75559 Policy Brief Issue 1 | September 2011 Gender Gaps at the Enterprise Level: Evidence from South Africa Francisco Campos Summary Female-owned small to medium businesses in the Western Cape Province in South Africa are less productive, generate lower revenues and have less employees than male-owned enterprises. In this brief, we use the baseline survey for an impact evaluation of a business development services program to identify why these differences exist and explore paths towards policy interventions to overcome them. We conclude that the concentration of businesses in low performing sectors, the lack of commitment to the business, the intertwining of household and business responsibilities, and access to finance can be important barriers to the growth of women-headed enterprises. We suggest targeted alternative interventions to address these constraints and recommend comparing their effectiveness through rigorous evaluations. Emerging small to medium firms in Sub-Saharan Africa are As such, we are using a rigorous impact evaluation to study the believed to play an important role in reducing poverty. In South gender disaggregated effects of a government-led business Africa, where unemployment rates are as high as 25% (Stats development services (BDS) program in the Western Cape SA, 2010), small businesses are seen as drivers for Province. This initiative aims to address market failures sustainable and inclusive economic growth (DTI, 2003). Small, including lack of knowledge by small firm owners about the medium, and micro enterprises (SMMEs) represent more than availability of BDS opportunities, difficulty by businesses in 95% of the total number of firms and employ more than 50% of valuing the benefits of BDS before actually pursuing them, and the workforce (World Bank, 2007). However, they represent credit constraints. The BDS hosted by an individual only about 35% of GDP (DTI, 2003). Given the relevance of government agency include subsidized access to a set of SMMEs for employment in South Africa, identifying and service providers for technical skills upgrading, business addressing their constraints can be an important engine in the advice and marketing tools. country‟s development path. In particular, due to higher female unemployment and the fact that women are more likely to hire Using the baseline data for the impact evaluation of this other women, improving the competitiveness of female-owned intervention, we estimate the size of the existing gender gaps enterprises can help reduce the high overall unemployment in the performance of the businesses, provide potential rates in the country. justifications for those differences and recommend interventions to overcome them. We argue that the gender differences identified in the performance of SMMEs in this Province of South Africa can be due to a combination of (1) the concentration of women-entrepreneurs in a small number of low-performing sectors, (2) firms being seen by entrepreneurs as an interim solution, (3) the intertwining of household and enterprise money, and (4) credit constraints. © Hdtpcar/Flickr The World Bank Group | Africa Region Gender Practice Policy Brief: Issue 1 1 Figure 1: Gender gaps for SMMEs in the Western Cape Province Av. value added per Annual turnover worker EBITDA margin (%) Annual Number of (R th) (R th, annual) net profit (R th) employees 17 7.5 551 18 74 11 10 30 3.6 99 Female Male Female Male Female Male Female Male Female Male *EBITDA= Earnings before interest, taxes, depreciation and amortization. Top and bottom percentile trimmed for data on annual turnover, profit and EBITDA. LARGE GENDER GAPS IN FIRM-LEVEL PERFORMANCE When analyzing the concentration of businesses at a gender The impact evaluation design allowed us to use a disaggregated level, we detect that while women own representative sample of 906 SMMEs in the Western Cape enterprises in 35 sub-sectors, male are present in 49. More Province (47% female-owned) and identify the existing gender importantly, the five most important sub-industries for women – gaps. As per Figure 1, the gender gaps between female and the sectors where women mostly operate – account for 75% of male owned enterprises are very large for the most commonly women‟s activities distribution, compared to 60% for a similar used outcomes of interest: average labor productivity, number account for male-owned enterprises. While female business of employees, turnover, profits, operating margins and owners are highly concentrated in the retail, personal services proportion of formal enterprises. Female-owned firms are more and hotel and restaurants sectors, male entrepreneurs tend to likely to be smaller, informal, with lower productivity, with lower take part in other modes of business including agriculture, sophistication than male-owned enterprises. These gaps are manufacturing and transportation services. prevalent within sector, race, and level of education of the entrepreneurs. The problem with the limited “choices� of activities is that the productivity levels are significantly lower in the sectors where Nonetheless, once we account for a set of enterprise (sector, women are concentrated than in the remaining industries. In investment, etc) and business owner observable the activities where women operate their business, the initial characteristics (race, education, etc), the gender gaps found in investments are low, but the market potential is also limited. i the main outcomes of interest are substantially reduced – For instance, 85% of retail customers in the Western Cape are becoming statistically insignificant – and in some cases fully concentrated within a radius of 50 km, over 10 percentage disappear. This implies that if the right constraints are points more than for the remaining sectors. addressed, there is an opportunity for meaningfully reduce the differences in the performance of male and female enterprises. The female concentration in limited sectors could be due to specific constraints that women face. These include credit WHY WOMEN-OWNED SMMEs ARE LESS PRODUCTIVE constraints and the lack of knowledge about the benefits of Identifying these bottlenecks is hence crucial for addressing experimenting with new activities, which can be linked to the problems faced by women-entrepreneurs in South Africa. limited access to business skills. Theoretically, it could also be We identify four potential determinants, which combined, are due to the need to balance work with family life, although, as likely contributing to the lower productivity and size of women- discussed later in this brief, this seems to be a problem for only owned enterprises. a small proportion of female entrepreneurs. 1. Concentration of women entrepreneurs in a small number of 2. Enterprise as an interim solution low-performing sectors Secondly, firms in the Western Cape, especially women- One important factor for the lower performance of female owned, have very low levels of sophistication and initial capital entrepreneurs is the concentration of women in a few sectors, investment, which could indicate – in addition to credit usually with low levels of productivity, often using a limited constraints – a lack of commitment to the business. Banerjee number of managerial and technical tools, and regularly (2010) mentions that firm-level activity could be a way of targeting small markets. “buying� a job or, in other words, a means of remaining occupied with a firm instead of unemployed. Women may face the barrier of having no valid alternative to starting a business. Policy Brief Issue 1 | September 2011 enterprises. To assess sophistication levels, we segmented the firms in our sample into five groups in accordance with the PROFILE: degree of sophistication of the business (A to E, where A is the SMALL BUSINESSES IN THE WESTERN CAPE highest level of sophistication). In this segmentation, businesses were grouped by the equipment, tools and ii Over 85% of the SMMEs in the Western Cape resources used. Almost 80% of women-owned enterprises are are micro-enterprises (with less than 5 concentrated in Segments D and E (compared to 67% for employees) and approximately 45% of the firms male), where very limited sophistication seems to be the norm. are self-employed individuals (no workers). The A further illustration of the potential lack of commitment of median number of years of operation of SMMEs women to their current businesses is that for the lower in the Western Cape is three and approximately segments of sophistication within each sector, women invest 70% of the enterprises invested less than less than men. R10,000 ($1,500) to start the business. Furthermore, female entrepreneurs are more likely than male The main sector of activity for SMMEs is the retail sector, which represents 51%t of the (49% vs. 38%) to justify having started the business out of SMMEs in the Province. Other relevant sectors reasons of necessity such as "couldn't find a job" rather than include hair salons and other personal services having identified an opportunity such as “found a gap in the enterprises, followed by agriculture, and hotel market.� Women are also more willing to accept a formal job if and restaurants. Businesses in the retail sector offered one, notably within the group of self-employed (36% vs. are mostly operating as informal spaza shops, 30% for male). Furthermore, firms‟ death rates in South Africa taverns and „fruit & veg‟ stalls in townships and iii are high. There are strong indications that women start their suburban areas. businesses in the Western Cape without very large commitments, which naturally reduces their investment and Seventy one percent of the SMMEs in the potential for growth. Western Cape are not registered with CIPRO (Companies and Intellectual Property 3. Intertwining of household and enterprise money Registration Office). For female-owned A third related factor is that it seems that female enterprises businesses, the proportion of unregistered businesses is 80%, 16 percentage points above are more likely to intertwine household and business money, that of male-owned enterprises. Approximately which can negatively impact their financial management and 80% of the businesses report not paying taxes – hence their business competitiveness. an additional indicator of informality. We compare male and female investment in businesses and their degree of sophistication to understand whether men and women demonstrate different levels of commitment to the business. In terms of capital investment, more than 80% of the female entrepreneurs invested less than R10,000 ($1,500) to start the business, compared to 59% for male-owned © Hdtpcar/Flickr The World Bank Group | Africa Region Gender Practice Policy Brief: Issue 1 3 Figure 2: Proportion of SMMEs with access to the following tools and equipment 39 41 36 33 29 26 23 21 20 21 18 14 14 8 Business bank Insurance plan Business plan Written budget Email Company car Security system account Female Male While recent research suggests that general financial literacy The discussion on the importance of access to finance has knowledge is not a binding constraint for small entrepreneurs been vast but far from conclusive. On the one hand, micro- (Cole et al., 2009), experiments on access to financial services lending does not seem to address any major barrier to growth suggest that this can have large positive effects on firms‟ (Banerjee et al., 2009, for example), on the other, access to performance, particularly for women. For instance, Dupas and finance continues to be commonly identified as a key constraint Robinson (2009) find that simple savings accounts can for small business development (Beck, 2007. Additionally, significantly improve the welfare of self-employed female there are indications that expanding credit to new borrowers entrepreneurs while Ashraf et al. (2008) find that they can can have positive effects for individuals (Karlan and Zinman, increase women‟s household decision-making power. Although 2008). In spite of access to finance being advocated in all these studies were not necessarily addressing the effects of private sector development forums – in a recent World Bank separating business from household money, it is generally IEG workshop in South Africa, government delegations from a believed that there is value in having money solely used for number of African countries identified access to finance as the business purposes. These benefits include tracking revenue main aspect to focus on for women's economic empowerment. and costs accurately, managing working capital better, and Issues like cost of financing, informational problems and costs reducing the risk of lack of self-control on the usage of money. of screening have prevented more businesses from accessing iv This last factor was identified by Fafchamps et al. (2010) as a credit. key reason for why there were limited returns to grants given to entrepreneurs in Ghana, where 83% of women and 69% of While the rates of formal loans for both male and female men operate businesses from home. entrepreneurs are too low to draw any definitive conclusions in v the Western Cape, being well-financed is identified by women In the Western Cape, as per figure 2, only 26% of female- as a relatively more important factor than by men. Additionally, owned firms have a business bank account compared to 39% women rate loans as the most useful service for a development of male-owned enterprises and, concurrently, female agency to provide their business with while male entrepreneurs entrepreneurs are more likely to use their personal accounts for prefer business advice followed by legal support. business (70% vs. 59% of male-owned). Furthermore, 79% of the female-owned SMMEs are managed from home (vs. 64% The combination of limited access to financial services, for male). These numbers indicate that women are more likely constrained opportunities for investment and barriers in to intertwine household and enterprise responsibilities, which accessing credit seem to help explain the lower productivity of can have important effects on the success of the business. women within all sectors of activities in the Western Cape. 4. Credit constraints PERHAPS NOT AS IMPORTANT Coupled with the limited number of activities where women Other factors could be raised as well to explain the differences operate and the low investments in the business, access to in productivity and size of the male and female-headed credit seems to help explain the differences between male and businesses in the Western Cape. In this section, we mention female-headed enterprises. Access to credit is two common explanations and why we believe they are not as disproportionally identified by women within both micro and relevant in this setting. medium enterprises as the main obstacle for the growth of their businesses. Policy Brief Issue 1 | September 2011 that are male-owned. Nonetheless, there is no indication that informality is a determinant for the lower performance of PROFILE: women-owned firms. It seems more the other way around: ENTREPRENEURS IN THE WESTERN CAPE small firms with low productivity cannot afford becoming formal. Usually firms do not register and do not pay taxes because Forty seven percent of the owners of SMMEs in they believe their business is too small. There is no difference the Western Cape are female. That compares between male and female owned-enterprises in terms of the with a 52% female population in the Province proportion of firms that claim they do not register due to the (Stats SA, 2009). Female entrepreneurs are process being too complicated or the costs associated with younger and less experienced than male registered being too high. Still, there may be a lack of business owners. vi knowledge about any potential positive benefits of formality. In terms of population group, 39% of the SMME entrepreneurs are African, 35% are Coloured and Work-life balance almost 24% are White. The latter are older than Another common explanation for why women-owned firms are entrepreneurs from other population groups. operated with low sophistication levels and a low productive nature is the need to balance the management of business with The average age of small business personal life. While the fact that they operate at home may entrepreneurs in the Western Cape is 41. This is indicate that this factor seems to matter, the analysis of the older than the age range of the adult population baseline data in the Western Cape indicates that women and in the Western Cape. Although 50% of the adult men dedicate the same amount of time to the business population in the Province is between 15 and 34 (approximately 45 hours on average per week). As discussed years old (Stats SA, 2009), only 29percent of the above, women are also more willing to accept a formal job – by SMMEs‟ business owners are in that age group. definition outside the home – if offered one. The work-life balance could be a constraint though for a sub-group of More than 96% of SMME owners in the Western women-owned entrepreneurs with children and elderly care Cape have completed primary school and half of them have completed at least high-school (with responsibilities and without much family support (over 40% of Matric or above). Forty eight percent of female women entrepreneurs have young children and operate from vii business proprietors have completed high home, and out of these, approximately 20% are not married). school, which compares to 54% for male entrepreneurs. Approximately 8% of SMME POLICY IMPLICATIONS owners have a university degree. In addition to In sum, the combined effect of limited commitment to the the formal education received, 22% of the business, credit constraints, lack of variety in type and size of entrepreneurs report having completed other business, and difficulties in managing resources can deter the training, and around 50% of this additional success of women entrepreneurs in the Western Cape training was of at least one-year long.. Province. However, this is not sufficient to rigorously determine which barrier is most binding. These constraints should be tested through pilot experiments, ideally by comparing different alternative programs to overcome them. Cost-benefit analyses Informality of distinct interventions will allow for the identification of the Informality, defined as not being registered as a business or most important obstacles faced by women-owned not paying corporate taxes, is greatly associated with lower entrepreneurs in improving their productivity, as well as business performance in the Western Cape. Eighty percent of ascertain which policy initiatives are worth scaling-up. female-owned firms are not registered versus 64% of those The World Bank Group | Africa Region Gender Practice Policy Brief: Issue 1 5 One option to address the first constraint concerning the concentration of women in a limited number of business activities with low productivity levels is to establish a risk- sharing mechanism like a fund dedicated to investing in new sectors of activities or in methods of re-targeting current business activities. The objective is to reduce the risk for entrepreneurs and limit the need for outside capital. This idea could mean, for example, supporting a caterer to reposition her business and start servicing hotels through capital investment. This fund could also be focused on supporting entrepreneurs so that they try to penetrate unexplored industries in their market. In terms of the enterprise being an interim solution, it is important to explore alternatives to household- entrepreneurship as the apparent „single‟ solution to the unemployment problem. As an increased number of people initiate self-employed activities due to lack of alternatives (Fox, 2011) with the increased risk of over-supplying the market, one solution, as per Gelb et al. (2009), is to expand formal job opportunities in the more-established segments of the economy through initiatives like temporary wage subsidies. This is about to be launched in South Africa in 2012 (MoF Budget Speech, 2011) under a new large youth wage subsidy. An alternative to this would be training programs to teach new technical skills. Aimed at reducing the strong interconnection between household and enterprise usage of money, one option is a training program on the benefits of separating resources (addressing the potential information problem). A second, perhaps complementary, intervention is a financial incentive to open a business-specific bank account (addressing the eventual cost problem), which could drive a change in behavior, and spur improved management of enterprise resources. To provide increased opportunities to entrepreneurs in accessing finance – not only microfinance but larger scale loans – one possibility is to implement new methods of screening. An example is an entrepreneur-level automated test that replaces credit history, need for detailed business plan, and need for collateral. Incentives like credit guarantees for banks to reach new entrepreneurs could also be useful. Testing these different options through pilot interventions will help shed light on the most important barriers faced by women- owned firms and contribute to finding sustainable solutions that will promote the development of small businesses in South Africa and other countries in the region. Policy Brief Issue 1 | September 2011 i Fifty six percent of women operate in the retail sector. ii The segmentation was produced using a range of information including Private Sector Development & Gender in Africa whether the SMMEs have access to electricity, telephone, Internet, bank accounts, budgets, among a number of other variables. The segmentation This brief is part of on-going analytical work on process was conducted from top to bottom, grouping the businesses in gender and private sector development by the Africa accordance with the tools and equipment available. Region Gender Practice seeking to identify gaps that iii Some references suggest numbers as high as 90% of SMMEs die in the first persist across gender in this sector and provide two years of operation. rigorous evidence on effective interventions to reduce iv The share of new enterprises with loans in Africa is less than 10%, those gaps. Working in partnership with governments according to Hallward-Driemeier (2010). and the Africa Region FPD Unit, the interventions v This led us to change the access to finance modules in other studies we are under study – matching grant, technical and conducting in South Africa and other countries in order to improve the entrepreneurial skills training, transition to formality, measurement of informal mechanisms of accessing credit. A similar financial literacy, microfinance, and networking conclusion was taken by Collins et al. in the “Portfolios of the Poor� book platform - aim to address specific barriers to (2009). development faced by businesses and their vi We are conducting an impact evaluation in Malawi on the value (benefits- entrepreneurs, particularly by women, including credit costs) disaggregated by gender for enterprises of becoming formal. constraints, limited access to markets, poor access to vii Women with young children not married and operating from home (just 38 business services and to skills upgrading observations) work less 25% less than women in similar situation working opportunities, and restrictions in accessing outside the home (with only 13 observations) . technology. The impact evaluations under this work program include on-going studies in DRC, Ethiopia, Ghana, Malawi, Mozambique, Nigeria, Tanzania, Togo, South Africa, For questions and more information about the region‟s gender program, please contact Sabrina Roshan at sroshan@worldbank.org The World Bank 1818 H St. NW Washington, DC 20433 USA The World Bank Group | Africa Region Gender Practice Policy Brief: Issue 1 7