Engineering.” Paper (8141). Francisco Campos, Aidan Coville, Markus Goldstein, of Intimate Partner Violence.” The Lancet 359, no. American Sociological Review 76(5), pages 641–66. Gareth Roberts, Sandra Sequeira, Josaphat Kweka, 9314, pages 1331–36. Croson, Rachel, and Uri Gneezy. (2009). “Gender Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa and Mahjabeen Haji. (2013). “Women and trade in Chakravarty, Shubha, Smita Das, and Julia Vaillant. Differences in Preferences.” Journal of Economic Africa: realizing the potential.” World Bank Group. Campos, Francisco, Markus Goldstein, Obert (2017). “Gender and youth employment in Sub- Literature 448–74. Pimhidzai, Mattea Stein, and Bilal Zia. (2013). Saharan Africa: a review of constraints and effective Breza, Emily and Cynthia Kinnan. (2017). “Measuring “Financial management and vocational training in interventions.” World Bank Policy Research Working Cull, Robert and Jonathan Morduch. (2017). the Equilibrium Impacts of Credit: Evidence from Uganda: Impact and network effects in informal Paper (8245). “Microfinance and Economic Development.” World the Indian Microfinance Crisis.” Northwestern industrial clusters.” World Bank Group – Enhancing Bank Policy Research Working Paper (8252). Institute for Policy Research. financial capability and behavior in low-and middle- Cherchi, Ludovica, Markus Goldstein, James income countries. Habyarimana, Joao Montalvão, Michael O’Sullivan, Dahal, Mahesh and Nathan Fiala (2018). “What do Brixiová, Zuzana and Thierry Kangoye. (2015). and Christopher Udry. (2018). “Incentives for Joint we know about the impact of microfinance? The “Gender and Constraints to Entrepreneurship in Campos, Francisco, Aidan Coville, Ana Fernandes, Land Titling: Experimental Evidence from Uganda.” problems of power and precision.” Ruhr Economic Africa: New Evidence from Swaziland.” IZA DP No. Markus Goldstein and David McKenzie. (2014). Draft paper. Papers, No. 756, ISBN 978-3-86788-880-6, RWI – 9273. “Learning from the experiments that never Leibniz-Institut für Wirtschaftsforschung, Essen. happened: Lessons from trying to conduct Cho, Yoonyoung and Maddalena Honorati. (2013). Brooks, Wyatt, Kevin Donovan, Terence randomized evaluations of matching grant “Entrepreneurship programs in developing Davidsson, Per and Benson Honig. (2003). “The Johnson (forthcoming). “Mentors or Teachers? programs in Africa,” Journal of the Japanese and countries: A Meta regression analysis.” World Bank role of social and human capital among nascent Microenterprise Training in Kenya.” American International Economies, 33, pages 4-24. Social Protection Discussion Papers (77168). entrepreneurs.” Journal of Business Venturing, 18, Economic Journal: Applied Economics. pages 301-331. Campos, Francisco, Markus Goldstein, and Cirera, Xavier and Qursum Qasim. (2014). Appendix & Endnotes Brüderl, Josef and Peter Preisendörfer. (1997). David McKenzie. (2015). “Short-Term Impacts of “Supporting Growth-Oriented Women De Mel, Suresh, David McKenzie, and Christopher “Network Support and the Success of Newly Formalization Assistance and a Bank Information Entrepreneurs: A Review of the Evidence and Woodruff. (2009). “Are Women More Credit Founded Business.” Small Business Economics, 10, Session on Business Registration and Access to Key Challenges.” Innovation, Technology & Constrained? Experimental Evidence on Gender pages 213-225. Finance in Malawi.” World Bank Policy Research Entrepreneurship Policy Note. and Microenterprise Returns.” American Economic Working Paper (7183). Journal: Applied Economics 1.3. Cirera, Xavier and William Maloney. (2017). “The Innovation Paradox: Developing-Country Capabilities 188 189 Demirguc-Kunt, Asli, Thorsten Beck, and Patrick Duvendack, Maren, Richard Palmer-Jones, James Fiala, Nathan. (2017). “Business is tough, but family Glaub, Matthias and Michael Frese. (2011). “A Critical Honohan. (2008). “Finance for All?: Policies and Copestake. Lee Hooper, Yoon Loke, Yoon, and Nitya is worse: Household bargaining and investment in Review of the Effects of Entrepreneurship Training Pitfalls in Expanding Access.” Washington, D.C.: Rao. (2011). “What is the evidence of the impact of microenterprises in Uganda.” Working Paper. in Developing Countries’ Enterprise.” Development World Bank. microfinance on the well-being of poor people?” and Microfinance 22 (4), pages 335–53. London: EPPI-Centre, Social Science Research Field, Erica, Seema Jayachandran, and Rohini Deininger, Klaus, Aparajita Goyal, and Hari Unit, Institute of Education, University of London. Pande. (2010). “Do Traditional Institutions Glaub, Matthias, Michael Frese, Sebastian Fischer, Nagarajan. (2013). “Women’s inheritance rights Constrain Female Entrepreneurial Investment? and Maria Hoppe. (2014). “Increasing Personal and intergenerational transmission of resources in Emran, Shahe, Mahbub Morshed, and Joseph A Field Experiment on Business Training in Initiative in Small Business Managers/Owners India.” Journal of Human Resources 48(1), pages Stiglitz. (2011). “Microfinance and Missing Markets.” India.” American Economic Review Papers and Leads to Entrepreneurial Success: A Theory- 114-141. MPRA Paper 41451, University Library of Munich, Proceedings. based Controlled Randomized Field Intervention Germany. for Evidence Based Management.” Academy of Diwan, Faizan, Grace Makana, David McKenzie, and Field, Erica, Rohini Pande, John Papp, and Y. Management Learning and Education 13, pages Silvia Paruzzolo. (2014). “Invitation Choice Structure Fafchamps, Marcel and Bart Minten. (1999). Jeanette Park. (2012). “Repayment Flexibility Can 354-379. Has No Impact on Attendance in a Female Business “Relationships and Traders in Madagascar.” Journal Reduce Financial Stress: A Randomized Control Training Program in Kenya.” PLOS ONE. 9(10): of Development Studies 35(6), pages 1-35. Trial with Microfinance Clients in India.” PLoS One, The Global Entrepreneurship and Development e109873. Vol. 7 (9): e45679. Institute – GEDI (2018). “The 2018 Global Fafchamps, Marcel. (2001).”Networks, Communities, Entrepreneurship Index.” Donald, Aletheia, Gayatri Koolwal, Jeannie Annan, and Markets in Sub-Saharan Africa: Implications Field, Erica, Seema Jayachandran, Rohini Pande, Kathryn Falb, and Markus Goldstein. (2017). for Firm Growth and Investment.” Journal of African and Natalia Rigol. (2016a). “Friendship at Work: Can Global Entrepreneurship Research Association – “Measuring women’s agency.” World Bank Policy Economies, 10 (2), pages 109-142. Peer Support Stimulate Female Entrepreneurship?” GERA. (2018). Global Entrepreneurship Monitoring Research Working Paper (8148). American Economic Journal: Economic Policy, 8(2), 2016-2017. Fafchamps, Marcel and Simon Quinn. (2013). “Social pages 125-53. Donald, Aletheia, Julia Vaillant, Francisco Campos, Networks and Business Practices: Evidence from a Goldstein, Markus, Kenneth Houngbedji, Florence and Maria Emilia Cucagna. (2018). “Caring about Randomized Experiment with Microentrepreneurs.” Field, Erica, Rohini Pande, Natalia Rigol, Simone Kondylis, Michael O’Sullivan, and Harris Selod. Carework : Lifting Constraints to the Productivity of Draft paper. Schaner, and Charity Troyer Moore. (2016b). “On (2016). “Formalizing rural land rights in West Africa: Women Farmers in the Democratic Republic of the Her Account: Can Strengthening Women’s Financial early evidence from a randomized impact evaluation Congo.” World Bank, Washington, DC. Fafchamps, Marcel, David McKenzie, Simon Quinn, Control Boost Female Labor Supply?.” Draft paper. in Benin.” World Bank Policy Research Working Simon, and Christopher Woodruff. (2014). “Micro- Paper (7435). Doyle, Kate, Ruti Levtov, Gary Barker, Gautam enterprise Growth and the Flypaper Effect: Evidence Fortin, Nicole, Thomas Lemieux, Sergio Firpo. Bastian, Jeffrey Bingenheimer, Shamsi from a Randomized Experiment in Ghana.” Journal (2011). “Decomposition Methods in Economics.” in Gneezy, Uri, Kenneth Leonard, and John List. Kazimbaya, Anicet Nzabonimpa, Julie Pulerwitz, of Development Economics 106, pages 211–26. “Handbook of Labor Economics”. Elsevier, pages (2009). “Gender Differences in Competition: Felix Sayinzoga, Vandana Sharma, Vandana, and 1-102. Evidence from a matrilineal and a patriarchal Dominick Shattuck. (2018). “Gender-transformative Fafchamps, Marcel and Simon Quinn. (2016). society.” Econometrica, 77 (5), pages 1637–1664. Bandebereho couples’ intervention to promote male “Networks and Manufacturing Firms in Africa: Frese, Michael, Stefanie Krauss, Nina Keith, engagement in reproductive and maternal health Results from a Randomized Field Experiment.” The Susanne Escher, Rafal Grabarkiewicz, Siv Luneng, Granovetter, Mark. (1973). “The Strength of Weak and violence prevention in Rwanda: Findings from World Bank Economic Review, lhw057. Constanze Heers, Jens Unger, and Christian Ties.” American Journal of Sociology 78 (6), pages a randomized controlled trial.” PLOS ONE. 13 (4): Friedrich. (2007). “Business owners’ action planning 1360-1380. e0192756. Fafchamps, Marcel and Christopher Woodruff. and its relationship to business success in three (2016). “Identifying gazelles: expert panels vs. African countries.” Journal of Applied Psychology, Granovetter, Mark. (1983): “The Strength of Weak Drexler, Alejandro. (2014). “Learning Exercise: surveys as a means to identify firms with rapid 92 (6), pages 1481-1498. Ties: A Network Theory Revisited.” in Randall Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa High-Potential Women Entrepreneurship Training growth potential.” World Bank Policy Research Collins (ed.), Sociological Theory, San Francisco, CA: Overview of Existing Programs.” World Bank Group. Working Paper (7647). Frese, Michael. (2009). “Towards a psychology of Jossey-Bass, pages 201–233. entrepreneurship - An action theory perspective.” Duflo, Esther (2003). “Grandmothers and FAO – Food and Agriculture Organization of the Foundations and Trends in Entrepreneurship, 5 (6), Greif, Avner. (1997): ‘‘Contracting, Enforcement, Granddaughters: Old-Age Pensions and United Nations. (2017). Gender and Land Rights pages 435-494. and Efficiency: Economics Beyond the Law.’’ in Intrahousehold Allocation in South Africa.” The Database. Annual World Bank Conference on Development World Bank Economic Review 17, no. 1, pages 1–25. Friedson-Ridenour, Sophia and Rachael Economics, Bruno and Pleskovic, eds. Fawole, Olufunmilayo, Ademola Ajuwon, and Kayode Pierotti (2018). “Competing Priorities: Women’s Duflo, Esther and Christopher Udry (2004). Osungbade. (2005). “Evaluation of interventions to Microenterprises and Household Relationships.” Grown, Caren, Mary Porter-Peschka, Natalie Louat, “Intrahousehold Resource Allocation in Côte prevent gender-based violence among young female World Bank Policy Research Working Paper (8580). Henriette Kolb, Gaelle Olivier, and Garance Wattez- d’Ivoire: Social Norms, Separate Accounts and apprentices in Ibadan, Nigeria.” Health Education, Richard. (2017). “She for shield: insure women to Consumption Choices.” National Bureau of 105(3), pages 186-203. Gertler, Paul (2018): “Presentation to World Bank better protect all.” World Bank Group. Economic Research. on “Making Entrepreneurs: The Returns to Training Ferraz, Claudio, Frederico Finan, and Dimitri Youth in Soft versus Hard Business Skills from an Grosse, Niels and Gerhard Riener. (2010). Dupas, Pascaline and Jonathan Robinson. (2013). Szerman. (2016). “Procuring Firm Growth: The At-Scale Field Experiment.” May 16, 2018. “Explaining gender differences in competitiveness: “Savings Constraints and Micro-Enterprise Effects of Government Purchases on Firm Gender-task stereotypes.” Jena Economic Research Development Evidence from a Field Experiment.” Dynamics.” Working Paper. Giné, Xavier and Ghazala Mansuri. (2014). “Money Papers, No. 2010.017. AEJ: Applied Economics 5(1), pages 163-92. or Ideas? A Field Experiment on Constraints to Appendix & Endnotes Fiala, Nathan. (2015). “Stimulating Microenterprise Entrepreneurship in Rural Pakistan.” World Bank Hagen-Zanker, Jessica, Luca Pellerano, Francesca Dupas, Pascaline, Dean Karlan, Jonathan Robinson Growth: Results from a Loans, Grants and Training Policy Research Working Paper (6959). Bastagli, Luke Harman, Valentina Barca, Georgina and Diego Ubfal. (2018). “Banking the Unbanked? Experiment in Uganda.” Grants and Training Sturge, Tanja Schmidt and Calvin Laing. (2016). Evidence from three countries.” American Experiment in Uganda. Gennari, Floriza, Diana Arango, and Nidia Hidalgo. “The impact of cash transfers on women and girls.” Economic Journal: Applied Economics Vol. 10 (2), (2015). “Violence against women and girls resource London: Overseas Development Institute. pages 257–97. guide: finance and enterprise development brief.” World Bank Group. 190 191 Hallward-Driemeier, Mary. (2013). “Enterprising International Finance Corporation. (2014). “Women- Kevane, Michael and Bruce Wydick. (2001). Mayoux, Linda. (1999). “From Access to Women: Expanding Economic Opportunities in Owned SMEs: A Business Opportunity for Financial “Microenterprise Lending to Female Entrepreneurs: Empowerment: Gender Issues in Micro-Finance.” Africa.” Rep. Washington, DC: World Bank – AFD. Institutions.” Sacrificing Economic Growth for Poverty CSD Virtual Conference. Alleviation?” World Development 29 (7), pages 1225- Hallward-Driemeier, Mary and Ousman Gajigo. International Labor Organization. (2018). “Gender 36. McKenzie, David. (2017). “Identifying and Spurring (2013). “Strengthening Economic Rights and and Entrepreneurship Together (GET Ahead).” High-Growth Entrepreneurship: Experimental Women’s Occupational Choice: The Impact of Women’s Entrepreneurship Development Kim, Seon Mi and Margaret Sherraden. (2014). Evidence from a Business Plan Competition.” Reforming Ethiopia’s Family Law.” World Bank Programme. “The Impact of Gender and Social Networks on American Economic Review 107 (8), pages 2278- Policy Research Working Paper (6695). Microenterprise Business Performance.” Journal of 2307. Instituto Promundo and CARE International in Sociology and Social Welfare 29 (4), pages 404-417. Hallward-Driemeier, Mary, Tazeen Hasan, and Anca Rwanda. (2012). “Journeys of Transformation: A McKenzie, David and Anna Luisa Paffhausen. (2017). Bogdana Rusu. (2013). “Women’s Legal Rights over Training Manual for Engaging Men as Allies in Klyver, Kim and Siri Terjesen. (2007). “Small Firm Death in Developing Countries.” CEPR 50 Years: What Is the Impact of Reform?” World Women’s Economic Empowerment.” “Entrepreneurial network composition: An analysis Discussion Paper No. DP12401. Bank Policy Research Working Paper (6617). across venture development stage and gender.” Instituto Promundo and International Center for Women in Management Review, 22 (8), pages 682- McKenzie, David and Christophe Woodruff. (2013). Harari, Mariaflavia (2018). “Women’s Inheritance Research on Women. (2014). “Engaging Men to 688. “What Are We Learning from Business Training Rights and Bargaining Power: Evidence from Prevent Gender-Based Violence: A Multi-Country and Entrepreneurship Evaluations around the Kenya.” Economic Development and Cultural Intervention and Impact Evaluation Study.” Koop, Sabine, Tamara De Reu, and Michael Frese. Developing World?.” World Bank Research Change. (2000). “Sociodemographic factors, entrepreneurial Observer. 29 (1), pages 48-82. International Trade Centre. (2015). “Unlocking orientation, personal initiative and environmental Hardy, Morgan, and Gisella Kagy. (2018a). “Mind markets for women.” problems in Uganda.” In: M. Frese (Eds.), Success McKenzie, David and Christopher Woodruff. (2017). The (Profit) Gap: Why Are Female Enterprise and Failure of Microbusiness Owners in Africa: A “Business practices in small firms in developing Owners Earning Less than Men?” AEA Papers and International Trade Centre. (2014). “Empowering Psychological Approach, 55-76. countries.” Management Science, 63 (9), pages Proceedings, 108: 1-4. Women through Public Procurement.” 2967-81. Lokshin, Michael, Elena Glinskaya and Marito Hardy, Morgan, and Gisella Kagy. (2018b). “It’s Iqbal, Sarah, Asif Mohammed Islam, Rita Ramalho, Garcia. (2004). “The Effect of Early Childhood McKenzie, David and Susana Puerto. (2017). Getting Crowded in Here: Experimental Evidence and Alena Sakhonchik. (2016). “Unequal before the Development Programmes on Women’s Labour “Growing Markets through Business Training for of Demand Constraints in the Gender Profit Gap.” law: measuring legal gender disparities across the Force Participation and Older Children’s Schooling Female Entrepreneurs: A Market-Level Randomized Draft paper. world.” World Bank Policy Research Working Paper in Kenya.” Journal of African Economies, Centre for Experiment in Kenya.” World Bank Policy Research (7803). the Study of African Economies (CSAE), vol. 13(2), Working Paper (7993). Healy, Andrew and Jennifer Pate. (2011). “Can Teams pages 240-276. Help to Close the Gender Competition Gap?” The Jack, William and Tavneet Suri. (2016). “Study: Mckenzie, David and Dario Sansone. (2017). “Man Economic Journal, 121 (555), pages 1192-1204. Mobile-money services lift Kenyans out of poverty.” Loscocco, Karyn, Shannon Monnat, Gwen Moore, vs. machine in predicting successful entrepreneurs: MIT News Office. and Kirsten Lauber. (2009). “Enterprising Women: A evidence from a business plan competition in Heckman James, Jora Stixrud, and Sergio Urzua. Comparison of Women’s and Men’s Small Business Nigeria.” World Bank Policy Research Working (2006). “The effects of cognitive and noncognitive Jakiela, Pamela and Owen Ozier. (2016). “Does Networks.” Gender & Society June 2009 23: pages Paper (8271). abilities on labor market outcomes and social Africa Need a Rotten Kin Theorem? Experimental 388-411. behavior.” Journal of Labor Economics, 24 (3), Evidence from Village Economies.” Review of McKinsey Global Institute. (2016). “Lions on pages 411–482. Economic Studies, 83 (1), pages 231-268. Luke, Nancy and Kaivan Munshi. (2011). “Women as the Move II: Realizing the Potential of African agents of change: Female income and mobility in Economies.” McKinsey and Company. Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Hicks, Joan Hamory, Michael Kremer, Isaac Mbiti, Kabeer, Naila. (1999). “Resources, Agency, India.” Journal of Development Economics 94 (1), and Edward Miguel. (2011). “Vocational Education Achievements: Reflections on the Measurement of pages 1-17. McMillan, John. (1997). ‘‘Markets in Transition.’’ Voucher Delivery and Labor Market Returns: A Women’s Empowerment”. Development and Change Advances in Economics and Econometrics: Theory Randomized Evaluation among Kenyan Youth.” 30 (3), pages 435-464. Lundeberg, Mary, Paul Fox, and Judith Punćcohaŕ. and Applications, Kreps and Wallis, eds. Cambridge Report for Spanish Impact Evaluation. (1994). “Highly confident but wrong: Gender University Press. Kamas, Linda, and Anne Preston. (2010). “Social differences and similarities in confidence Hoel, Jessica, Melissa Hidrobo, Tanguy Bernard, preferences, competitiveness and compensation: judgments.” Journal of Educational Psychology, 86 McMillan, John and Christopher Woodruff. (1999). and Maha Ashour. (2018) “Productive efficiency and Are there gender differences?.” Draft paper. (1), 114. “Interfirm Relationships and Informal Credit In cooperation between spouses: Evidence from dairy Vietnam.” The Quarterly Journal of Economics, 114 farming in Senegal. Draft paper. Kantor, Paula. (2005). “Determinants of Women’s Marcus, Rachel and Caroline Harper. (2014). (4), pages 1285-1320. Microenterprise Success in Ahmedabad, India: “Gender justice and social norms – processes of Hundley, Gregory. (2001). “Why women earn less Empowerment and Economics.” Feminist change for adolescent girls.” ODI. Meager, Rachael. (2016). “Understanding the than men in self-employment.” Journal of Labor Economics 11 (3), pages 63-83. Average Impact of Microcredit Expansions: A Research 22, 817-829. Martinez, Sebastian, Sophie Naudeau and Vitor Bayesian Hierarchical Analysis of 7 Randomized Kelly, Tim and Rachel Firestone (2016). “How Pereira. (2012). “The Promise of Preschool in Africa: Experiments.” Draft paper. Hussam, Reshmaan, Natalia Rigol, and Benjamin Tech Hubs are helping to Drive Economic Growth A Randomized Impact Evaluation of Early Childhood Roth. (2016). “Targeting High Ability Entrepreneurs in Africa.” World Bank WDR Digital Dividends Development in Rural Mozambique.” World Bank / Mhede, Edwin, Yuki Higuchi, Tetsushi Sonobe. Using Community Information: Mechanism Design background paper. Save the Children. (2018). “Medium-run Impact of Management Appendix & Endnotes In The Field.” MIT Economics. Working Paper. Training: Evidence from a Randomized Controlled Kes, Alisha, Krista Jacobs, and Sophie Namy. Mateo Díaz, Mercedes and Lourdes Rodriguez- Experiment in Tanzania.” Draft paper. International Telecommunications Union. (2016). (2011). “Gender differences in asset rights in central Chamussy. (2016). “Women, Childcare, and “World Telecommunications/ICT Development Uganda.” International Center for the Rights on Prosperity in Latin America and the Caribbean.” Microcredit Summit Campaign. (2015). “Mapping Report database.” Women. Washington, DC. Inter-American Development Bank. Pathways out of Poverty.” RESULTS Educational Fund. 192 193 Morris, Michael, Nola Miyasaki, Craig Watters, Constraints and Effective Interventions.” World Intrafamiliar en Colombia.” Universidad de los Inefficient Strategic Savings.” American Economic and Susan Coombes. (2006). “The Dilemma of Bank Policy Research Working Paper (8250). Andes-Cede. Journal: Applied Economics, 7 (2), pages 135-174. Growth: Understanding Venture Size Choices of Women Entrepreneurs.” Journal of Small Business OECD. (2016). “Tracking the Money for Women’s Ricardo, Christine, Marci Eads, and Gary Barker. Schaner, Simone. (2017). “The Cost of Convenience? Management, 44 (2),pages 221-244. Economic Empowerment: Still a Drop in the (2011). “Engaging boys and young men in the Transaction Costs, Bargaining Power, and Savings Bucket.” OECD, Paris. prevention of sexual violence: A systematic and Account Use in Kenya.” Journal of Human Morrison, Andrew and Maria Orlando. (2004). global review of evaluated interventions.” Pretoria: Resources, 52 (4), pages 919-945. “The Costs and Impacts of Gender-Based Overseas Development Institute. (2015). “How do South Africa: Oak Foundation. Violence in Developing Countries: Methodological Gender Norms Change?” Simavi, Sevi, Clare Manuel, and Mark Blackden. Considerations and New Evidence.” World Bank Ridgeway, Cecilia and Shelley Correll. (2004). (2010). “Gender Dimensions of Investment Report – 36151. Paryavi, Maliheh. (2016). “Representing Others in “Unpacking the Gender System: a Theoretical Climate Reform: A Guide for Policy Makers and Competitions: Why Women Lean Back.” Harvard Perspective on Gender Beliefs and Social Practitioners.” World Bank. Müller, Susan, Thierry Volery, and Björn Von University, Graduate School of Arts and Sciences. Relations.” Gender & Society 18 (4). Siemens. (2012). “What do entrepreneurs actually Soll, Jack and Joshua Klayman. (2004). do? An observational study of entrepreneurs’ Paryavi, Maliheh, Francisco Campos and Indhira Rodgers, Yana and Menon Nidhiya. (2012). “A Meta- “Overconfidence in interval estimates.” Journal of everyday behavior in the start-up and growth Santos. (2018). “Gender and willingness to compete: Analysis of Land Rights and Women’s Economic Experimental Psychology: Learning Memory, and stages.” Entrepreneurship Theory and Practice, 36 lab experiments in Kenya.” Draft paper. Well-Being.” Draft paper. Cognition, 30 (2), pages 299-314. (5), pages 995-1017. Patel, Payal. (2014). “Effectiveness of Robb, Alicia, Alexandria Valerio, and Brent Parton. Sutter, Matthias and Daniela Rützler. (2010). Munshi, Kaivan and Mark Rosenzweig. (2006). Entrepreneurship Development Interventions (2014). “Entrepreneurship Education and Training: “Gender differences in competition emerge early in “Traditional institutions meet the modern world: for women entrepreneurs.” An ILO-WED Issue Insights from Ghana, Kenya and Mozambique.” life.” Working Papers in Economics and Statistics, Caste, gender, and schooling choice in a globalizing Brief. International Labor Organization, Women World Bank. No. 2010-14. economy”. American Economic Review 96 (4), pages Enterprise Development. 1225-1252. RSA National Treasury (2017). “Preferential Tabbert, Falko. (2009). “Child Care and Female Platteau, Jean-Philippe. (2000). “Institutions, Procurement Policy Framework Act, 2000: Labor Force Participation in Developing Countries – Niederle, Muriel, and Lise Vesterlund. (2007). Social Norms, and Economic Development.” Preferential Procurement Regulations, 2017.” Quasi- Experimental Results from Togo.” Faculty of “Do women shy away from competition? Do men Fundamentals of Development Economics Series. Government Gazette no. 40553. Economics and Management, Leibniz-University of compete too much?” The Quarterly Journal of Amsterdam: Harwood Academic Publishers. Hannover. Economics, 122, pages 1067-1101. Roy, Sanchari. (2015). “Empowering women? Pierotti, Rachael. (2016). “Barriers for female Inheritance rights, female education and dowry Tarozzi, Alessandro, Jaikishan Desai, and Kristin Niederle, Muriel and Lise Vesterlund. (2008). entrepreneurs’ access to finance: a qualitative payments in India.” Journal of Development Johnson. (2015) “The Impacts of Microcredit: “Gender Differences in Competition.” Negotiation assessment in Ethiopia.” Presentation to World Economics, 2015, vol. 114, issue C, pages 233-251. Evidence from Ethiopia.” American Economic Journal (October), pages 447-463. Bank Africa Gender Innovation Lab. Journal: Applied Economics. 7 (1), pages 54-89. Ruble, Diane, Carol Martin, and Sheri Berenbaum. Niederle, Muriel and Lise Vesterlund. (2010). Pulford, Briony and Andrew Colman. (1997). (2006). “Gender Development.” In W. Damon & R. Udry, Christopher. (1996). “Gender, Agriculture “Explaining the Gender Gap in Math Test Scores: “Overconfidence: Feedback and item difficulty M. Lerner (Eds.), Handbook of Child Psychology: Production, and the Theory of the Household.” The Role of Competition.” Journal of Economic effects.” Personality and Individual Differences, 23, Social, emotional, and personality development, Vol. Journal of Political Economy 104 (5). Perspectives, 24(2), pages 129-144. pages 125-133. 3, pages 858–932. United Nations. (2017). “Leave No One Behind: Niederle, Muriel and Lise Vesterlund. (2011). Raj, Anita, Mohan Ghule. Julie Ritter, Battala Rudman, Laurie. (1998). “Self-Promotion as a Taking Action For Transformational Change On Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa “Gender and Competition.” Annual Review of Madhusudan, Velhal Gajanan, Saritha Nair, Anindita Risk Factor for Women: The Costs and Benefits of Women’s Economic Empowerment.” United Economics, 3(1), pages 601-630. Dasgupta, Jay Silverman, Donta Balaiah, and Counter stereotypical Impression Management.” Nations. Niranjan Saggurti. (2016). “Cluster Randomized Journal of Personality and Social Psychology, 74 (3), Nix, Emily, Elisa Gamberoni, and Rachel Heath. Controlled Trial Evaluation of a Gender Equity and pages 629-645. Uzzi, Brian. (1996). “The Sources and Consequences (2015). Bridging the Gender Gap: Identifying What Family Planning Intervention for Married Men and of Embeddedness for the Economic Performance Is Holding Self-employed Women Back in Ghana, Couples in Rural India.” PLoS One. 11 (5): e0153190. Rudman, Laurie and Peter Glick. (1999). “Feminized of Organizations: The Network Effect.” American Rwanda, Tanzania, and the Republic of Congo.” The Management and Backlash Toward Agentic Women: Sociological Review 61, pages 674–698. World Bank Economic Review, 30(3), pages 501{521. Rankin, Katherine. (2001). “Social capital, Hidden Costs to Women of a Kinder, Gentler Image microfinance, and the politics of development.” of Middle Managers.” Journal of Personality and Uzzi, Brian. (1997). “Social structure and Nordman, Christophe Jalil and Julia Vaillant. Feminist Economics, 8 (1), pages 1-24. Social Psychology, 77 (5), pages 1004-1010. competition in interfirm networks: The paradox of (2014). “Inputs, Gender Roles or Sharing Norms? embeddedness.” Administrative Science Quarterly, Assessing the Gender Performance Gap Among Rauch, Andreas and Michael Frese. (2007). Ruiz Abril, Maria Elena. (2008). “Liberia: Girls 42, pages 35–67. Informal Entrepreneurs in Madagascar.” IZA “Let’s put the person back into entrepreneurship vulnerability assessment.” Government of Liberia, Discussion Papers 8046, Institute for the Study of research: A meta-analysis on the relationship Nike Foundation, and World Bank. Valdivia, Martin. (2015). “Business training plus Labor (IZA). between business owners’ personality traits, for female entrepreneurship? Short and medium business creation, and success.” European Journal Shattuck, Dominick, Brad Kerner, Kate Gilles, term experimental evidence from Peru.” Journal of O’Sullivan, Michael, Arathi Rao, Raka Banerjee, of Work and Organizational Psychology, 16 (4), pages Miriam Harmann, Thokozani Ng’ombe, and Greg Development Economics. 113, pages 33-51. Kajal Gulati, and Margaux Vinez. (2014). “Levelling 353-385. Guest. (2011). “Encouraging Contraceptive Uptake Appendix & Endnotes the field: improving opportunities for women by Motivating Men to Communicate About Family Valerio, Alexandria, Brent Parton, and Alicia Robb. farmers in Africa.” World Bank Group / One Renzulli, Linda A., Howard Aldrich, and James Planning: The Malawi Male Motivator Project.” (2014). “Entrepreneurship Education and Training Campaign. Moody. (2000). “Family Matters: Gender, Networks, American Journal of Public Health 2011. 101(6), Programs around the World: Dimensions for and Entrepreneurial Outcomes.” Social Forces 79.2. pages 1089-1095. Success.” World Bank. O’Sullivan, Michael. (2017). “Gender and Property Rights in Sub-Saharan Africa: A Review of Ribero, Rocio and Fabio Sánchez. (2004). Schaner, Simone. (2015). “Do Opposites Detract? Van der Sluis, Justin, Mirjam van Praag, and Wim “Determinantes, Efectos y Costos de la Violencia Intrahousehold Preference Heterogeneity and Vijverberg. (2008). “Education and Entrepreneurship 194 195 Endnotes Selection and Performance: A Review of the Literature.” Journal of Economic Surveys 22 (5), pages 795–841. Vyas, Seema and Charlotte Watts. (2009). “How does economic empowerment affect women’s risk of intimate partner violence in low and middle income countries? A systematic review of published evidence.” Journal of International Development. 21, pages 577–602. Wang, Qing. (2015). “Child care, work-family policy 1. The Global Entrepreneurship and Development Institute - GEDI women. (2018). 38. See Ashraf et al. (forthcoming) on the importance of large scale and female entrepreneurship.” Social Science 2. Global Entrepreneurship Research Association - GERA (2018). institutional interventions to overcome biases on women’s Research Network. 3. United Nations (2017). education participation. 4. This is based on the authors calculations using the World Bank 39. See Deininger, Goyal, and Nagarajan (2013), and Roy (2015) on the Group’s Enterprise Surveys. importance of changes in property rights on investment in girls’ Wolf, Kathrin and Michael Frese. (2018). “Why 5. The following datasets are of micro-enterprises: Benin, Ghana education. husbands matter: Review of spousal influence on (1), Malawi, Togo, Uganda (1) and Uganda (2). The following 40. See Munshi and Rosenzweig (2006), Field, Jayachandran and women entrepreneurship in sub-Saharan Africa.” datasets are of SMEs: Mozambique, Nigeria (1), Nigeria (2), and Pande (2010), Luke and Munshi (2011) on the importance of South Africa. DRC is based on a census of firms. Ghana (2) is a solutions to women’s mobility constraints in South Asia. The Africa Journal of Management, vol. 4, pages 1-32. survey of firms in tailoring. Ethiopia and Ghana (3) are firms in evidence on these issues in Africa is more limited. manufacturing. 41. See Croke, Goldstein and Holla (2017) about sector shifting. World Bank - Independent Evaluation Group. (2011). 6. See for example, Hallward-Driemeier (2013). Women with traditional views regarding occupational segregation 7. The Global Entrepreneurship and Development Institute - GEDI see larger improvements. “Evidence and Lessons Learned (2018). 42. See Doyle et al. (2018) about a 15-week gender-transformative from Impact Evaluations on Social Safety Nets.” 8. Chakravarty, Das, and Vaillant (2017). group education program that engages new and expectant fathers 9. Hardy and Kagy (2018b). and their partners. Analysis of data 21 months post-baseline 10. Bardasi, Sabarwal, and Terrell (2011); Campos et al. (2015); Alibhai, shows significant differences between the intervention and control World Bank. (2012). “World Development Report: Buehren, and Papineni (2015). groups across a range of health and development outcomes, Gender Equality and Development.” Washington, 11. Evidence from other regions demonstrates that business sector including women’s reports of physical and sexual intimate partner sex segregation is not limited to Africa. In Sri Lanka, investment violence, women’s and men’s reports of harsh discipline against DC: World Bank. rates and returns to investment are lower in sectors characterized children, contraceptive use and antenatal care attendance, men’s by a higher share of female entrepreneurs (De Mel, McKenzie and dominance in decision-making, and men’s participation in and World Bank - WLSME. (2013). “Female Woodruff, 2009). In the US, women’s concentration in the personal time spent on domestic and caregiving tasks. services sector explains as much as 14% of the gender-based 43. Campos et al. (2017). Entrepreneurship: Program Guidelines and Case self-employment earnings differential (Hundley, 2001). 44. Alibhai, Buehren, and Papineni (2016). Studies.” 12. McKenzie and Woodruff (2017). 45. Batista and Seither (2018). 13. Benhassine et al. (2018). 46. McKenzie and Puerto (2017). An impact evaluation of the same 14. Campos, Goldstein, and McKenzie (2018). program in Vietnam also found positive impacts on firm growth World Bank. (2014). “Doing Business in Nigeria 2014: 15. Paryavi, Campos and Santos (2018). (Bulte, Lensink, and Vu, 2017). Understanding regulations for Small- and Medium- 16. Some surveys from the report’s impact evaluation database stand 47. Dupas and Robinson (2013). However, it is not possible in this Sized Enterprises.” as exceptions. For example, in Malawi, female entrepreneurs are experiment to distinguish between the effects of gender and the 38% more likely than male business owners to agree with the effects of sector of operation on the impact of the intervention. following statement: “If paid employment was offered to me at 48. Dupas et al. (2018). World Bank - Infodev. (2016). “Growth roughly the same level that I take home in this business, I would 49. Campos, Goldstein and McKenzie (2015). take such employment and close the business”. 50. Campos, Goldstein and McKenzie (2018). Entrepreneurship in Developing Countries: A 17. World Bank Group (2018). 51. Aker et al. (2016); Field et al. (2016b). Preliminary Literature Review.” Washington, DC: 18. This is as measured by the World Bank Group’s Women, Business 52. Jack and Suri (2016). The World Bank and the Law. These countries are South Africa, Zimbabwe, and 53. Bastian et al. (2018a). Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa Rwanda. 54. McKenzie (2017). 19. Cech et al. (2011). 55. Fafchamps and Quinn (2016). World Bank. (2018). “The Little Data Book on 20. Correll (2001; 2004). 56. United Nations (2017). Financial Inclusion 2018.” World Bank, Washington, 21. Blair-Loy (2003); Ridgeway and Correll (2004). 57. The Global Entrepreneurship and Development Institute - GEDI 22. Rudman (1998); Rudman and Glick (1999). (2018). DC. 23. Amanatullah and Morris (2010); Bowles (2012). 58. Global Entrepreneurship Research Association - GERA (2018). 24. Fawole, Ajuwon, and Osungbade (2005); Morrison and Orlando 59. Kelly and Firestone (2016). World Bank Group. (2018). “Women, Business, and (2004). 60. Bardasi, Sabarwal and Terrell (2011); Hallward-Driemeier (2013); 25. Ruiz Abril (2008). Nordman and Vaillant, 2014; Brixiová and Kangoye (2015); Nix, the Law 2018.” Washington, DC: World 26. World Bank (2012). Gamberoni, and Heath (2015); McKenzie and Woodruff (2017); Bank. License: Creative Commons Attribution CC BY 27. Beyer (1990); Pulford and Colman (1997); Soll and Klayman (2004). Hardy and Kagy (2018a). 3.0 IGO. 28. Niederle and Vesterlund (2007). 61. The following datasets contain survey information on micro- 29. Grosse and Riener (2010); Kamas and Preston (2010); Lundeberg, enterprises: Benin, Ghana (1), Malawi, Togo, Uganda (1) and Fox, and Punćcohaŕ (1994). Uganda (2). The following datasets contain information on SMEs: Wozniak, David, William Harbaugh, and Ulrich Mayr. 30. https://globalfindex.worldbank.org/ Mozambique, Nigeria (1), Nigeria (2), and South Africa. DRC data 31. Loscocco et al. (2009); Kim and Sherraden (2014); Renzulli, is based on a census of firms. Ghana (2) data is a survey of firms (2014). “The Menstrual Cycle and Performance Aldrich, and Moody (2000); Klyver and Terjesen (2007); Rankin in the tailoring sector. Ethiopia and Ghana (3) datasets contain Feedback Alter Gender Differences in Competitive (2001); Fafchamps and Minten (1999); Granovetter (1973, 1983). information on manufacturing firms. Choices.” Journal of Labor Economics 32 (1), pages 32. Udry (1996). 62. Throughout the report, the words enterprise, firm, company, and 33. World Bank (2012). business are used interchangeably. 161–198. 34. The importance of household responsibilities is in line with the 63. This is based on the authors’ calculations using the Impact findings in other regions. For instance, in Guatemala (Kevane Evaluation Surveys. Zhao, Hao and Scott Seibert. (2006). “The big five and Wydick, 2001) and India (Kantor, 2005), women’s sector 64. This is based on the authors’ calculations using the WBG concentration responds to the time available for home production, Enterprise Surveys. personality dimensions and entrepreneurial status: Appendix & Endnotes and constraints on location of business. 65. This is based on the authors’ calculations using the Impact A meta-analytical review.” Journal of Applied 35. Clark et al. (2017); Tabbert (2009); Barros et al. (2011); Berlinski Evaluation Surveys. Psychology, 91 (2), pages 259-271. and Galiani (2007). 66. Value-added is obtained by subtracting direct costs (costs of 36. The analysis draws on two surveys conducted specifically for this inputs, raw materials, and goods purchased for resale) from firm report: Togo (December 2016) and Malawi (February 2017). The sales. questionnaires used in these surveys are very similar, so results 67. This is based on the authors’ calculations using the Impact are comparable. It also uses other impact-evaluation surveys Evaluation Surveys. from Malawi and Ghana. Most of the questions that relate to 68. Figures exclude 0 and negative profits. social norms in the impact-evaluation surveys are different from 69. All data using Impact Evaluation Surveys. Similar patterns were those asked in the Togo and Malawi surveys for this report. obtained using Household Surveys. 37. Gender-biased norms are used in this context as those against 70. Chakravarty, Das, and Vaillant (2017). 196 197 71. Chakravarty, Das, and Vaillant (2017). is no information on tenure, the report uses the firm’s age as a 138. Benhassine et al. (2018). 185. Morris et al. (2006). 72. Hardy and Kagy (2018b). proxy. 139. Campos, Goldstein and Mckenzie (2018). 186. Arias, Evans and Santos (2017). 73. Emran et al. (2011). 105. When excluding Ghana manufacturing census (3), which does 140. Datasets in Latin American countries do not have information on 187. Heckman, Stixrud and Urzua (2006). 74. Chakravarty, Das, and Vaillant (2017). not control for any as many variables as the other datasets, share of ownership by gender. 188. Müller, Volery and Siemens (2012); Rauch and Frese (2007); Zhao 75. Bardasi, Sabarwal and Terrell (2011). the average gender gap is 21% when accounting for firm and 141. Gneezy, Leonard, and List (2009). and Seibert (2006). 76. Fafchamps et al. (2014); Nordman and Vaillant (2014). household differences. 142. Paryavi, Campos and Santos (2018). 189. Crant (1995); Frese (2009); Frese et al. (2007); Koop, De Reu, and 77. Niederle and Vesterlund (2007, 2008); Balafoutas and Sutter 106. This survey has less detail than the other ones and analysis only 143. Niederle and Vesterlund (2007, 2008). Frese (2000); Campos et al. (2017). (2010); Healy and Pate (2011); Niederle and Vesterlund (2010); controls for capital and labor. See Appendix 2 for a table with 144. Balafoutas and Sutter (2010); Healy and Pate (2011); Niederle and 190. The difference between Mozambique and Togo could be related Sutter and Rützler (2010); Wozniak, Harbaugh and Mayr (2014). surveys used and modules available in each. Vesterlund (2010); Sutter and Rützler (2010); Wozniak, Harbaugh to the firms in Mozambique being larger than those in Togo. 78. Bossuroy et al. (2013). 107. The part of the gap that can be explained using this statistical and Mayr (2014). Female entrepreneurs in the study in Mozambique may be already 79. McKenzie and Woodruff (2017). approach varies substantially by dataset, going from over 100% 145. Niederle and Vesterlund (2011). a selected sample, not representative of the population of those 80. OECD (2016). explained in Benin and one of the datasets in Uganda, down 146. Niederle and Vesterlund (2007). self-employed. 81. Kabeer (1999). to zero for one of the datasets in Ghana. If one excludes Togo 147. Ruble, Martin, and Berenbaum (2006). 191. Beyer (1990); Pulford and Colman (1997); Soll and Klayman (2004). 82. Donald et al. (2017). where the gap is very small to start with, the Oaxaca-Blinder 148. Gneezy, Leonard, and List (2009). 192. Niederle and Vesterlund (2007). 83. The countries included are Benin, the Democratic Republic of decomposition can account for about 50% of the gender gap in 149. Women may be even less likely to compete when representing 193. Grosse and Riener (2010); Kamas and Preston (2010); Lundeberg, Congo, Ethiopia, Ghana, Kenya, Malawi, Mozambique, Nigeria, business performance. the interests of others, like their employees. Beckmann and Fox, and Punćcohaŕ (1994). South Africa, Togo, and Uganda. 108. Hardy and Kagy (2018a) show that in the garment industry in Menkhoff (2008) found female mutual-fund managers to be more 194. Analysis for this report using data from Ghana Impact Evaluation. 84. For example, for Benin, the survey asks: “In the last month, what Ghana, male-owned firms earn nearly twice as much profit as averse to competition compared to their male colleagues. Paryavi 195. Kamas and Preston (2010); Grosse and Riener (2010). was the profit generated by your business? It is the difference female-owned firms and the gap cannot be explained by a set of (2016) found that women were significantly less likely to enter 196. Paryavi (2016). between all the income of your company and all the costs firm and owner characteristics. into competition when charged with representing the financial 197. Croson and Gneezy (2009). (salaries of employees, materials, taxes, rents...)”. 109. This finding assumes that in Ghana and Uganda differences in the interests of someone else than if they were to only represent 198. Paryavi, Campos and Santos (2018). 85. The Democratic Republic of Congo, Ethiopia, Ghana Census, and levels of capital investment are not contributing to the gender gap themselves. 199. Berge, Juniwaty, and Sekei (2016). In the study, willingness to take Nigeria GEMS. in profits. In Ghana and Uganda, the report has multiple studies 150. An additional lab experiment in Uganda (Buehren, Goldstein risks as a group is captured by letting groups decide jointly, face 86. As found in the Democratic Republic of Congo, Nigeria, and Togo. and these do not get to the same conclusion. Otherwise this and Montalvão, 2016) found no gender differences in competition to face, whether to invest in a risky asset or not. 87. This is from the Impact Evaluation surveys. Data on education statistic would be 70%. amongst adolescent boys and girls, but the experimental design 200. Aterido, Beck, and Iacovone (2011). levels excludes impact evaluations in Nigeria, which targeted 110. Mckenzie (2017). had a couple of important constraints: participants could not 201. https://globalfindex.worldbank.org/ a population where 80 to 90% of the sample received higher 111. Fafchamps et al. (2014) see who else they were competing with, and the task chosen for 202. Mayoux (1999); Demirguc-Kunt, Beck and Honohan (2008). education. 112. Banerjee, Karlan, and Zinman (2015). the competition experiment was seen as gender neutral (sorting 203. Demirguc-Kunt, Beck and Honohan (2008). 88. In Ghana, firm-level profits of male entrepreneurs represent 45% 113. Fafchamps et al. (2014). building blocks), whereas the literature has found that women are 204. It is important to take into account that most datasets have only a of their household expenditures. 114. Banerjee, Karlan, and Zinman (2015). less competitive in stereotypically male tasks such as math. few of the metrics on conditions of loans. 89. Data from the Impact Evaluation surveys show that female- 115. Bernhardt et al. (2017). 151. Some surveys from the report’s impact evaluation database stand 205. In Malawi, the positive difference (women more likely to need to owned businesses have been in operation on average for 8 years 116. Within individual sectors identified in each dataset, there may as exceptions. For example, in Malawi, female entrepreneurs are present collateral) disappears after controlling for the size of the and male businesses have been in operation for 9 years. The be further sex-segregation with women mostly selling to other 38% more likely than male business owners to agree with the loan. In Benin, there is no information on size of loan and the female-owned (male-owned) firm’s age rises to 11 (13) years when women and men to other men. These differences may not be following statement: “If paid employment was offered to me at unconditional difference is negative. In the other datasets, there is using the Enterprise Surveys for the informal sector. Among the reflected in the sector composition, limiting the depth of the roughly the same level that I take home in this business, I would no information on the need for presenting collateral. group of formal large firms covered in the Enterprise Surveys, analysis on the importance of sector choice. take such employment and close the business”. 206. https://globalfindex.worldbank.org/ female-owned businesses have been in operation around 20 years 117. Nordman and Vaillant (2014). 152. Paryavi, Campos and Santos (2018) 207. Africa is the leader region in the use of mobile money accounts and male-owned businesses almost 22 years, which is relatively 118. Fafchamps et al. (2014). 153. Berge et al. (2015). where the 21% of adults have a mobile money account. World similar to other regions for firms of this size. 119. McKenzie and Woodruff (2017). 154. Similarly, a large-scale business plan competition in Nigeria had Bank (2018). 90. Data from the Impact Evaluation surveys. It is 24 days for 120. Bardasi, Sabarwal, and Terrell (2011); Costa and Rijkers (2012); 24,000 entrants in round one, of which only 18% were women. 208. Campos, Goldstein and McKenzie (2015). male-owned firms. These figures may be over-representing Campos et al. (2015). In response, the second year was restricted only to women, 209. Kes, Jacobs and Namy (2011). entrepreneurs with full-time businesses, as impact evaluations 121. Bardasi, Sabarwal, and Terrell (2011). and it received 65,000 entrants. This in part reflects increasing 210. https://globalfindex.worldbank.org/ often select participants who have an interest in doing something 122. Alibhai, Buehren, and Papineni (2015). knowledge of the program (the program achieved more than 211. https://globalfindex.worldbank.org/ for their business, which may exclude those who run part-time 123. Campos et al. (2015). 100,000 applicants in year three when it went back to both 212. Granovetter (1973); Bertrand, Luttmer and Mullainathan (2000). businesses. 124. Evidence from other regions demonstrates that gender business genders), but provides an example of how design can support 213. Davidsson and Honig (2003); Audretsch and Feldman (2004); 91. Estimate based on median value from Impact Evaluation surveys. sector segregation is not limited to Africa. In Sri Lanka, participation in competition. Renzulli, Aldrich, and Moody (2000). On average, female-owned firms in the datasets analyzed have investment rates and returns to investment are lower in sectors 155. World Bank Group (2018). 214. Qualitative work in Ethiopia suggests that large family and friend five people working in their business and male-owned firms have characterized by a higher share of female entrepreneurs (De Mel, 156. A similar analysis for Europe indicates that 20 out of 39 countries networks are important for having enough collateral for large nine. McKenzie and Woodruff, 2009). In the US, women’s concentration fully comply with the 22 items of formal law that treat women and loans. (Pierotti, 2016) 92. Data from the Impact Evaluation surveys shows that 41% of in the personal services sector explains as much as 14% of the men equally. 34 out of 39 countries (87%) in Europe comply with at 215. Uzzi (1997); McMillan and Woodruff (1999); Fafchamps (2001). female-owned businesses have electricity, while this proportion gender-based self-employment earnings differential (Hundley, least 21 items of formal law that treat women and men equally. 216. Field et al. (2016a). is 56% for male-owned firms. These figures increase to 53% and 2001). 157. World Bank Group (2018). 217. Brüderl and Preisendörfer (1997); Uzzi (1996). 62% respectively in Enterprise Surveys of informal firms. Female 125. As for the profit gap, in this analysis the following datasets are of 158. World Bank Group (2018). 218. Uzzi (1997). and male-owned large firms experienced on average 8 and 7% micro-enterprises: Benin, Ghana (1), Malawi, Togo, and Uganda 159. Platteau (2000); Bicchieri (2006). 219. This reports’ analysis based on impact evaluation data. Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa respectively of losses due to electricity cuts. (1). The following datasets are of SMEs: Mozambique, Nigeria (1), 160. Marcus and Harper (2014). 220. This reports’ analysis based on impact evaluation data. 93. 16% in Impact Evaluation surveys, compared to 23% in Enterprise Nigeria (2), and South Africa. DRC is based on a census of firms. 161. Rudman (1998); Rudman and Glick (1999). 221. Alibhai, Buehren, and Papineni (2015); Campos et al. (2015). Surveys. Ghana (2) is a survey of firms in tailoring. Ethiopia and Ghana (3) 162. Amanatullah and Morris (2010); Bowles (2012). 222. Cai and Szeidl (2018). 94. According to the median value of capital stock from the Impact are firms in manufacturing. 163. Cech et al. (2011). 223. Loscocco et al. (2009); Kim and Sherraden (2014); Renzulli, Evaluation surveys. 126. As mentioned above, the “core specification” includes firm 164. Friedson-Ridenour and Pierotti (2018). Aldrich, and Moody (2000); Klyver and Terjesen (2007); Rankin 95. Data from the Impact Evaluation surveys. These percentages characteristics (number of workers, capital stock, business 165. Correll (2001, 2004). (2001); Fafchamps and Minten (1999). decrease to 34% and 37% in Enterprise Surveys, and 34% and practices, formalization, innovation, financial services, firm has 166. This report’s analysis in Malawi and Togo suggests that a large 224. Granovetter (1973, 1983). 35% in Enterprise Surveys of informal firms. a loan, firm has at least one additional worker, and sector) and share of entrepreneurs believe men are better managers than 225. Field et al. (2016a). 96. This large difference across countries on having borrowed may owner characteristics (age, tenure, marital status, household size, women. 226. World Bank (2012). be due also to measurement issues as measurement work in number of children per adult, and education). 167. Blair-Loy (2003); Ridgeway and Correll (2004). 227. Beaman, Keleher and Magruder (2017). multiple countries has shown that the way the question is asked 127. Impact evaluations include often businesses that operate on a 168. Croke, Goldstein, and Holla (2017). 228. In an environment with weak contract enforcement, networks (allowing for listing all sources of credit including household and full-time basis, so these metrics may be excluding the dynamics 169. Alibhai, Buehren, and Papineni (2015). can be instrumental to facilitate credible commitments (Greif, community level versus questions just focused on formal loans among those operating businesses on a part-time basis. 170. Campos et al. (2015). 1997; McMillan, 1997; Fafchamps, 2001). Networks may play a from commercial banks) can lead to different statistics (World 128. This finding is also compatible with men hiring more workers 171. Fawole, Ajuwon, and Osungbade (2005). particularly important role in reducing transaction costs, by Bank - WLSME, 2013). The conclusions are mostly driven by until the returns to labor are the same across gender. 172. Morrison and Orlando (2004). fostering trust, information sharing, and supporting alternate gender comparisons within countries using the same questions 129. Similarly to capital investment, this analysis controls for firm 173. Ruiz Abril (2008). enforcement mechanisms. and focusing on the typical and average results. characteristics (number of workers, capital stock, business 174. Analysis for this report using Malawi and Uganda Impact 229. In Ghana, entrepreneurs were randomly incentivized to participate 97. Given the significant differences across countries analyzed in the practices, formalization, innovation, financial services, firm has Evaluation data. in a joint activity requiring daily collaboration (Fafchamps and amount of outstanding loans, which would affect the averages a loan, firm has at least one additional worker, and sector) and 175. Analysis for this report using Tanzania Impact Evaluation data. Quinn, 2013). This intervention— seen as being given a business (showing a larger gender gap) but not the conclusion of the owner characteristics (age, tenure, marital status, household size, 176. While this difference between own and others may reflect network— had a positive effect on business performance, and importance of this issue, the figures presented in this table for number of children per adult, and education). underreporting, it can also be due to framing and anchoring, as led to the diffusion of business practices among entrepreneurs loans outstanding only includes the countries where the amounts 130. Mckenzie and Woodruff (2017). no one would answer with non-integral figures out of 10 people assigned to the same team. Researchers found women were have lower variance: DRC, Malawi, Togo, Uganda. It is hence more 131. Cirera and Maloney (2017). (like 0.5 out of 10), and a low number out of 10 may end up by less likely than men to have interacted with members of their conservative than if using all countries in datasets. 132. Cirera and Maloney (2017). “higher” than out of 100. assigned team. An experiment in Ethiopia, Zambia and Tanzania Appendix & Endnotes 98. The survey conducted in Kenya includes only women, so it is 133. 16% of female-owned businesses in Kenya are planning to 177. Campbell (2002). found a positive effect on diffusion of business practices for not possible to compute gender gaps. As a result, this survey is introduce a new product and 3% are planning to improve existing 178. Ribero and Sánchez (2004). the managers of small firms assigned to collaborate with large excluded from the analysis presented in this section. products in the next 2 years. 179. Doyle et al. (2018). firms (Fafchamps and Quinn, 2016). The study, however, did not 99. Measured in hours of work of all people working in the business. 134. This difference is large in magnitude and percentage terms (40%), 180. Vyas and Watts (2009). find evidence of diffusion of business practices among average 100. This includes inventory plus equipment and property/land value. but not statistically significant. 181. World Bank (2012). participants. 101. Using an index of business practices within data availability as per 135. Results available upon request. 182. This may be due to social norms or to lower labor market 230. Udry (1996). McKenzie and Woodruff (2017). 136. Identifying differences in innovation in impact evaluation datasets returns to women’s education, which may itself be the result 231. Nordman and Vaillant (2014). 102. Firm is registered and pays local/national taxes and fees. could be specific to the group of firms analyzed. More data needs of discrimination in hiring or sex segregation of the workforce. 232. De Mel, McKenzie and Woodruff (2009). 103. Firm has a bank account and separates business from household to be collected throughout the region on firm-level innovation to These issues may limit women to lower paying jobs and sectors. 233. Pierotti (2016). money. take more appropriate conclusions. 183. Van der Sluis, Van Praag and Vijverberg (2008). 234. Schaner (2015). 104. Tenure is measured in years managing business. When there 137. Data on formal firms from World Bank Enterprise Surveys. 184. Hallward-Driemeier (2013). 235. Fafchamps et al. (2014). However, a study from urban Madagascar 198 199 shows low within-household differences in returns to capital 269. This uses the approximation “exp(ß )-1” for interpreting the inverse 308. Hallward-Driemeier, Hasan, and Rusu (2013). 362. Fafchamps and Quinn (2016). in households operating informal businesses, which suggests hyperbolic sine-dummy relationship as per Bellemare and 309. This study uses a difference-in-differences approach taking 363. Brooks, Donovan, and Johnson (forthcoming). an unbiased allocation of capital among household members Wichman (2018). advantage of the sequencing of the reform rollout to estimate 364. Among those who buy inventory at least once a week, the study (Nordman and Vaillant, 2014). 270. Renzulli, Aldrich, and Moody (2000). impact. found a substantial benefit from mentorship, but none from the 236. Friedson-Ridenour and Pierotti (2018). 271. This is based on the 2011 Malawi survey. 310. Hallward-Driemeier and Gajigo (2013). class treatment. 237. Bernhardt et al. (2017). 272. This is based on the 2013 Togo survey. 311. Harari (2018). 365. Bastian et al. (2018b). 238. Duflo (2003); Duflo and Udry (2004). 273. Bossuroy et al. (2013). 312. World Bank (2014). 366. The mentors had to appoint mentees before their own 239. Fafchamps et al. (2014). 274. Fafchamps and Quinn (2013). 313. World Bank Group (2018). entrepreneurship training program. Mentors were then randomly 240. Nordman and Vaillant (2014). 275. Fafchamps and Quinn (2016). 314. Hallward-Driemeier (2013). assigned to either participate in the mentee support intervention 241. Fiala (2015). 276. This is based on the 2011 Malawi dataset. 315. Rodgers and Menon (2012). or not, allowing for comparing the impacts of the intervention 242. World Bank (2012). 277. Brüder and Preisendörfer (1997) explore this possibility but do not 316. FAO (2017). across mentees in both groups. 243. World Bank (2012). reach any empirical conclusion. 317. O’Sullivan (2017). 367. Bastian et al. (2018b). 244. The importance of household responsibilities is in line with the 278. Fiala (2015). 318. O’Sullivan (2017). 368. McKenzie and Puerto, 2016. findings in other regions. For instance, in Guatemala (Kevane 279. In Kenya, firm-level profits of female entrepreneurs represent on 319. Ali, Deininger, and Goldstein (2014). 369. Valdivia (2015); Giné and Mansuri (2014). and Wydick, 2001) and India (Kantor, 2005), women’s sector average 65% of their household income. In Malawi, 38% of money 320. Ali et al. (2014). 370. Field et al. (2016a). concentration responds to the time available for home production, earned in business goes towards daily household expenses. 321. Cherchi et al. (2018). 371. Bernhardt et al. (2017). and constraints on location of business. 280. Wolf and Frese (2018). 322. Goldstein et al. (2016). 372. Buvinic and Furst-Nichols (2014). 245. Donald et al. (2018). 281. Schaner (2015). 323. Agyei-Holmes et al. (2018). 373. Microcredit Summit Campaign (2015). 246. Blackden and Wodon (2006). 282. McKenzie and Paffhausen (2017). 324. Ali et al. (2014). 374. Tarozzi, Desai, and Johnson (2015). 247. Martinez, Naudeau and Pereira (2012). 283. The median chop money is about 25% higher than the median 325. Bruhn and McKenzie (2013). 375. Banerjee, Karlan, and Zinman (2015). 248. Clark et al. (2017). monthly profits. Both the distribution of chop money and of 326. Cost-free assistance included visiting business owners and offer 376. Dahal and Fiala (2018). 249. Tabbert (2009). profits have high variance, requiring some caution in this type of assistance in registering their businesses, while conveying to 377. Dahal and Fiala (2018). 250. see Lokshin, Glinskaya and Garcia (2004) for Kenya. comparison. them a single-page information flyer on the potential benefits 378. Meager (2016). 251. Barros et al. (2011). 284. Fafchamps et al. (2014). offered by registration. For those who were interested, the team 379. Patel (2014); Duvendack et al. (2011) 252. Berlinski and Galiani (2007). 285. Nordman and Vaillant (2014); Fiala (2015). assisted them in filling out the business registration form, 380. A recent study from India found more promising benefits on 253. The analysis draws on two surveys conducted specifically for this 286. Berge et al. (2012). took the required photo, and delivered their entire application, micro-enterprise performance (Breza and Kinnan, 2017). report: Togo (December 2016) and Malawi (February 2017). The 287. While over 40% of husbands overestimate how much female including paying the registration fee on their behalf. Once ready – 381. Banerjee et al. (2015b). questionnaires used in these surveys are very similar, so results entrepreneurs make, which would suggest should disclose their on average certificates take two weeks to be prepared – the team 382. Cull and Morduch (2017). are well comparable. It also uses other impact-evaluation surveys income, there are few reasons otherwise: women do not know delivered the Business Registration Certificates back to these 383. Field et al. (2012). from Malawi and Ghana. Most of the questions that relate to that their husbands over-estimate, not all husbands over- firms. Thus, the only cost to these firms was the time it took to fill 384. Berge et al. (2012). social norms in the impact evaluation surveys are different from estimate, and even if husbands are over-estimating, women have out the registration form (where they were assisted by the team). 385. Fafchamps et al. (2014). those asked in the Togo and Malawi surveys for this report. been managing with their current income level, so disclosure 327. Campos, Goldstein and McKenzie (2018). 386. De Mel, McKenzie and Woodruff (2009). 254. Using WBG Enterprise Surveys. might not lead men to conclude that they need to increase their 328. Benhassine et al. (2016). 387. Fafchamps et al. (2014). 255. Gender-biased norms are used in this context as those against contributions. 329. The lower take-up of formalization in Benin due to the 388. This finding applies to small cash grants to small businesses. See women. 288. De Mel, McKenzie and Woodruff (2009). combination of tax and business registration means that Table 7 for the effect of grants in other contexts. 256. Bertrand, Kamenica, and Pan (2015). 289. Friedson-Ridenour and Pierotti (2018). there was no statistical power to detect impacts on business 389. Similar programs include World Bank financed government 257. Still, in Togo, women are more likely than men (47% versus 39%) 290. Angelucci and Garlick (2016); Hoel et al. (2018). performance. programs in Côte D’Ivoire, Guinea-Bissau, Kenya, and Senegal; to agree that they should work in sectors that are considered 291. The policies also include bundled interventions that seek to 330. Campos, Goldstein and McKenzie (2018). Dfid’s ENGINE competition in Ghana; the African Innovation Prize; suitable for them, while in Malawi women are less likely to hold address multiple constraints simultaneously and these are 331. http://www.enterprisesurveys.org/data/exploretopics/regulations- Enablis Entrepreneurial Network’s Business Plan Competition in such views than men (39% versus 49%). In Togo, the women’s included whenever impact evaluation evidences is available. and-taxes Ghana; the Darecha Business Ideas Competition in Tanzania; the view that they should work in suitable sectors seems to be 292. Chakravarty, Das, and Vaillant (2017). 332. Overseas Development Institute (2015). SEED Awards in Ethiopia, Kenya, Malawi, Morocco, Mozambique, driven by those in traditionally female-dominated sectors: In 293. While arguably the high number of subsistence-level micro- 333. Policy interventions that combine access to capital and assets Namibia, South Africa, Tanzania and Uganda; the StartUp Cup non-female-dominated sectors, 37% of women and 38% of men enterprises in Africa reflects a lack of wage employment with training and other support, in what is often called “bundled in Cameroon, Ghana, Kenya, Rwanda and Zambia. This list is believe women should work in suitable sectors, compared with opportunities, there is limited evidence of the relative impact of services” or “wrap-around services” are reviewed under Policy compiled mostly from Fafchamps and Quinn (2016). 50% of women (and 42% of men) in female-dominated sectors. programs to support micro-enterprise development compared Area 4. 390. McKenzie (2017). 258. This difference is statistically different at the 5% level. to wage employment. Assessing the impact of enterprise 334. McKenzie and Woodruff (2013); Patel (2014); Cho and Honorati 391. In the first round of the business plan competition in Nigeria, only 259. A similar set of vignettes on business vs. household priorities and development policies vis-à-vis those seeking to increase wage (2013). 18.5% of applicants were women. A second round targeting only decision-making was recently posed to female entrepreneurs in employment is beyond the scope of this report. 335. Cirera and Qasim (2014). female entrepreneurs was then conducted. Ghana. Women were effectively gender-blind in their responses 294. Adoho et al. (2014). 336. Robb, Valerio, and Parton (2014). 392. Fafchamps and Quinn (2016). to questions on spending priorities and on consulting the spouse. 295. Buvinic and O’Donnell (2016). 337. McKenzie and Woodruff (2013). 393. Fafchamps and Quinn (2016). This may be partly due to the fact that the stakes involved in the 296. Banerjee et al. (2015a). 338. An exception includes a study in Uganda, which also found no 394. Mckenzie and Sansone (2017). Ghana vignettes were higher than those in Togo. 297. Women represented 86% of participants. positive impacts for female or male entrepreneurs (Campos et al. 395. Dupas and Robinson (2013). However, it is not possible in this 260. In Togo, the maximum bias is 0.83, because the most gender- 298. Blattman et al. (2013). 2013). experiment to distinguish between the effects of gender and the biased individuals in this sample believe in 5 of the 6 gender-bias 299. World Bank - Independent Evaluation Group (2011). 339. McKenzie and Woodruff (2013). effects of sector of operation on the impact of the intervention. Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa questions. 300. Hagen-Zanker et al. (2016). 340. Cirera and Qasim (2014); Drexler (2014). 396. Campos et al. (2015). 261. Similar results are obtained when using the Malawi 2014 survey. 301. O’Sullivan et al. (2014). 341. Berge et al (2012). 397. Campos et al. (2018). 262. The small sample size and large variance of capital investment 302. International Finance Corporation (2014). 342. McKenzie and Woodruff (2017). 398. Schaner (2017). measures seem to be driving the non-significance, as the 303. African Development Bank’s African Women in Business 343. The other countries are Bangladesh, Chile, Ghana, Kenya, Mexico, 399. The study (Dupas et al., 2018.) also included Chile. difference on average is still quite large. (AWIB) Initiative https://www.afdb.org/fileadmin/uploads/afdb/ Nigeria, and Sri Lanka. 400. Aker et al. (2016). 263. The analysis obtains similar results for profits with the vignette Documents/Generic-Documents/d%C3%A9pliant%20AWIB%20 344. Buvinic and O’Donnell (2016). 401. Field et al. (2016b). responses from Togo. The more conservative views women have ENGL.pdf; Goldman Sach’s 10,000 Women Project https://www. 345. Valdivia (2015) suggests that there are higher drop-out rates in 402. Jack and Suri (2016). (particularly the more they believe that a female entrepreneur goldmansachs.com/citizenship/10000women/#overview; IFC longer training programs. 403. Bastian et al. (2018a) should consult with her husband before making important and Goldman Sachs Women Entrepreneurs Opportunity Fund 346. Berge et al. (2012). 404. International Telecommunications Union (2016). decisions / the less they believe that a female entrepreneur (WEOF) https://www.ifc.org/wps/wcm/connect/news_ext_content/ 347. Valerio, Parton, and Robb (2014); Glaub and Frese (2011). 405. Arraiz et al. (2017). should hire an employee to increase profits), the higher profits ifc_external_corporate_site/news+and+events/news/perspectives/ 348. Botha, Nieman and van Vuuren (2006). 406. Schaner (2015). they make compared to male entrepreneurs. perspectives-i1c6; Women Entrepreneurs Finance Initiative (We- 349. Campos et al. (2017). 407. Jakiela and Ozier (2016); Baland et al. (2016); Boltz, Marazyan and 264. See Ashraf et al. (forthcoming) on the importance of large scale fi) https://we-fi.org/; Affirmative Finance Action for Women in 350. Glaub et al. (2014). Villar (2017); Ashraf (2009); Castilla and Walker (2013). institutional interventions to overcome biases on women’s Africa (AFAWA) Cherie Blair Foundation for Women’s Mentoring 351. Gertler (2018). 408. Fiala (2017). The game was done two years after the treatments education participation. Women in Business http://www.cherieblairfoundation.org/ 352. Batista and Seither (2018). and could reflect their outcomes rather than prior conditions. 265. See Deininger, Goyal, and Nagarajan (2013) and Roy (2015) on the programmes/mentoring/; African Development Bank’s “50 million 353. “The training was provided by the Digital Opportunity Trust (DOT), 409. http://blogs.worldbank.org/impactevaluations/what-happens- importance of changes in property rights on investment in girls’ women speak” https://www.afdb.org/en/news-and-events/going- a social enterprise that provides entrepreneurship training to when-business-training-and-capital-programs-get-caught-web- education. digital-afdbs-50-million-women-speak-digital-platform-set-to- clients of the Women Entrepreneurship Development Project intrahousehold-dynamics. 266. See Munshi and Rosenzweig (2006), Field, Jayachandran and create-digital-value-that-underpins-africas-future-economic- (WEDP) in Mekelle, Ethiopia. The DOT ReachUp! training 410. Bernhardt et al. (2017). Pande (2010), and Luke and Munshi (2011) on the importance of transformation-17455/; Power Africa and USAID’s Women in program takes an innovative approach to entrepreneurship 411. A first attempt to evaluate matching grant programs in Africa has solutions to women’s mobility constraints in South Asia. The African Power https://www.usaid.gov/powerafrica/gender; development, through a 120-hour course that helps inexperienced proven unsuccessful (Campos et al., 2014). The study attempted evidence on these issues in Africa is more limited. Standard Chartered Women in Technology Incubator Kenya entrepreneurs to learn basic technology and business skills, and to implement randomized experiments to evaluate the impact of 267. See Croke, Goldstein and Holla (2017) about sector shifting. Program https://vc4a.com/ibizafrica/standard-chartered-women- to foster the self-esteem and entrepreneurial spirit needed to seven matching grant programs offered in six African countries, Women with traditional views regarding occupational segregation in-technology-incubator-kenya-3/; USAID Accelerating Women build sustainable livelihoods.” (Alibhai, Buehren, and Papineni, but in each case the experimental evaluation was not possible. Appendix & Endnotes see larger improvements. Entrepreneurs Award https://www.opportunitiesforafricans. 2016) The study suggests three main proximate reasons for lack 268. See Doyle et al. (2018) about a 15-week gender-transformative com/usaid-accelerating-women-entrepreneurs-prize-2018/; 354. Alibhai, Buehren, and Papineni (2016). of implementation: continued project delays, politicians not group education program that engages new and expectant fathers CocaCola 5by20 initiative https://www.coca-colacompany.com/ 355. International Labor Organization (2018). willing to allow random assignment, and low program take-up; and their partners. Analysis of data 21 months post-baseline our-company/5by20-what-were-doing; Africa Women Innovation 356. McKenzie and Puerto (2017). An impact evaluation of the same and then delves into the underlying causes of these occurring: shows significant differences between the intervention and control and Entrepreneurship Forum http://awieforum.com/; program in Vietnam also found positive impact on firm growth political economy, overly stringent eligibility criteria, lack of groups across a range of health and development outcomes, 304. The World Bank Doing Business report identified 781 reforms in (Bulte, Lensink, and Vu, 2017). attention to detail in “last mile” issues, incentives facing project including women’s reports of physical and sexual intimate partner regulations governing business start-up, operations, property 357. Bastian et al. (2018b). implementation staff, and the way impact evaluations are funded. violence, women’s and men’s reports of harsh discipline against rights, and debt resolution in Africa between 2006 and 2017. 358. Mhede, Higuchi, and Sonobe (2018). 412. Grown et al. (2017). children, contraceptive use and antenatal care attendance, men’s 305. Simavi, Manuel, and Blackden (2010). 359. Campos et al. (2017). 413. Martinez, Naudeau and Pereira (2012); Barros et al. (2011); Tabbert dominance in decision-making, and men’s participation in and 306. Hallward-Driemeier (2013). 360. Campos et al. (2018). (2009); Berlinski and Galiani (2007); Clark et al. (2017). time spent on domestic and caregiving tasks. 307. Iqbal et al. (2016). 361. Fafchamps and Quinn (2013). 414. Clark et al. (2017). 200 201 415. Wang (2015). 416. Lokshin, Glinskaya and Garcia (2004). 417. Mateo Díaz and Rodriguez-Chamussy (2016). 418. Mateo Díaz and Rodriguez-Chamussy (2016). 419. Most interventions thus far have focused on changing behaviors, and not norm change. Substantial behavior change might eventually lead to norm change, but that is harder to measure and harder to change in individual experimental work. 420. Ricardo, Eads, and Barker (2011); Instituto Promundo and CARE International in Rwanda (2012); Instituto Promundo and International Center for Research on Women (2014); Shattuck et al. (2011); Raj et al. (2016). 421. Doyle et al. (2018). 422. https://www.poverty-action.org/study/making-cash-grants-work- small-scale-women-entrepreneurs-ghana 423. Campos et al. (2015); Alibhai, Buehren, and Papineni (2015). 424. World Bank Group (2018). 425. Campos et al. (2015); Alibhai, Buehren, and Papineni (2015). 426. Nordman and Vaillant (2014). 427. Campos et al. (2015). 428. Alibhai, Buehren, and Papineni (2015). 429. Hicks et al. (2011). 430. Campos et al. (2015). 431. Alibhai, Buehren, and Papineni (2015). 432. Croke, Goldstein and Holla (2017). 433. Gennari, Arango, and Hidalgo (2015). 434. Brenton et al. (2013). 435. Brenton et al. (2013). 436. Croke et al. (2018). 437. Brenton et al. (2013). 438. Cadot et al. (2015). 439. Atkin, Khandelwal, and Osman (2014). 440. Networking capital can be understood as the ability for a firm to draw on social and professional networks to gain access to information and business opportunities that can improve its productivity. 441. Brenton et al. (2013). 442. International Trade Centre (2015). 443. McKinsey Global Institute (2016). 444. Ferraz, Finan, and Szerman (2016). 445. International Trade Centre (2014). 446. South Africa is one example where government procurement rules include preference points and subcontracting requirements for contracts above a certain threshold to promote black economic empowerment, including female-owned micro and small businesses. 447. RSA National Treasury (2017). 448. Valerio, Parton, and Robb (2014). 449. Campos et al. (2015). 450. World Bank - Infodev (2016). 451. Banerjee et al. (2015b). 452. Hussam, Rigol, and Roth (2016). 453. McKenzie (2017). 454. Fafchamps and Woodruff (2016). Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 455. McKenzie (2017). 456. Bardasi et al. (2017). 457. Valerio, Parton, and Robb (2014). 458. McKenzie and Woodruff (2013). 459. Bardasi et al. (2017) 460. Anderson, Chandy, and Zia (2016). 461. Berge et al. (2012). 462. Paryavi, Campos and Santos (2018). 463. Paryavi (2016). 464. Diwan et al. (2014). 465. One example of a business training program delivered via SMS is the Business Women service developed by the Cherie Blair Foundation, in partnership with Exxon Mobile Foundation and Nokia, which reached over 100,000 women in Indonesia, Nigeria, and Tanzania. Appendix & Endnotes 202 203 Profiting from Parity | Unlocking the Potential of Women’s Businesses in Africa 205