92210 Public Disclosure Authorized Innovation, Technology & Entrepreneurship Policy Note September 2014  •  Number 5 Supporting Growth-Oriented Women Entrepreneurs: Public Disclosure Authorized A Review of the Evidence and Key Challenges ABSTRACT In recent years, support programs for women entrepreneurs have gained traction and prominence as a means to create jobs and boost productivity at the national and regional levels. However, disparities in initial resource endowments of male—and female-led firms, sector sorting into low productivity activities, social norms, and institutional arrangements, constrain Public Disclosure Authorized the growth of female-led enterprises. This note reviews the outcomes of programs supporting female growth entrepreneurs and draws lessons from available evidence to inform the design of more effective programs. The review shows that most programs are primarily geared toward microenterprises, making it difficult to draw conclusions about program design for growth-oriented entrepreneurs, but some early findings point the way forward. Management practices appear to improve as a result of business education, but there is little robust evidence to prove that support programs lead to significant improvements in business performance outcomes. Furthermore, in programs with both male and female participants, firm performance improves in some cases for male-led firms only, not for female-led firms. The note concludes by suggesting the need for more experimentation in the design and delivery of services and a Public Disclosure Authorized new focus on strengthening the engendering of support programs to more specifically address gender-specific constraints such as social norms, entrepreneurial preferences, and institutional arrangements, changing public discourse, and paying more attention to factors that induce female entrepreneurs to diversify into higher value-added activities. Offering mentoring, Xavier Cirera Innovation Technology and Entrepreneurship, Trade networking, and other consulting services, and Competitiveness Global Practice, the World Bank E-mail: xcirera@worldbank.org in addition to education on basic business practices and strengthening critical areas such Qursum Qasim Africa Region, Finance and Markets Global Practice, the as gender-specific content, can potentially World Bank E-mail: qqasim@worldbank.org increase the effectiveness of these programs. Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 1 1. INTRODUCTION are unable to grow their businesses from micro or Poverty reduction and shared prosperity can only small to medium or large productive enterprises be achieved with the full economic participation with transformational economic impact. Therefore, of both men and women. Yet almost one billion empowering female entrepreneurs, especially those in women have the potential to contribute more fully high-growth sectors, has the potential to create jobs, to their economies but are unable to do so. Of these increase incomes, lift millions out of poverty, and lead 812 million live in the developing world, where their to greater economic and social transformation. contributions, as workers and job creators, is greatest.1 Female entrepreneurial activity is concentrated in low- The last decade has seen a burgeoning of productivity sectors with limited potential for growth entrepreneurship support programs aimed at in income and employment and that often operate unleashing the potential of female entrepreneurs. informally.2 In many cases, female entrepreneurs Evidence on the impact of these programs is limited, and the few impact evaluations that have been conducted suggest that the impact of these programs on business growth outcomes is mixed at best. Thus, the question of how to effectively design support TABLE OF CONTENTS programs that facilitate female entrepreneurs to move 1. Introduction . . . . . . . . . . . . . . . . . . . . 2 into growth sectors with potential for job creation and productivity gains remains unresolved. 2. Entrepreneurship and the Gender Gap in Performance . . . . 3 This note reviews the empirical literature analyzing 2.1  Performance of Female-led the performance gaps between male and female Firms vs. Male-led Firms? . . . . . . . . . . . . . . 4 entrepreneurs and the impact evaluations of 2.2  What Explains the Performance Gap? . . . . . 6 programs that support female entrepreneurship. 3. Taking Stock: Lessons from Its aim is to enhance the effectiveness of these Existing Women Entrepreneurship programs by drawing lessons from current and past Support Programs . . . . . . . . . . . . . . . 8 support programs, identifying gaps in knowledge, and proposing areas of focus for program design 3.1 Selection . . . . . . . . . . . . . . . . . . . . . . . . . . 9 going forward. The note focuses on female growth 3.2  Business Education . . . . . . . . . . . . . . . . . . 10 entrepreneurs, that is, those with the potential to 3.3  Networking and Mentoring . . . . . . . . . . . 12 create new jobs and generate productivity gains3 3.4  Other Services . . . . . . . . . . . . . . . . . . . . . 13 rather than “necessity” entrepreneurs, who are 3.5  Access to Finance . . . . . . . . . . . . . . . . . . 14 unlikely to generate substantial growth in terms of job 3.6  Monitoring and Evaluation . . . . . . . . . . . 15 creation and broad economic impact.4 Female growth entrepreneurs are defined here as those who wish to 4. Looking Forward: Areas of Focus grow their firms—not only “high-growth” firms, or for Female Entrepreneurship gazelles, but also small firms and microenterprises with Support Programs . . . . . . . . . . . . . . 15 4.1  Engendering Programs . . . . . . . . . . . . . . . 15 1 Aguire et al. (2012). 4.2  Supporting Crossovers . . . . . . . . . . . . . . . 16 2 Bardasi et al. (2011). 3 Antoinette Schoar (2010) “The Divide between Subsistence and 4.3  Changing the Narrative and Transformational Entrepreneurship,” in Josh Lerner and Scott Reforming Legal Institutions . . . . . . . . . . 16 Stern Innovation Policy and the Economy, Volume 10 pages 57–81 NBER Books, National Bureau of Economic Research. 5. Conclusion . . . . . . . . . . . . . . . . . . . . 17 4 It is also likely that the combination and intensity of skills and References . . . . . . . . . . . . . . . . . . . . . . 18 support that growth entrepreneurs require is different from necessity entrepreneurs, and as a result interventions support- ing both groups should be differentiated. 2 Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 FIGURE 1 Entrepreneurship Prevalence Rates by Gender in Selected Countries (gender gap declining from left to right) 70% 60% 50% 40% 30% 20% 10% 0 Jordan Egypt India Colombia Turkey Norway China Brazil United States Indonesia Sweden Germany Uganda Philippines Peru France Malaysia South Africa Angola UK All Male Female Source: Authors’ calculations from GEM data 2001–2008. Entrepreneurship rate is defined as the share of nascent, early-stage entrepreneurs and owner-managers in total population. *Lowest gender gap: Angola. *Uganda has a lower gender gap than USA, UK, and Germany. growth potential. Little is known about the relative there is a performance gap, what explains it? The merits of using entrepreneurship programs to support answer to the first question determines whether “necessity” entrepreneurs, self-employed and other specific focus on female entrepreneurs is justified. The groups, especially vis-à-vis other support programs to answer to the second question determines what type facilitate integration into the labor market or social of intervention, if any, is required. assistance. More evidence is needed to understand how better to support these groups in order to improve their The emergence of gender-disaggregated cross- incomes, and whether entrepreneurship programs are country entrepreneurship surveys, such as the Global the best instrument to do so. Entrepreneurship Monitor (GEM),5 several national firm-level surveys, and access to national business This note is structured as follows: Section 2 summarizes registries, has led to empirical studies characterizing the main facts about the gender gap in business female entrepreneurial activities. Most of this literature performance and the explanations that have been has focused on OECD countries, but an increasing put forth. Section 3 reviews the impact of existing number of studies analyze female entrepreneurship in support programs and draws some lessons that could developing countries. These datasets are not perfectly inform the design of programs to support female designed for studying entrepreneurship dynamics, entrepreneurs. Section 4 suggests new areas of focus since they capture neither the decision to become for these programs, and Section 5 concludes. an entrepreneur nor firm dynamics for existing entrepreneurs. However, they provide a starting point ENTREPRENEURSHIP AND THE 2.  GENDER GAP IN PERFORMANCE 5 GEM (2012) for example, provides a rich overview of female entrepreneurship in 67 countries, and several papers have Should there be entrepreneurship support programs summarized some of the evidence regarding the gender gap in targeted specifically at female entrepreneurs? The developing countries (See Minnitti and Naude, 2010 or Klapper answer largely depends on the answers to two and Parker, 2010). One weakness of the dataset, however, is the reliance on self-reported measures of entrepreneurship. additional questions: how do female-led firms compare Thus, interviewees self-report whether they are entrepreneurs to male-led firms in terms of performance? And, if without verification of the enterprise. Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 3 for studying entrepreneurial activity in developing opportunities)8 rather than in pursuit of profit and countries. growth opportunities. Performance of Female-led Firms vs. 2.1  Male-led Firms Women-led Firms are Concentrated in Low- productivity, Low-technology, Low-growth Sectors Lower Entrepreneurship Prevalence Rates among Women not only lead fewer businesses; they also tend Women to concentrate in less profitable sectors. In developed Across the world, entrepreneurship prevalence rates countries, women entrepreneurs are concentrated tend to be lower among women than men, but this in the sales, retail, and service sectors (Klapper and gap is reduced in regions with lower income per capita Johnson, 2012), with little participation in high-growth income (see Figure 1).6 Sub-Saharan Africa (SSA) has or high-technology sectors (Menzies, Diochon, and the lowest gender gap as well as the highest rate of Gasse 2004). A similar sectoral distribution is observed entrepreneurial activity among women, and in some in developing countries. For the SSA region, Hallward- countries, female entrepreneurs are more prevalent Driemeier (2013) shows that women are more likely than male entrepreneurs. These high entrepreneurial to operate in traditional, informal and lower value- rates in SSA are likely explained by the large entry costs, added sectors. Global Entrepreneurship Monitor especially for women, into labor markets.7 (GEM) data reveal two interesting results. First, women entrepreneurs tend to have higher levels of sector Entrepreneurship prevalence rates, however, are only concentration than men, although the concentration part of the story; a deeper look at the data reveals gap varies greatly by country. Second, as Figure 2 that a larger share of female than male entrepreneurs shows, there are male-dominated sectors where male tend to be “necessity” entrepreneurs. That is, entrepreneurs have larger shares, such as real estate more women are driven to entrepreneurship out of or construction, and others that are clearly female- necessity (for example, due to lack of employment dominated, mainly services such as wholesale and retail or personal and other services. These female-dominated sectors are traditionally lower-productivity sectors. FIGURE 2  Gender Distribution of Early-stage Entrepreneurs across Sectors Women-owned Firms are Significantly Smaller Female entrepreneurs Closely associated with the higher incidence of necessity entrepreneurship among women and their 16% concentration in lower productivity sectors is the fact 30% 17% that women-owned firms are significantly smaller than 23% 3% male-owned firms, in terms of employment and sales. 4% 2% 2% 0% Data from World Bank enterprise surveys show that Male entrepreneurs 1% 7% 1% 5% 0% 5% 0% women-owned enterprises have lower overall sales 1% 0% 2% volumes than male-owned firms in Europe and Central 3% 18% 6% 8% 0% 6% 0% 8% 13% 7% 6 The Middle East and North Africa region seems to be an excep- 0% 9% 0% tion and female entrepreneurship rates are much lower than male, likely due to a stronger effect of social norms. Other Agriculture & Forestry Construction 7 Hallward-Driemeier (2013) Education Electricty, gas, water Finacial intermediation 8 It is of course entirely possible that necessity entrepreneurs Fishing Health and social work Hotels and restaurants graduate to opportunity entrepreneurship—but this transition in itself requires certain characteristics which necessity entre- Source: Authors’ calculations from GEM data, 2001–2008. preneurs may or may not have. 4 Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 Asia, Latin America, and SSA.9 They also suggest that entrepreneurs, for example, is mixed. Fairlie and Robb the number of both “gazelles” 10 and high-growth (2009) document higher exit rates for women-owned firms is larger among male-owned firms.11 firms in the United States, but Koellinger et al. (2013) find similar ratios for men and women in 17 OECD and emerging countries.15 Kalleberg and Leicht (1991) and Women-led firms Experience Lower Returns to Bruderl and Preisendorfer (1998) also find evidence for Capital and Lower Profitability the idea that firm survival rates are not different between Studies of microenterprises in Sri Lanka12 and male and female entrepreneurs in developed economies. Madagascar13 find that women-owned firms experience lower returns to capital and lower profitability.14 For example, Fafchamps et al. (2014) find that although In Similar Sectors, Women-led Firms Perform as returns to capital for female-led microenterprises in Well as Peer Firms Led by Men Ghana were high, they were not as high as those for While female-led enterprises tend to be less productive male-led enterprises. Interestingly, Ghanian female- on average than male-led enterprises, these differences led enterprises that have sales above the mean (for all disappear in some countries when comparing firms) are found to have returns to capital similar to male—and female-led enterprises within the same male-led counterparts—which suggests that the gender sector. A review of firms in Central and Eastern Europe16 gap in capital returns may be smaller for women- and Madagascar17 finds that female-owned businesses owned enterprises with initial high profits. tend to be less productive than male-owned firms in the same sector. Other studies present contrasting findings. Bardasi et al. (2007) find that women-owned firms in Performance Gaps are Likely to be Larger at Lower Africa tend to be as productive in terms of value added Income per capita Levels per worker and total factor productivity (TFP) as male- The evidence on the gender performance gap in owned firms. Comparing labor productivity between developed countries tends to be more mixed than in male and female businesses in SSA for the same developing countries. For example, Watson (2002) industry, size, and capital intensity, Hallward-Driemeir shows that Australian women business owners earn (2013) finds no gender gap in productivity. similar rates of return on equity and assets as their male counterparts, but they have less startup capital, which explains their lower incomes and profits compared to 9 Bardasi et al. (2011) male business owners. Kepler and Shane (2007) find no 10 Gazelles are defined here as young firms of less than 6 years old that grow at a rate of 20% per year for two consecutive years, significant gender differences in terms of performance since the enterprise surveys only provide sales information for outcomes in nascent entrepreneurs in the United a three year period. High growth firms on the other hand are States. Other studies suggest that women-owned defined as firms that also grow at 20% for two years but from a size of at least 10 workers or more, to correct for the large enterprises perform as well as male peers in terms of growth bias of very small firms and that are of any age. employment creation in OECD countries (Fischer et al. 11 Looking across all the countries with data available, the 1993; Chaganti and Parasuraman, 1996). weighted average number of high growth firms is 99.83 for male firms and 66.45 for female firms, and for gazelles this is 25.37 and 14.07 respectively. 12 De Mel, McKenzie, and Woodruff (2008) No Gender Differences in some Performance 13 Normand and Vaillant (2013) Indicators 14 Aterido and Hallward-Driemeier (2011); Bardasi et al. (2011); The outlook for women-led enterprises is not uniformly De Mel et al. (2009); Nichter and Goldmark (2009) 15 Analyze the difference between male and female exit rates by gloomy and not all performance outcomes are looking at the ratios between nascent and established entre- significantly different by gender. The evidence on the preneurs in the GEM dataset. differences in firm survival between male and female 16 Bardasi and Terrell (2008) 17 Normand and Vaillant (2013) Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 5 Using data from the harmonized enterprise surveys over others? Understanding the reasons will help inform (ES), this study estimates the gender gap premium on the design of programs and enhance their impact. TFP18 for firms in 82 countries, that is, whether there are systematic differences in productivity between Traditionally, the literature has differentiated between male—and female-led enterprises, controlling two sets of potential explanations (Bardasi et al., for sector participation. There are no statistically 2011): constraint-driven factors, that is, external significant differences in productivity between female barriers stemming from institutions, the regulatory and male firms in 90 percent of the countries in the environment, and social norms, and preference-driven sample.19 This suggests that, conditional on entry into factors arising from internal motivations such as risk entrepreneurship and sector sorting, there are no tolerance and subjective preferences. In practice, significant productivity differences between female and however, it is difficult to disentangle the two types male entrepreneurs in most countries. of determinants, since cultural norms also shape preferences, and preferences inform social attitudes. Women-led Businesses Have a Higher Share of Female Employment Constraint-driven Factors Another interesting trend that emerges from the analysis Regarding constraint-driven factors, initial conditions are of the ES is that female-led firms tend to employ more the first important consideration. Female entrepreneurs female workers as a share of the total workforce, often start businesses with less startup capital and less after controlling for sector, age, and type of firm.20 access to credit, less experience, and less schooling, and In 49 percent of the countries in the sample, there are no significant differences in the share of female employment between male—and female-led firms. 18 We calculate TFP for each country as the residual of the follow- ing Cobb-Douglas production function, where Y is the loga- However, in the remaining 51 percent of countries, the rithm of sales or value added, K is the log of capital, L is the average difference is statistically significant and positive, log of employment and M log of material inputs. Only 8.5% of the countries in the sample have a statistically significant indicating that female-led firms tend to have a larger gender gap. share of female workers than firms led by men. log(T FPi ) = log(Yi ) − α log(K i ) − β log(Mi ) − ∑ S k Then, we estimate the gender gap in productivity by regressing for each country separately firm TFP on a dummy for firms that Female-led Firms Do Underperform! have at least one female owner and controlling for sector and Summing up and answering the question about whether firm size. One caveat of the analysis is the fact that our mea- female-led firms underperform, posed at the beginning of sure of female ownership is imperfect, since we do not know the extent of female owners’ control over the firm. the section, the evidence points to a gender performance 19 The countries with negative productivity gap, firms with gap between female and male entrepreneurs. This is at least one women owner less productive, are Botswana, manifested primarily as lower entry into entrepreneurship Indonesia, Jamaica and Argentina. The countries with a posi- tive gender gap are China, Iraq and Bulgaria. As a further and sorting of existing female entrepreneurs into lower- check and to control for the fact that female owned firms may productivity business sectors and activities. As a result, not be able to grow and therefore may have smaller size and productivity levels, we match male and female owned firms some of these differences disappear once performance is according to age, sector and type of firm (part of a group or compared within the same sector. limited liability company) and look at differences in productivity between matched firms for each country. We find that in most countries (84.4% of cases) there are no statistically differences in productivity between male and female owned firms, once 2.2  What Explains the Performance Gap? firms are matched by sector and firm size. If performance gaps are significantly reduced or even 20 We use propensity matching techniques and for each country disappear entirely when controlling for initial conditions we match male and female led firms based on sector, age of the firm and type of firm (public,..). Then we compute the and sectors of operation, the critical question becomes, average difference in female employment shares analogously What motivates women to operate in certain sectors to the average treatment effect. 6 Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 lower level of management skills, which constrains their Bank’s Women, Business and the Law (WBL) measures growth and chances of success. Fairlei and Robb (2008), the extent to which women and men have equal for example, find that women-owned businesses in access to economic opportunities. Research shows the United States are less successful than male-owned that the higher the number of gender-based legal businesses because they have less startup capital, less restrictions, the lower the female participation in firm business experience, and less human capital. This is ownership (Figure 3). For example, the rights of married confirmed by Cohoon et al. (2010) in their study of a and unmarried women affect women entrepreneurs’ selective cohort of 540 successful entrepreneurs in high- capacity to register businesses, sign contracts, open tech companies, which finds no differences in education bank accounts, as well as their mobility, measured and access to capital between male and female by their ability to choose a residence or obtain a entrepreneurs. Thus, when initial conditions are similar, passport. According to WBL, 200 million women live outcomes are also similar, regardless of gender. in countries which legally require formal permission from the husband or a male relative to start a business. A far more complex barrier to female entrepreneurship There has been some progress in introducing reforms are social norms, which induce female entrepreneurs to aimed at reducing gender-based economic restrictions. choose socially acceptable sectors and can impact their In Ethiopia, for example, reforms in family law that perceptions about what they are capable of achieving.21 allowed women to pursue economic opportunities Social norms also limit educational opportunities and without requiring the husband’s permission and joint access to finance, affecting initial conditions for female- administration of marital property led to increases in led enterprises, and influencing intra-household choices women’s participation in work outside the home, full- that constrain entrepreneurial decisions. Normand and time work, and higher-skilled work.22 Most countries, Vaillant (2013) find that informal female entrepreneurs however, are still lagging behind in addressing legal in Madagascar self-select into industries in which they constraints to female entrepreneurship. can combine market-oriented and domestic activities. Comparing male and female entrepreneurs in India in households where one or both members may have an FIGURE 3 G  ender-based Legal Restrictions and Female enterprise, Field, Pande, and Rigol (2014) find much Participation in Business Ownership weaker performance in terms of profits and higher 80 concentration in low-profitability sectors for female % of firms with female participation in ownership entrepreneurs in households where both men and 60 women have businesses. They attribute this to the fact that the household is the unit of decision for female entrepreneurs when making economic choices. It is 40 consistent with findings that female entrepreneurs may invest loans or grants in male-led enterprises if they 20 are the more profitable enterprise in the household. Although these female enterprise decisions might be optimal for the household, they can substantially limit 0 the potential for enterprise growth and in some cases 0 5 10 15 20 25 women’s empowerment. More research on these Number of legal restrictions complex interactions within the household is needed in Sources: Women, Business and the Law database, Enterprise Surveys database, World Development Indicators database. order to help design support programs. Barriers to female entrepreneurship are also embedded in legal frameworks and institutions, especially in 21 Field et al. (2012) developing countries (Bardasi et al., 2007). The World 22 Hallward-Driemeier (2010) Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 7 Preference-driven Factors TAKING STOCK: LESSONS 3.  A second set of interlinked elements that determine FROM EXISTING WOMEN entrepreneurial choices are subjective preferences and ENTREPRENEURSHIP SUPPORT personal traits. The psychology literature suggests that PROGRAMS there is a set of subjective traits that are important The trends and facts outlined above combined with predictors of entrepreneurial activity, such as perceptions the growth imperative in developing countries have led of opportunity, self-confidence, and fear of failure to an increasing number of support programs aimed (Koellinger et al., 2007). In many cases, women rank at facilitating the growth of female-led enterprises. lower than men in these perceptions. For example, GEM The following section reviews the lessons from these data show that across countries, early-stage female programs. entrepreneurs tend to exhibit significantly higher fear of failure than male entrepreneurs. Minniti (2010b), A review of entrepreneurship programs is also using the GEM dataset, estimates that subjective complicated by the dearth of impact evaluations perceptions about one’s own skills, likelihood of failure, of programs supporting growth entrepreneurs in and the existence of opportunities explain a significant developing countries. Most evaluations focus on the portion of the gender gap in entrepreneurial activity. In a microenterprise sector, which includes both micro- review of the experimental literature on gender differences entrepreneurs with a desire to grow their business in risk, social, and competitive preferences, Croson and “necessity” entrepreneurs. Therefore, it is difficult and Gneezy (2009) find that the bulk of the evidence to extract concrete lessons for the design of support highlights fundamental differences between male and programs targeted primarily at growth entrepreneurs. female preferences in relation to entrepreneurial activity. They also find that women are more risk averse than men and less likely to engage in competitive situations.23 Impact of Support Programs is Mixed at Best A meta-analysis of impact evaluations of Gender differences have also been found when entrepreneurship programs for SMEs and examining motivations for business ownership. Fairlei and microenterprises in developing countries (Cho and Robb (2008) find evidence that women entrepreneurs in Honorati 2013) finds an overall positive impact on the United States work fewer hours and have different the business knowledge and management practices goals for their businesses. In their study of successful of entrepreneurs in the programs. But impact on high-tech entrepreneurs in the United States, Cohoon business growth is limited (for example, positive et al. (2010) find gender differences in motivation business performance outcomes were observed for for starting a business. For female entrepreneurs, youth and entrepreneurs with higher education). More encouragement from peers is the most important specifically for female entrepreneurs, the authors find motivation in starting a business, and networks are cited improvements in business attitudes but no significant as critical for success. Knowing other entrepreneurs and impacts on business growth outcomes. McKenzie and access to networks are especially important for women Woodruff (2012), focusing on business education in entrepreneurs in poorer countries.24 their analysis, find weak evidence of positive effects of these programs on male-owned businesses and zero or Is nature or nurture the source of these differences? Croson negative impact on women-owned firms. and Gneezy’s (2009) review supports both explanations and suggests that the critical element is to understand the The lack of impact of support programs on business relative weight that should be given to each determinant. outcomes of female-led enterprises, even as intermediate The fact that social norms and institutions limiting female outcomes improve, also appears to be a common finding entrepreneurial activity are stronger in developing countries might suggest greater influence of nurture on these 23 Croson and Gneezy (2009) preferences in these countries. 24 Minniti (2010) 8 Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 in the empirical literature on microenterprises. Berge specifically at growth entrepreneurs. Depending on the et al. (2012) find positive effects of business training local context, programs could opt to focus exclusively and financing conditional on gender for micro firms on women-led small and medium-size firms with high in Tanzania. Although women entrepreneurs gained growth potential. In environments where women business knowledge, it failed to translate into improved entrepreneurs are highly constrained, there may be a business performance. De Mel et al. (2014) evaluate a concurrent focus on female-led microenterprises that program combining standard business training with cash can be scaled up to SMEs. grants and find that business training alone results in improved business practices but has no significant impact Therefore, it is necessary to carefully delineate the on business profits. However, when business training target segment for a proposed support program and is combined with a grant, there is a positive impact on then develop a selection process to identify members of profits, which dissipates after the first eight months. the segment from among the applicants to a program. Not all studies, however, find a lack of impact of support programs on female business outcomes. For Selecting Growth-oriented Entrepreneurs is Complex example, Valdivia (2011) finds a positive impact of Selecting growth-oriented entrepreneurs running business education on sales among women-owned microenterprises and/or SMEs is a highly complex microenterprises in Peru when the educational package is task with few templates for success. One critical combined with technical assistance. This impact, however, principle when designing selection is the need to was measured only once, at the four-month mark, with accept some degree of failure, since it is impossible no measurement of the medium-to-long-term impact. ex ante to determine firm success with certainty.26 However, there are several distinct personality traits Three main themes emerge from the evaluations of these that are correlated with entrepreneurial ability,27 programs. First, these programs appear to be successful in including high need for achievement, calculated risk- improving the management skills of beneficiaries—both taking, internal locus of control, a problem-solving men and women. Second, there appears to be little to orientation, interpersonal reactivity, and assertiveness. no impact on firm growth among female entrepreneurs. A study of rural entrepreneurs in India28 finds that Third, there is a gender differential in attrition rates, personal characteristics like self-efficacy and need business outcomes, and overall program efficacy. for achievement explained a significant part of the variation in sales. Entrepreneurial orientation can be Two reasons may account for the lack of effectiveness identified through questionnaires and panel interviews. of entrepreneurship programs for women. First, beneficiaries of such programs may be primarily In addition to these non-cognitive skills, cognitive necessity entrepreneurs, who are less likely to grow skills may also be good predictors of entrepreneurial their businesses or be motivated to do so. There is success. A study in Sri Lanka29 finds that “higher-ability some evidence that impact is enhanced when larger SME owners are more likely to add employees,” with enterprises are targeted.25 Second, the programs ability proxied by scores on multiple cognitive tests30 evaluated may not be designed to address the main binding constraints to female enterprise growth. 25 Valdivia (2011) 26 According to National venture capital association (http://www. nvca.org/index.php?Itemid=147), only 40% of VC backed 3.1  Selection firms experience high growth, or even as low as 20% in some cases (Mulcahy et al, 2012). Having determined that not all potential or 27 Croson and Gneezy (2009) existing entrepreneurs have either the desire or the 28 Acharya et al. (2007) entrepreneurial traits to grow their businesses, there is 29 De Mel et al. (2010) a strong argument to be made for programs targeted 30 Digit span, Raven, and cognitive reflection. Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 9 and years of schooling. Also, the authors find that In general, more experimentation with different owners with a greater need for achievement and who methods of selecting female growth entrepreneurs is are more willing to give up control (power motivation) needed, especially given the limitations of the models were more likely to generate employment. A survey of to take into account particular cultural contexts. As a entrepreneurs in Brazil found that school achievement result, combining model predictions based on ability (i.e., cognitive ability) and father’s higher education (cognitive and non-cognitive skills) and expert panels (i.e., family characteristics) were significant predictors of may be desirable, especially with a focus on identifying entrepreneurial success (proxied by sales growth).31 entrepreneurs with no desire to grow their business, rather than predicting success. Programs that support growth entrepreneurs have mainly used panel interviews with experts, sometimes in combination with specific questions on the application 3.2  Business Education form to identify growth potential. In a review of support programs for growth-potential women entrepreneurs in Business Practices Are Positively Impacted but developing countries, Drexler (2014) suggests that panel Business Growth is Not opinions tend to have the most influence in selection.32 General business education or training35 is the most In a recent impact evaluation of a program supporting popular component of entrepreneurship support growth entrepreneurs in Ghana, Fafchamps and Woodruff programs. Most evidence shows that business (2014) test the predicted value of expert panel opinions in training programs yield positive changes in business forecasting firm growth. They find that cognitive skills are practices (most notably in record keeping) but are a good predictor of firm growth, but this prediction can largely ineffective in terms of actual business growth be enhanced by considering the scores from expert panels, (measured in sales, profits, and employment growth) especially on bad performers. Experts’ opinion did not add for women entrepreneurs.36 Some limited effects on much explanatory power in predicting best performers. profits and sales were, however, found in a broader Predictive Models can be Helpful but Require 31 Djankov et al. (2007) Further Experimentation 32 The use of questionnaires that identify entrepreneurial traits during selection is still not widespread, and it is commonly Some predictive models using information about restricted to baseline data collection for impact evaluations. the personality traits of applicants are gathering 33 See for example the Entrepreneurial Finance Lab http://www. momentum. For example, several psychometric tests33 hks.harvard.edu/centers/cid/programs/entrepreneurial-finance- lab-research-initiative or psychometrictest.org.uk and statistical models34 have been developed to capture 34 De Mel et al. (2012); Grimm et al. (2012) entrepreneurship capabilities. But much work remains 35 here we take training to mean less intense interactions of to be done to determine the range of applicability of shorter duration than academically grounded, more MBA like these methods. business education . 36 One exception is the evidence from a randomized controlled trial in rural Mexico, (Calderon et al., 2013). The program Considering that the evidence on selection is still too provided a six week long training to women entrepreneurs in- narrow to make generalizations, it is important to cluding modules on costing and pricing, formalizing a business, business management, marketing and sales. Medium term recognize the inherent uncertainty in identifying growth results (1 and 2.5 years after the program) showed a 23% in- entrepreneurs. One cautious but more resource-intensive crease in daily profits and 29% increase in revenues. The use of approach to selection that minimizes exclusive reliance formal accounting practices rose by 4.7 percentage points and formal business registration (a theme specifically addressed dur- on information from the application process is to have ing the training) increased by 8.6 percentage points while there a higher proportion of applicants automatically selected were suggestions that participants changes the product mix in for a “light training” and base acceptance on reviewing their enterprises to add low cost goods with higher revenues. Results also showed that lower quality entrepreneurs—proxied their commitment to pursuing growth and reviewing by below-median pre-intervention profits—were more likely to their growth plans developed during the training. close down their businesses after the training. 10 Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 FIGURE 4 Percentage of Entrepreneurship Programs* with Selected Components 100 90 80 70 60 50 40 30 20 10 0 Innovation Business Business Financial Accounting Marketing HR Customer Networking Exports Access to Gender Country Managing plan/ management Planning analysis management service capital/ specific specific a family Strategy Loans challenges challenges business *programs included in analysis by Mckenzie and Woodruff (2012) and Drexler (2014). Categories pooled together by authors. study of entrepreneurship education and training also confirms this heterogeneity. Most programs are programs conducted by the World Bank (2013).37 grounded in similar approaches that combine business education and entrepreneurial skills.41 Some have actually been standardized for delivery across several Programs are Highly Heterogeneous countries.42 However, as Figure 4 illustrates, the specific An analysis of more than 80 entrepreneurship training topics covered—accounting, financial planning, and education programs by Robb et al. (2014) illustrates marketing, and business management—vary greatly the content of these programs. Broadly speaking, even though the context of their implementation (i.e. general business skills, including financial skills, are most low-income countries) is fairly similar. Topics such as common, followed by marketing and sales training. human resources or customer service are less frequently Only three programs (for both men and women) have offered, and an even smaller number of programs cover industry-specific content and a mere eight incorporate topics specifically relevant for female entrepreneurs, content that takes into account the specific context such as business networking or managing family of the participants. The duration of programs varies businesses, reflecting the absence of gender-specific greatly—from less than two weeks to more than a year. content in programs ostensibly catering to women A more limited sample of 20 women-only programs entrepreneurs. adds nuance to the analysis.38 Most of these are aimed at existing entrepreneurs and only four are specifically oriented toward stimulating gains in confidence and communication in combination with general business 37 Framing the Global Landscape of Entrepreneurship Education skills. Similarly, McKenzie and Woodruff (2012)39 note and Training Programs, World Bank (2013) heterogeneity in terms of length of training, course 38 Robb et al. (2014) content, and delivery mechanisms (through professional 39 McKenzie et al. (2012) trainers in some instances, microfinance staff in others, 40 Drexler (2014) and practicing entrepreneurs). 41 Some of these approaches include Babson College, EMPRETEC methodology developed by David McClelland at Harvard University and others. Drexler’s (2014) review of programs aimed specifically 42 UNCTAD EMPRETEC, ILO’s Start and Improve Your Business, at women entrepreneurs with growth potential40 IFC’s Business Edge Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 11 Robb et al. (2014) suggest that the design of the 3.3  Networking and Mentoring business education component matters. They find that Networking and mentoring play a critical role in programs combining both general business education improving women’s entrepreneurial decisions and and soft skills such as leadership and teamwork, yield performance. Yet, as the GEM 2010 Women’s Report significant positive impacts in both management found, female entrepreneurs across a range of practices and firm performance outcomes. In South countries have smaller and less diverse networks and Africa, for example, where the program also provided tend to rely more heavily on personal sources of advice mentoring services, participants were able to generate than male entrepreneurs. significant increases in annual sales/turnover, success of the business, number of employees, and number of customers, according to a rigorous impact evaluation. Networking and Mentoring are Important Resources Another program in Nicaragua observed a significant for Female Entrepreneurs increase in revenues and number of employees hired A study based on surveys in Nigeria and Argentina in addition to improved business skills and increased suggests that mentors and networks provide invaluable confidence. Indeed, the program also increased market information to female entrepreneurs, facilitate mentoring among businesswomen and the formation access to finance, and provide intangible benefits in of personal and professional networks using social terms of support and increased confidence. More than networking tools. 80 percent of women in Argentina and more than 90 percent in Nigeria reported significant benefits from Very few programs include strategic planning in their networking, including mentoring, access to information, curriculum, and not many programs seek to influence and professional visibility. Half of the women surveyed entrepreneurial mindsets. Networking was an explicit depended on personal contacts for access to finance, focus in only two programs, in India and Mexico. Both and more than 35 percent in both countries received programs also offered soft skills in the curriculum, and advice regularly from other women in the network.44 both resulted in increased revenues for participants.43 These programs also offered industry-specific Networking and mentoring have also been found content. It is not possible to know which of these to be critical in motivating women entrepreneurs to add-ons—networking, soft skills, or industry-specific lead crossover enterprises, that is, firms operating in content—had the most to do with positive business more productive, male-dominated sectors. Campos outcomes. This gap in knowledge merits further et al. (2013) investigate the attributes of women who investigation. successfully cross over in Uganda. While human capital does not appear to be a significant determinant, access In the absence of conclusive evidence about a business to a role model, in addition to initial capital, is a critical education approach that works best for growth- determinant of crossovers. Women who had male role oriented women entrepreneurs, heterogeneity of models were between 55 and 74 percent more likely to business education content and approaches allows cross over into higher-productivity sectors than women also for useful experimentation and calibration of who had no such access. Fifty-four percent interacted topics that are most pertinent for a specific segment of with other business owners at least once a month, while women entrepreneurs. New approaches in engendering only 39 percent of non-crossovers did. These results the curricula, topics on market expansion, and technical training are likely to be important but require 43 However the evaluations of these programs did not have an further experimentation in order to better understand appropriate counterfactual. how best to incorporate these elements in support 44 The Benefits of Women’s Business Networks (Vital Voices in programs. ´ et al., 2012) Buvinic 12 Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 stress the importance of including in support programs program for both male and female-managed SMEs, regular productive interactions with role models and Bruhn et al. (2012), using a randomized controlled inductions into industry networks, while also providing trial in Mexico,46 find that entrepreneurs who received access to finance to reduce the gap in initial capital.45 customized consulting services47 over a period of one year not only improved business practices but were also able to raise sales by 80 percent and monthly profits by Effective Provision of These Services Requires 120 percent. Further Experimentation There is significant anecdotal evidence pointing In a study of microenterprises, Valdivia (2011) finds towards the positive impact of mentoring and positive impacts of additional services in Peru. Female networking on women’s entrepreneurship activities, owners of microenterprises were randomly assigned to and many entrepreneurship programs have two treatment groups.48 One received general business incorporated the provision of these services. For training (over four weeks) and the other received example, 70 percent of growth entrepreneur programs both the general training and customized technical reviewed by Drexler include mentoring components. assistance (an additional three months). The latter Several approaches have been tested in existing recorded sales 19 percent higher than their peers in the programs. High-impact entrepreneurship programs, general training group and improvements in business such as Endeavor, typically rely on experienced and practices (formality of the business and management of successful volunteer mentors who operate within the business/household account). In contrast, the group guidelines to prevent potential conflicts of interest. that received only general business training was more Other programs focusing on micro and small-scale likely to close loss-making businesses rather than make entrepreneurs in developing countries encounter improvements in enterprise management. difficulties recruiting appropriate mentors. Some use consultants, while others find faculty or entrepreneurs Focusing on larger textile firms in India (with only one who need to be paid in order to keep them engaged in woman-owned enterprise in the cohort), Bloom et the program. The relative merits of these approaches al. (2013) find that adopting management practices have yet to be directly compared, and further learned through consulting services increased the experimentation is needed to structure the delivery of productivity of treated firms by 17 percent in the first networking and mentoring services for impact. year, in line with the growing literature emphasizing the importance of management quality for firm productivity Anecdotal evidence also indicates that networking (Bloom and Van Reenen, 2007).49 This evidence for events with other female entrepreneurs can be inspiring. However, access to male networks can 45 Breaking the metal ceiling: Female entrepreneurs who succeed probably help women access valuable market in male-dominated sectors in Uganda (Campos et al. in Buvinic ´ et al., 2012) information and facilitate crossover to a greater degree. 46 The Impact of consulting Services on Small and Medium Enterprises: Evidence from a Randomized Controlled Trial in Mexico, Bruhn et al. (2012) 3.4  Other Services 47 Consulting services included a diagnosis of the growth con- straints for individual enterprises, suggested solutions to over- come the constraints, and implementation of the solutions. Additional Business Services can have Positive 48 Training or technical assistance for female entrepreneurship? Impact Evidence from a field experiment in Peru, Valdivia (2011) 49 Bloom, Nicholas, and John Van Reenen, ‘‘Measuring Other services, such as the availability of consultancy or and Explaining Management Practices across Firms and technology extension services, can enhance the impact Countries,’’ Quarterly Journal of Economics, 122, no. 4 (2007), of entrepreneurship support programs. In a mixed 1341–1408. Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 13 larger firms, however, contrasts with the evidence female-led microenterprises in Ghana, with in-kind for microenterprises presented above. While support grants (equipment) producing significantly higher programs for microenterprises appear to improve monthly profits for both groups and cash grants management, this does not translate into enterprise producing higher profits only for male-led enterprises.51 growth. The findings of Bruhn et al. (2012) for SMEs in Furthermore, the equipment transfers were significant Mexico suggest that customized business services may only for women-led firms that were larger in size (in be more effective in improving management quality. terms of capital stock) at the baseline, even for firms More evidence is needed in this case, especially for in similar sectors. This appears to bolster the idea that female entrepreneurs. the returns to capital of female enterprises are lower, although female enterprises of larger size could benefit substantially from greater access to finance. There are Important Supply-side Constraints in the Quality of these Services Not all of the evidence on access to finance for women Despite the potential enabling impact of these entrepreneurs is discouraging. For example, Klinger and complementary services, Drexler’s (2014) review Schündeln (2011) analyze the impact of TechnoServe’s suggests that most programs for women-owned mixed-gender business plan training program in businesses with growth potential do not offer these Guatemala, El Salvador, and Nicaragua and show that services, mainly because of the cost. Consultancy the effect on starting/expanding a business is stronger costs in some programs are four times the cost of as a result of receiving a cash grant, although the effect the business education component, and high-quality is smaller for women.52 Attanasio et al. (2011b) find consultants in developing countries are in short supply. that financing through group borrowing has a positive This suggests that programs seeking to offer such impact on the creation and survivability of women-led services need to consider aspects related to the demand microenterprises in Mongolia.53 as well as addressing supply-side constraints. While it is clear from the assessment of constraints faced by female entrepreneurs that some form of 3.5  Access to Finance access to finance is necessary for female-led enterprises Female entrepreneurs experience more difficulties in to achieve growth, the evidence on the impact of accessing finance than male entrepreneurs, and thus exclusively finance-oriented support programs is limited tend to start their businesses undercapitalized. It may and mostly focused on microenterprises. Studies on therefore be critical to incorporate access to finance to microenterprises suggest that access to finance alone female entrepreneurs into support programs. has a limited impact on productivity and that, when combined with other training programs, outcomes are stronger when provided in the form of in-kind Access to Finance has Mixed Impact, Depending on assistance rather than cash grants or group borrowing. Firm Characteristics A significant amount of evidence has been produced in the microfinance literature that analyzes the integration 50 Other studies finding no impact of combining access to micro- finance and business education are Gine and Mansouri (2012) of business education with access to finance for for Pakistan or Berge et al. (2011) for Tanzania. some female-led microenterprises. Outcomes have 51 When is capital enough to get female microenterprises been mixed. In a study on Peru, Karlan and Valdivia growing? Evidence from a randomized experiment in Ghana, (2011), for example, find no evidence of impact on Fafchamps (2011) 52 Can Entrepreneurial Activity be Taught? Quasi-Experimental key outcomes such as business revenue, profits, and Evidence from Central America (2007) employment.50 Fafchamps et al. (2011) compare 53 Group lending or individual lending? Evidence from a random- the effects of cash and in-kind grants for male and ized field experiment in Mongolia, Attanasio et al. (2011) 14 Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 There is some evidence that larger female enterprises LOOKING FORWARD: AREAS 4.  are more likely to benefit from access to finance. More OF FOCUS FOR FEMALE research on how to best structure the provision of ENTREPRENEURSHIP SUPPORT finance for growth-oriented women entrepreneurs is PROGRAMS also needed. The preceding review sheds light on some of the constraints faced by women entrepreneurs and suggests some potential solutions. These constraints 3.6  Monitoring and Evaluation include initial conditions (at the startup stage), such Underlying this entire discussion is the glaring absence as access to finance, education, information, and of a critical mass of robust evidence on what works for networks, which lead to sorting into low-productivity growth-oriented women entrepreneurs. Considering and low growth-potential sectors. Subjective the importance of support programs for women preferences and the overall business environment, entrepreneurs to the global agenda for economic including institutional and legal frameworks, also growth and shared prosperity, there is an urgent influence managerial choices and firm decisions, need to incorporate robust evaluation mechanisms, which has a direct bearing on business performance specifically to invest in impact evaluations, in order to outcomes. Further experimenting and robust determine what types of support are most effective, evaluations are necessary to enrich program design which women entrepreneurs benefit the most, and to grounded in evidence. what extent. One potential explanation of why support programs for women entrepreneurs have weaker outcomes Better Impact Evaluations and Longer Term than those for their male peers is that the programs Measurement of Outcomes is Needed tend to focus on improving initial resource conditions Establishing key performance indicators that monitor but rarely address the remaining binding constraints ongoing changes in both business practices (i.e., record to female enterprise growth, such as concentration keeping and business accounts) and business outcomes in low-productivity and low-growth activities, intra- (i.e., sales, profit, and job creation) is an integral part household choices, and the role of institutions and of an effective evaluation. Additionally, measuring legal frameworks in shaping the overall business participant satisfaction with the various components environment for women entrepreneurs. Addressing can increase the responsiveness of program design. these constraints requires designing programs more However, more evaluations using the appropriate holistically and combining instruments and services counterfactual methodologies are needed. oriented to addressing these barriers. There are three areas that can potentially increase the effectiveness Currently, most impact evaluations stop collecting of these programs and where more evidence and data one year after implementation of the experimentation is required to calibrate programs to program. However, firm growth is not a short- specific segments of female growth entrepreneurs. term phenomenon, nor is there necessarily a linear relationship between accumulating business knowledge and improved performance outcomes. 4.1  Engendering Programs The evidence regarding the positive impact of these How can programs be engendered so that most of programs on business practices is quite robust. Ideally, the constraints faced specifically by female growth program evaluations should attempt to measure both entrepreneurs are addressed? Existing support the longevity of these changes and the medium- programs tend to engender programs by limiting term impact of these positive changes on business participation to women only. Some programs also performance and firm growth. emphasize specific cognitive and non-cognitive skills Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 15 where female entrepreneurs are constrained, such as As Campos et al. (2013) suggest in the Uganda case negotiation, leadership, and communication skills. study, supporting crossovers requires programs to However, this does not seem to address all constraints. increase exposure of women entrepreneurs to these For example, mobility and opportunity cost constraints sectors by using appropriate mentors and facilitating are rarely addressed. The delivery mechanism also access to information. One example of a successful needs to take into account gender-specific constraints. crossover program in Bangladesh supported women Peer learning from other female (or male) peers and entrepreneurs’ entry into the clean cook stoves value linkages with mentors and relatable role models who chain by providing access to credit and incorporating can be inspirational for women are also potential other organizations, such as business associations, into engendering mechanisms that can be adopted. More the program.54 Reaching out to community leaders evidence is needed on how to engender the content was also an important strategy for increasing the of business education. Using women-centered case acceptance of women entrepreneurs in society (Buvinic ´, studies, employing instructors and trainers qualified to et al., 2012).55 address gender constraints, and emphasizing practical soft skills like negotiating in male-dominated markets. Although there is scant literature in this area, Penrose’s might make the support more effective. (1959) theory of the firm showed that firm capability is critical for firm growth and diversification. Female In addition, support programs should better integrate entrepreneurs will succeed in crossovers if they acquire household constraints. Only a few programs provide the capability to do so. This implies that entry in similar specific wraparound services that support women activities of higher value added within the same sector entrepreneurs by providing support for childcare or requiring similar capabilities may yield larger returns. joint sessions with spouses to discuss the woman’s Two types of programs may be useful in building enterprise and household responsibilities. Some of the crossover capabilities: (i) entrepreneurship programs self-evaluations of these programs suggest that such that incorporate more technical training and other wraparound services could have a positive impact on customized services to facilitate this process of activity empowering women and their business decisions, crossover, and; (ii) project-oriented programs that although more research is needed to understand how target specific sectors and activities and support links to better design these wraparound services. to particular supply chains with business and technical training and access to markets. Given the large potential impact of these crossover interventions, it is 4.2  Supporting Crossovers critical to experiment with these approaches. Stimulating female entrepreneurs’ entry into higher- productivity and growth activities, often male- dominated, is critical to narrow the performance gap Changing the Narrative and Reforming 4.3  between male and female entrepreneurs. In cases Legal Institutions of large concentration of female entrepreneurs in The final piece of the puzzle is the overall environment low-productivity and low-growth sectors, supporting for women entrepreneurs, including public perceptions crossovers is likely to be the only way to generate and the legal and institutional framework. Traditional significant growth in female-led enterprises. Crossing over in this context refers to both entry into different sectors that have higher productivity and within-sector 54 One additional benefit of the program was that women-led firms employed more women; almost 80% of their employees diversification into higher value-added activities. This were women compared to 10–15% female employment in type of intervention needs to be integrated into support male-led firms. programs for women entrepreneurs. ´ et al. (2012) 55 Dey in Buvinic 16 Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 narratives, for the most part, confine women to These findings suggest that there is significant room specific business categories, usually microenterprises, for improving entrepreneurship support programs, with little room for growth. Often this does not reflect by offering mentoring, networking, and other the complex reality of women-led enterprises, which consulting services, in addition to education on operate in all sectors and come in all sizes. Part of the basic business practices and by strengthening critical challenge of changing the self-perceptions of women areas that can potentially increase the effectiveness entrepreneurs is their limited access to successful of these programs. More experimentation is peers and role models in high-growth sectors. It is required in the areas of: (i) engendering programs critical for programs to try to change these narratives more effectively in order to overcome some of with public campaigns, events, and discussions with the constraints stemming from social norms and community leaders. Access to networks and mentoring subjective preferences, (ii) supporting women is a crucial part of this, especially to overcome the lack entrepreneurs to cross over into higher value-added of information that constrains women entrepreneurs’ and more productive activities, and (iii) addressing decisions to enter into specific sectors and subsequent legal and institutional constraints. growth ambitions. Given the significant growth of female Along with these efforts, support programs also entrepreneurship programs targeting various groups need to engage with government institutions and such as the self-employed, it is important to better chambers of commerce to reform legal frameworks understand whether entrepreneurship support is likely that constrain female entrepreneurial activity. This to be more effective in increasing incomes for these requires robust assessment of existing legal constraints groups than alternative programs to engage them and the creation of forums for dialogue between in labor market activities. Relatedly, there are some sectors, government, and civil society on female aspects of delivering services to women entrepreneurs entrepreneurship to encourage reforms. that the note has not touched upon due to an exclusive focus on program design. Primary among these is the question of scaling up of these programs 5. CONCLUSION in order to have a real and sustainable impact on Women entrepreneurs are now widely recognized the lives of millions of potential and practicing as critical engines of national, regional, and global entrepreneurs. Some of the components that appear growth. It is therefore important to design appropriate essential in the success of women entrepreneurs support programs to facilitate their move into higher- cannot uniformly be delivered to millions of growth and higher-productivity and activities. women as they are currently implemented. For this, technology offers an immensely valuable tool to scale A review of the evidence on the entrepreneurial up these programs—now the task is to understand gender gap and of the impact evaluations of female how that might happen and test innovative ideas entrepreneurship programs revealed that support around it. programs have yielded positive outcomes in improving management practices but do not seem to be Existing approaches to supporting growth-oriented producing similar results in firm performance in terms women entrepreneurs are heterogeneous in their design of sales, profits, or employment. The review found and delivery, and yield indicative evidence that points that women-led enterprises underperform under the way forward. There is evidence to suggest several certain conditions which have to do with social norms, areas of focus which can facilitate the growth and perceptions about abilities and opportunities, and expansion of female-led enterprises, creating jobs and entrepreneurial preferences. boosting shared prosperity. Further experimentation and Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 17 evaluation are critical to developing a cohesive, holistic Berge, Lars Ivar Oppedal and Kjetil Bjorvatn and Bertil approach toward supporting growth-oriented women Tungodden (2012) “Human and financial capital for entrepreneurs and realizing their immense potential in microenterprise development: Short-term and long- stimulating economic transformation. term evidence from a field experiment in Tanzania”. Dartmouth University. REFERENCES ´, Mayra , Rebecca Furst-Nichols, and Emily Buvinic Acharya, Viral and Abhilasha Rajan and Antoinette Courey Pryor (2013) “A Roadmap For Promoting Schoar (2007) “What Determines Entrepreneurial Women’s Economic Empowerment” . http:// Success? A Psychometric Study of Rural www.womeneconroadmap.org/sites/default/files/ Entrepreneurs in India”. http://goo.gl/AIJyn5 WEE_Roadmap_Report_Final.pdf (accessed May 13, (accessed May 13, 2014) 2014). Brüderl, Josef and Peter Preisendörfer (1998) Aguirre, DeAnne, Leila Hoteit, Christine Rupp, and “Network Support and the Success of Newly Karim Sabbagh (2012) “Empowering the Third Founded Businesses” Small Business Economics 10: Billion. Women and the World of Work in 2012.” 213–225. Booz and Company. Bruhn, Miriam and Bilal Zia (2011) “Stimulating Aterido, Ryes and Mary Hallward-Driemeier (2011) managerial capital in emerging markets: The “Whose business is it anyway?.” Small Business impact of Business and Financial literacy for young Economics 37(4): 443–464. entrepreneurs”. The World Bank. Attanasio, Orazio, Adriana Kugler, and Costas Meghir Bruhn, Mariam and Dean Karlan and Antoinette (2011a) “Subsidizing Vocational Training for Schoar (2012) ”The Impact of Consulting Services Disadvantaged Youth in Colombia: Evidence from on Small and Medium Enterprises: Evidence from a a Randomized Trial.” American Economic Journal: Randomized Trial in Mexico,“ Working Papers 1010, Applied Economics, 3(3): 188–220. Economic Growth Center, Yale University. Attanasio, Orazio and Britta Augsburg and Ralph Campos, Francisco and Markus Goldstein and Laura de Haas and Emla Fitzsimons and Heike Harmgart McGorman and Ana Maria Munoz Boudet and Obert (2011b) “Group lending or individual lending? ´, et al. (2012) “Breaking the Pimhidzai in Buvinic Evidence from a Randomised Field experiment in metal ceiling: Female entrepreneurs who succeed in Mongolia”. IFS Working Papers W11/20, Institute for male-dominated sectors in Uganda”. Fiscal Studies. Calderon, Gabriela and Jesse M. Cunha and Bardasi, Elena and C. Mark Blackden and Juan Carlos Giacomo De Giorgi (2013) “Business Literacy Guzman (2007) “Gender, entrepreneurship, and and Development: Evidence from a Randomized competitiveness in Africa”. In World Economic Controlled Trial in Rural Mexico,” NBER Working Forum, World Bank, and African Development Bank. Papers 19740, National Bureau of Economic Africa Competitiveness Report, 69–85. Research, Inc. Bardasi, Elena and Shwetlena Sabarwal and Katherine Cho, Yoonyoung and Maddalena Honorati (2013) Terrell (2011) “How do Female Entrepreneurs “Entrepreneurship programs in developing countries: Perform? Evidence from Three Developing Regions”. a meta regression analysis,” Social Protection Small Business Economics, 37(4), 417–441. Discussion Papers 77168, The World Bank. 18 Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 Cohoon, J. McGrath and Vivek Wadhwa and is capital enough to get female microenterprises Lesa Mitchell (2010) “Are Successful Women growing? Evidence from a randomized experiment Entrepreneurs Different from Men?”. Available at in Ghana,” NBER Working Papers 17207, National SSRN: http://ssrn.com/abstract=1604653 or http:// Bureau of Economic Research, Inc. dx.doi.org/10.2139/ssrn.1604653 Fafchamps, Marcel, David McKenzie, Simono Quinn, Croson, Rachel and Uri Gneezy (2009) “Gender and Christopher Woodruff (2014) “”Microenterprise Differences in Preferences,” Journal of Economic Growth and the Flypaper Effect: Evidence from Literature, American Economic Association, vol. a Randomized Experiment in Ghana”. Journal of 47(2), pages 448–74, June. Development Economics, vol. 106(C), pages 211–226. de Mel, Suresh, David McKenzie, and Christopher Georgellis, Yannis and Howard Wall (2005) “Gender Woodruff (2008) “Are Women More Credit differences in self-employment”. International Constrained? Experimental Evidence on Gender Review of Applied Economics 19: 321–337. and Microenterprise Returns.” American Economic Journal: Applied Economics, 1(3): 1–32. Hallward-Driemeier, Mary (2013) Enterprising Women: Expanding Economic Opportunities in Africa The de Mel, Suresh David McKenzie, Christopher Woodruff World Bank (2014) “Business training and female enterprise start- Kalleberg, Arne L. and Kevin T. Leicht (1991) “Gender up, growth, and dynamics: Experimental evidence and Organizational Performance: Determinants of from Sri Lanka”. Journal of Development Economics, Small Business Survivial and Success” Academy of 106(1), 199–210, ISSN 0304-3878, http://dx.doi. Management Journal, 34(1): 136–161. org/10.1016/j.jdeveco.2013.09.005. Karlan, Dean and Martin Valdivia (2011) “Teaching Dey, Nepal C. in Buvinic ´, et al. (2012) “Case Studies Entrepreneurship: Impact of Business Training on and Best Practice Research on Empowering Women Microfinance Clients and Institutions,” The Review of through Clean Cookstoves Projects in Bangladesh”. Economics and Statistics, MIT Press, vol. 93(2), pages 510–527, May. Djankov, Simeon and Yingyi Qian and Gérard Roland and Ekaterina Zhuravskaya (2007) “What Makes a Klinger, Bailey and Matthias Schündeln (2011) Successful Entrepreneur? Evidence from Brazil” http:// “Can Entrepreneurial Activity be Taught? Quasi- eml.berkeley.edu/~groland/pubs/brazilent0907.pdf Experimental Evidence from Central America,” World Development, Elsevier, vol. 39(9), pages 1592–1610, Drexler, Alejandro (2014) “Learning Exercise: High- September. Potential Women Entrepreneurship Training Overview of Existing Programs”. The World Bank. Koellinger, Philipp and Maria Minniti and Christian Schade (2007) “I think I can, I think I can … : A study Fairlie, Robert and Alicia Robb (2009) “Gender of entrepreneurial behavior”. Journal of Economic differences in business performance: evidence from Psychology, 28: 502–527. the Characteristics of Business Owners survey,” Small Business Economics, Springer, vol. 33(4), pages Koellinger, Philipp and Maria Minniti and 375–395, December. Christian Schade (2013) “Gender Differences in Entrepreneurial Propensity”. Oxford Bulletin of Fafchamps, Marcel and David McKenzie and Simon Economics and Statistics, 75: 213–234. doi: 10.1111/ R. Quinn and Christopher Woodruff (2011) “When j.1468-0084.2011.00689. Supporting Growth-Oriented Women Entrepreneurs: A Review of the Evidence and Key Challenges – September 2014 19 McKenzie, David and Christopher Woodruff (2012) than men? Evidence from Madagascar”. http://goo. “What are we learning from business training gl/0K1DNc and entrepreneurship evaluations around the developing world?”. The World Bank. http://dx.doi. Nichter, Simeon and Lara Goldmark (2009) “Small org/10.1596/1813-9450-6202 Firm Growth in Developing Countries”. World Development, 37(9), 1453–1464. Minniti, Maria (2010) “Entrepreneurship, poverty and institutions”. Manuscript. Field, Erica, Rohini Pande and Natalia Rigol (2014) “Social norms and female entrepreneurship in India” Minniti, Maria (2010) “Female Entrepreneurship forthcoming and Economic Activity”. European Journal of Development Research, 22, 294–312. doi:10.1057/ Robb, A., Alexandria Valerio and B. Parton (2014) ejdr.2010.18. “Entrepreneurship Education and Training: Insights from Ghana, Kenya and Mozambique”. Washington, Mulcahy, Diane and Bill Weeks and Harold Bradley DC. World Bank (Forthcoming May 2014). (2012) “We Have Met the Enemy…and He is Us: Lessons from Twenty Years of the Kauffman Valdivia, Martin (2011) “Training or technical Foundation’s Investments in Venture Capital Funds assistance for female entrepreneurship? Evidence and the Triumph of Hope Over Experience”. Available from a field experiment in Peru”. The World Bank. at SSRN: http://ssrn.com/abstract=2053258 or http:// http://siteresources.worldbank.org/INTGENDER/ dx.doi.org/10.2139/ssrn.2053258 Resources/336003-1303333954789/final_report_ bustraining_BM_march31.pdf Nordman, Christophe Jalil and Julia Vaillant (2014) “Inputs, Gender Roles or Sharing Norms? Assessing Valerio, Alexandria and Brent Parton and Alicia the Gender Performance Gap Among Informal Robb (2013) “Framing the global landscape of Entrepreneurs in Madagascar,” IZA Discussion Papers entrepreneurship education and training programs”. 8046, Institute for the Study of Labor (IZA). Washington DC: World Bank. http://documents. worldbank.org/curated/en/2013/06/18347783/ Nordman, Christophe J. and Julia Vaillant (2013) “Why framing-global-landscape-entrepreneurship- are female informal entrepreneurs less successful education-training-programs 20 www.worldbank.org 20