The impact of gender on SME characteristics and access to debt finance in South Africa

Abstract

This study aimed to determine whether South African small and medium enterprises (SMEs) are affected by gender differences in demand for debt and its availability. It also looked at whether there are gender differences in the firm and entrepreneurial characteristics of SMEs. The study was conducted by means of a survey using a self-administered questionnaire and statistical analyses that included descriptive statistics, a t-test and a logistic regression. Significant gender differences were found in SMEs' demand for debt finance but only insignificant differences in availability. The findings also revealed significant gender differences in some of the firm and entrepreneurial characteristics of SMEs. It appears that for SME owners in South Africa gender differences exist in the demand for debt finance but not in its availability. The policy recommendation is that commercial banks, government agencies and non-governmental organisations should aim to help and encourage female SME owners to apply for debt finance.

Keywords:

• gender
• debt
• small and medium enterprises (SMEs)
• firm
• entrepreneurial characteristics

1. Introduction

South Africa suffers from high levels of unemployment, poverty and income inequality. Recent findings paint a dismal picture: there is a significant gender difference in unemployment (Okojie, 2003); the Gini coefficient is 0.58, making South Africa one of the most unequal countries in the world (Demombynes & Özler, 2005:267); approximately 57% of the population live below the poverty income line of US$2 per day (Agüero et al., 2007); labour force participation is gender-skewed, with the average rate being 34.2% for women and 52.1% for men; and the feminisation of poverty is becoming increasingly apparent (Shisana et al., 2010). According to Leibbrandt et al. (2010) and Posel & Rogan (2012), poverty estimates are significantly and consistently higher for women (especially African women) than for men. The unemployment rate, 25.5% at the time of writing (Stats SA, 2011), is one of the highest in the world.

The development of entrepreneurship and small and medium enterprises (SMEs) is important for addressing these problems in South Africa (O'Neill & Viljoen, 2001). Herrington et al. (2009) point out that, given the failure of the formal and public sector to absorb the growing number of job seekers in South Africa, attention has focused increasingly on entrepreneurship and SMEs and their potential for contributing to economic growth and job creation. According to the Department of Trade and Industry, women make up 52% of the population in South Africa (DTI, 2005). However, business ownership has traditionally been viewed as a male preserve where women have been involved in ventures but often only as silent or invisible partners. Mahadea (2001) notes that women entrepreneurs in South Africa remain on the periphery of the national economy. The female total entrepreneurial activity index shows that participation rates for men tend to be on average 50% higher than those for women (Minniti et al., 2005). According to Maas & Herrington (2006), globally the average for female entrepreneurship is 7.72% of the population. South African women are performing at just 4.83%. The findings of the Global Entrepreneurship Monitor are that South Africa's position in the global ranking continues to deteriorate relative to other countries participating, particularly when compared to similar developing economies. In addition, male owned or managed enterprises have a lower failure rate than their female counterparts in South Africa (Von Broembsen et al., 2005).

Ladzani & Van Vuuren (2002) observe that the White Paper on small business development in South Africa (RSA, 1995) made special reference to the development of female entrepreneurs. One specific objective of the national small business policy is to facilitate equalisation of income, wealth and economic opportunities, with special emphasis on supporting the advancement of women in all business sectors. This emphasis on the development of female entrepreneurs is understandable considering that women represent more than 50% of the South African population but own only about 33% of existing businesses (O'Neill & Viljoen, 2001).

However, despite their importance to both male and female entrepreneurs, 75% of SMEs in South Africa fail within the first two years of operation, one of the highest failure rates in the world. One reason for the failure of both male- and female-owned businesses is the non-availability of external finance (Gundry et al., 2002:69). Access to external finance is needed to reduce the impact of cash flow problems (Atieno, 2009:33). The availability of finance for investment in positive net present value projects is vital to the sustainability and viability of SMEs. The vast majority of SMEs depend on internal finance (contribution from the owners, family and friends), but this is often inadequate for them to survive and grow. A large percentage fail because of inadequate capital structure or shortage of resources. Carpenter & Petersen (2002) find that the growth of SMEs is constrained by dependence on internal finance. In contrast, firms that make use of external funds exhibit growth rates far above what can be supported by internal finance. It is, however, increasingly difficult to keep costs within the constraints of self-financing. Therefore SMEs, both male- and female-owned, need capital from external sources.

According to Demirguc-Kunt et al. (2006) the two primary sources of external finance for SMEs are equity and debt. External equity in the form of venture capital or the stock exchange is usually not available to SMEs. The lack of equity funding obliges many SMEs to depend on debt finance but unfortunately access to debt finance is very limited for SMEs in South Africa and other developing countries. Commercial banks hesitate to grant credit to SMEs. In addition, some SMEs do not ask for credit for fear that their applications will be rejected. These are referred to as ‘discouraged borrowers’ (Jappelli, 1990; Kon & Storey, 2001). According to Robb & Wolken (2002), firm and entrepreneurial characteristics of SMEs can affect their debt financing, having significant effects on the demand for and availability of credit for both female-owned and male-owned SMEs.

The literature mentions a lower participation rate of women in business ownership and a higher failure rate of female-owned enterprises (Minniti et al., 2005; Herrington et al., 2009). Herrington et al. (2009) point out that not being able to access finance is one of the primary causes of the low firm creation and high failure rate of SMEs in South Africa. The fact that there are fewer female-owned than male-owned SMEs and that the former fail more often than the latter suggests that women do not have equal access to external capital. Gender differences in firm and entrepreneurial characteristics may also affect the availability of debt finance.

The findings of previous studies on the impact of gender on the demand for and availability of credit are inconclusive. Freel et al. (2010) and Mama & Ewoudou (2010) find that gender is not a factor in the availability of credit and discouragement about obtaining it. Roper & Scott (2009) and Mijid (2009), however, find that gender is a factor in the demand for and availability of credit. Furthermore, an analysis of the literature on the financing of SMEs in South Africa revealed that no study has investigated empirically whether there are gender differences in the demand for and availability of debt finance in this country, nor did related studies, such as DTI (2005) and Mashigo (2006), on South African women entrepreneurs empirically investigate discouragement. Given the inconclusive evidence on gender and discouragement, the gaps in the literature and the dearth of empirical studies in South Africa, the objectives of this study were to investigate empirically whether female-owned SMEs are significantly different from male-owned SMEs in terms of 1) credit demand and availability, and 2) firm and entrepreneurial characteristics. In other words, the study aimed to establish whether female entrepreneurs in South Africa are more prone to discouragement when demanding credit and whether they suffer gender discrimination in the availability of debt finance.

The findings of this study offer insights into the characteristics of female compared to male entrepreneurs and thus help to identify where women actively participate in the South African economy and where they are less active. This could help government agencies and non-governmental organisations (NGOs) to tailor policies to increase women's participation in the economy and empower them by improving their incomes. The study provides insights into the demand for and availability of credit to female entrepreneurs. The evidence about whether female entrepreneurs in South Africa are prone to discouragement, and about the firm and entrepreneurial characteristics that can cause discouragement, could help commercial banks to develop products specifically designed to give female entrepreneurs better access to finance and help them improve their credit worthiness. The findings of the study could also help government to identify focus areas for improving female entrepreneurs' access to credit. Finally, the study makes a significant contribution to the literature on gender entrepreneurship and access to finance.

2. Literature review

Modigliani & Miller's (1963) theory of capital structure states that firms will select the mix of debt and equity that minimises their weighted average cost of capital. Because interest expense is tax deductible, debt is less costly than equity as a source of capital. In principle, therefore, firms act to minimise the cost of capital and maximise the value of the firm by financing with debt. Myers (1984:575) proposes a modified version of the ‘pecking order’ theory of finance which states that firms use internally generated funds in the form of retained earnings before turning to external sources. Both these capital structure theories suggest that SMEs use external debt before external equity. According to Coleman & Cohn (1999), for many SMEs the decision to finance with debt rather than equity may be driven at least as much by necessity as by choice, because SMEs do not have the same access to capital that larger public firms do. External equity finance in the form of venture capital or a listing on the stock exchange is generally unavailable to SMEs. Frelinghaus et al. (2005) observe that for SMEs internal equity is inadequate and external equity is unavailable, hence their reliance on debt finance even though it is not readily available to them, especially in developing countries. Stiglitz & Weiss (1981:395) call this phenomenon ‘capital rationing’.

2.1 Demand for credit

In his seminal work (1990), Jappelli recognises that credit constraint may be related not only to non-availability of credit but also to failure to demand credit. ‘Discouraged borrowers’, i.e. business owners with a high probability of being denied a loan, may not apply because they perceive that if they do they will be refused (1990:219). Kon & Storey (2001) also point out that some good borrowers do not apply for a bank loan because they feel they will be rejected: they may offer perfectly reasonable business proposals but fail to apply for a bank loan.

Levenson & Willard (2000:85) argue that an analysis of credit denial is likely to be biased towards finding too small an incidence of being denied if one does not account for failure to apply for credit. They find that more than twice as many small firms are discouraged as are rejected for loans by financial institutions in the US, which implies that discouragement is more important than credit restrictions of the Stiglitz-Weiss type. Freel et al. (2010:1) posit that if the extent of discouragement is indeed large or significantly larger than rejection, then addressing the fears of discouraged borrowers may be a more appropriate means of intervention than traditional supply-side mechanisms. Han et al. (2009:416), however, challenge this view by showing that discouraged borrowers are usually riskier applicants. In that case, discouragement would be an efficient self-rationing mechanism. Risky borrowers have a small chance of being approved, and therefore do not incur the cost of applying in the first place.

Is gender a factor in discouragement? Freel et al. (2010:415) found that twice as many businesses were discouraged from applying for a bank loan than had their loan request denied in the UK. They identified several distinguishing characteristics of discouraged borrowers – strategy, sector, prior entrepreneurial experience and banking relationships – but did not find that gender was a discouraging factor. Mama & Ewoudou (2010) also did not find any significant gender effect. Watson et al. (2009:43) conclude that there is also no evidence that Australian SME owners (particularly female owners) are being discouraged from applying for loans from a financial institution because they believe their application will be rejected. The findings of these three studies suggest that other demand-side issues (particularly risk-taking propensity and desire to maintain control) play a more important role in SME owners' capital structure decision-making.

Roper & Scott (2009:149), however, find that women's start-up rates in the UK are being reduced by two factors: the perception among the general female population that they face tougher financial barriers to start-up, and women's unwillingness to seek external finance for business start-up. Mijid (2009) finds that in the US women and minority-owned firms have higher loan denial rates and lower application rates than their male and white-owned counterparts. The mean probability of applying for loans is 32.2% for women and 44.6% for men. This indicates that women apply for a loan at a 12.4% lower rate than men do. These findings are consistent with findings of other studies, such as those by Midill et al. (2006) and Brooksbank et al. (2007), that gender is a discouraging factor for credit demand by SMEs. In the present study we therefore hypothesised that there are gender differences in the demand for debt by SMEs.

2.2 Availability of credit

According to Stiglitz & Weiss's formulation (1981), credit rationing is said to occur if 1) among loan applicants who appear to be identical, some receive credit while others do not, or 2) there are identifiable groups in the population that are unable to obtain credit or can only obtain credit at much higher prices. The Stiglitz and Weiss theory therefore suggests that there are significant numbers of new SMEs that could use funds productively if they were available, but cannot obtain finance from the formal financial system.

Constantinidis et al. (2006) find a large difference between male and female start-up funding. In total, men used three times more start-up capital than women. They found pronounced gender differences in financial arrangements for the ongoing business. Women, while similar to men in the use of personal finance, were less likely to use institutional finance such as overdrafts, bank loans and supplier credit, suggesting credit constraint. In addition, female-owned businesses tend to start up with lower levels of overall capitalisation, have lower ratios of debt finance and a much lower likelihood of using private equity or venture capital (Brush et al. 2002:309).

In contrast, Coleman & Cohn (1999:12) find that the primary determinants of leverage are firm size, firm age and profitability. They say that they found ‘no significant differences in the usage of debt between men and women, and gender was not a significant predictor of financial leverage’. They say this suggests ‘that women do have equal access to sources of debt and that they use them’ and that ‘they are not, as has been suggested, more risk averse than men, nor are they victims of discrimination since they appear to have the same access to external loans’. Becker-Blease & Sohl (2010:1) suggest that women seek angel financing at rates substantially lower than those of men, but have an equal probability of receiving investment. Cole & Mehran (2008) say that female-owned firms are significantly more likely to be credit constrained because they are more likely to be discouraged from applying for credit, though not more likely to be denied credit when they do apply. Irvin & Scott (2006) find that ‘ethnic minority businesses, particularly black owner-managers, had the greatest problem raising finance and hence relied upon “bootstrapping” as a financing strategy’, but that there was no difference in refusal rates between men and women. Treichel & Scott (2006:53) note that ‘while there is a common belief that women-owned businesses face discrimination in securing loans from banks … after controlling for the firm size, years in business, industry, and other owner/firm characteristics, discrimination does not appear to exist’. This suggests that there is no gender discrimination in the availability of debt. We therefore hypothesised that gender differences do not exist in the availability of debt to SMEs.

2.3 Firm and entrepreneurial characteristics

Firm characteristics are traits or features specific to the firm which can affect the performance of the firm positively or negatively. Emphasising the role of firm characteristics has become an increasingly important consideration in the empirical studies examining firm performance. Robb & Wolken (2002) found that female-owned businesses were significantly smaller than male-owned businesses in terms of employment, assets and sales. Smaller businesses are considered to have greater difficulty in securing bank loans and they pay higher interest than larger businesses (Brau, 2002). This may have an adverse effect on the perceived capacity of women to service or to repay their loans, and so they may face greater difficulty in obtaining credit. Female-owned businesses are more likely to operate in retail and services than businesses owned by men. This may influence the need for credit and the type of credit used. Service-based businesses require little if any financing and retail businesses allow for the use of trade credit (Robb & Wolken, 2002). Female entrepreneurs are over-represented in the retail and service sectors, particularly in personal services. Male entrepreneurs are over-represented in manufacturing, wholesale trade and financial services (Watson & Robinson, 2003).

Entrepreneurial characteristics are those traits or attributes that are specific to the owner of the firm which can affect the performance of the firm negatively or positively. This study investigated the effect of gender on entrepreneurial characteristics affecting the performance of SMEs, such as networking (membership of trade association) and education. Studies such as Blumberg & Letterie (2008) suggest that education improves managerial competency. Bowen et al. (2009) and Leitao & Franco (2011) find that high levels of owner education have a positive effect on the performance of SMEs. Highly educated SME owners are more likely to develop the skills required to start and sustain a business. Storey (2004) finds that a higher owner education reduces the incidence of denial of loans and increases application for loans. Blumberg & Letterie (2008) note that the business owner's potential earning capacity in a subsequent job (i.e. should the business venture fail) is a signal to the bank whether the small business owner will be able meet his or her credit obligations even if the business fails. Highly educated business owners are more likely to have a high post-failure earning capacity than less educated people.

Networks are the mechanism through which resources (information and capital) are introduced by specific agents into a particular social field (Schirato & Webb, 2003). They are important to entrepreneurship because they facilitate the efficient movement of these resources to entrepreneurs, and of information about entrepreneurs to a wider community. Networking activities are assumed to lower barriers when acquiring bank loans (Verheul & Thurik, 2001). Kepler & Shane (2007) found that men tend to have more business experience than women prior to opening a business but similar educational backgrounds. Green et al. (2010) found that women entrepreneurs tend to be slightly less educated than male entrepreneurs. Male entrepreneurs tend to network more with external parties such as chambers of commerce, while female networks are usually family and friends. Consequently, we hypothesised that there are differences between the firm and entrepreneurial characteristics of female-owned and male-owned SMEs.

3. Research methods

The empirical study focused on four cities in the Eastern Cape Province of South Africa: East London, Port Elizabeth, Queenstown and King Williams Town. Data were collected through a survey using a self-administered questionnaire comprising both closed-ended and Likert scale questions. A sizeable percentage of SMEs in the province are located in the four cities, and a list of 1936 identifiable SMEs was extracted from the telephone directories for these cities. The Raosoft sample size calculator (statistical software that enables researchers to determine the sample size given the variables margin of error, confidence level, population and expected response distribution) was used to calculate the required sample size (Raosoft Inc., 2010). For our study the minimum recommended sample size was 321. However, because of the disadvantages associated with the use of a self-administered questionnaire, such as possible weak response rate, we distributed 480 questionnaires – 312 to male SME owners and 168 to female SME owners. Of these questionnaires, 316 were returned (after repeated visits and calls to the respondents), made up of 199 from male owners and 117 from females.

The data were collected by the authors between January and October 2010, with the assistance of three paid fieldworkers, and analysed using descriptive statistics, a t-test and a logistic regression. For brevity, only the parameter estimates, marginal effects, standard errors and t-statistics for the primary variable of interest, female-owned SMEs, are presented in the logistic regression (see Table 3 below). Validity and reliability were ensured by using a statistician and a panel of experts to evaluate the research instrument for conceptual clarity, pre-testing the research instrument in a pilot study and comprehensively reviewing the literature for theoretical constructs and empirical conclusions. The pilot study, covering 40 respondents, enabled us to fine-tune the questionnaire. The Kolmogorov–Smirnov test was used to determine the normality of the data, and the case-wise deletion method to treat missing values. The number of employees was used to measure the size of the SME. To measure demand for credit, the respondents were asked whether their firm had needed credit in the past three years. Binary number one was used to represent firms that needed credit and zero to represent those that did not need it. To measure application for credit (measure of discouragement), one was used to identify firms that needed credit and applied for it and zero to represent those that did not apply for fear of rejection. If firms applied for credit, one was used to measure those that were successful in their application and zero to measure those that were unsuccessful (measure of availability of credit).

The financing of SMEs comes from many different sources, such as government and its associated agencies, NGOs, venture capital, angels, and the equity market. A limitation of this study that must be acknowledged is that it focused only on debt finance and was further limited to examining the demand side of debt financing and not the supply side.

4. Results

Biographical information (age of the business, age of the owner, education, etc.) and other descriptive statistics for the 312 respondents, and firm characteristics such as average number of employees, possession of collateral by the business, business sector and business status, can be seen in Table 1.

Table 1: Descriptive statistics (distribution of female and male respondents)

Variables	Male	Female
Questionnaires returned	199	117
Firm characteristics
Average number of employees	16	9
Have fixed asset as collateral	108	49
Manufacturing	68	25
Retail and wholesale	57	59
Services	74	33
Close corporation	116	49
Partnership	36	23
Sole proprietorship	47	45
Average age of firm	16	13
Entrepreneurial characteristics
Average age of owner	45	41
Experience	119	73
Degree/diploma	97	43
Membership of trade association	55	19
Credit
Need credit	155	91
Applied for bank loan	94	40
Did not apply for bank loan	61	51
Application successful	54	22
Application unsuccessful	40	19

4.1 Firm characteristics

The results of the t-test and the logistic regression (see Tables 2 and 3) indicate that female-owned SMEs in our sample were significantly smaller than male-owned SMEs. Female entrepreneurs were also significantly less likely to have collateral than male entrepreneurs. Empirical studies by Blumberg & Letterie (2008) and Abor (2008) find a positive relationship between the possession of collateral and the demand and availability of credit to SMEs. Collateral helps to reduce informational asymmetries and moral hazard problems that arise between banks and entrepreneurs.

Table 2: Mean and t-test results

Variable	All firms		Male-owned		Female-owned
	Mean	Standard error	Mean	Standard error	Mean	Standard error	t-statistic
Firm characteristics
Average number of people employed	13.895	0.342	16.146	0.349	9.243	0.241	5.74^a
Asset for collateral	0.597	0.08	0.688	0.369	0.427	0.269	3.48^c
Manufacturing	0.354	0.004	0.484	0.004	0.103	0.008	3.157^b
Retail and wholesale	0.129	0.007	0.080	0.008	0.222	0.016	–3.399^c
Services	0.281	0.004	0.442	0.005	0.547	0.007	4.393
Close corporation	0.177	0.005	0.184	0.009	0.149	0.018	2.899
Partnership	0.057	0.006	0.061	0.008	0.052	0.008	2.321
Sole proprietorship	0.394	0.008	0.401	0.04	0.543	0.07	4.277
Age of firm	14.174	0.175	14.653	0.341	12.231	0.09	3.201
Entrepreneurial characteristics
Membership of professional association	0.330	0.005	0.418	0.002	0.162	0.003	0.82^c
Degree	0.243	0.006	0.329	0.007	0.0216	0.013	4.07^b
Experience	0.644	0.125	0.679	0.003	0.616	0.018	3.69
Age of owner	47.288	0.174	47.459	0.185	45.267	0.236	4.22
Demand and availability of credit
Need for credit	0.825	0.025	0.767	0.377	0.721	0.024	5.275
Application for bank loan	0.632	0.005	0.771	0.007	0.308	0.002	3.514^b
Success of application (bank loan)	0.585	0.004	0.630	0.011	0.605	0.023	1.473
^aSignificant at 1% level. ^bSignificant at 5% level. ^cSignificant at 10% level.

Table 3: Multivariate differences in male-owned and female-owned SMEs (results of logistic regression)

Variable	Parameter estimate	Marginal effect	Standard error	t-statistic
Intercept	–0.680		0.388	–1.75^b
Firm characteristics
Average number of people employed	0.280	–0.031	0.087	3.22^b
Asset for collateral	0.140	–0.013	0.034	4.17^c
Manufacturing	0.860	0.008	0.247	3.50^b
Retail and wholesale	0.111	0.026	0.235	0.45^b
Services	–0.170	–0.005	0.187	–0.91
Close corporation	0.130	–0.018	0.127	1.03^c
Partnership	0.121	0.019	0.155	0.75
Sole proprietorship	0.090	0.013	0.089	0.97
Age of firm	0.010	–0.002	0.009	1.17
Entrepreneurial characteristics
Membership of professional association	0.041	–0.006	0.008	4.97^b
Degree	0.200	0.034	0.163	–1.20^c
Experience	0.057	0.042	0.022	2.59
Age of owner	0.820	0.057	0.311	2.65
Demand and availability of credit
Need credit	0.050	–0.065	0.117	0.45
Did not apply for bank loan	0.038	0.012	0.039	3.22^b
Denied bank loan	–0.066	–0.038	0.221	1.75
Notes: Female-owned = 1, male-owned = 0. ^aSignificant at 1% level. ^bSignificant at 5% level. ^cSignificant at 10% level.

Our results showed that the male-owned SMEs were mostly in manufacturing and the female-owned SMEs mostly in the retail and service sector. This is consistent with the findings of Watson & Robinson (2003), who note that male and female entrepreneurs tend to work in different sectors. Female entrepreneurs were over-represented in the retail and service sectors, particularly in personal services, while male entrepreneurs were over-represented in manufacturing, wholesale trade and financial services. The sector that a business was in had an effect on the demand for and availability of credit. Abor (2007) finds that service businesses are less likely to be candidates for bank loans because they often lack assets that can be used as collateral, while on the other hand businesses that are highly capital intensive, such as manufacturing, transportation and construction, may be more likely to use external capital.

Our results also showed that the male-owned SMEs were more likely to be organised as close corporations, while the female-owned SMEs were more likely to be organised as sole-proprietorships. Cassar (2004) notes that banks may perceive incorporation as a clear sign of the credibility and formality of operations. The difference in firm age was not found to be significant in our sample.

4.2 Entrepreneurial characteristics

Entrepreneurial characteristics considered were membership of professional associations, education to degree level and owner's age. The results in Tables 2 and 3 show that there were significant differences in membership of professional associations between the male and female business owners in our sample. Male entrepreneurs were more likely to join trade associations than their female counterparts. Not joining could affect the development of social capital by female entrepreneurs. Robb & Fairlie (2008) and Ngoc et al. (2009) find that networking plays an important role in spreading knowledge about a firm's existence and its practices. The female entrepreneurs were likely to be less educated than their male counterparts. Blumberg & Letterie (2008) point out that education can improve investment readiness and access to finance. The results did not indicate any significant difference between the ages of female and male entrepreneurs.

4.3 Demand for credit

Demand for credit is divided into two areas: need for credit and application for credit. The results shown in Tables 2 and 3 do not show any significance difference between female and male entrepreneurs' need for credit. The results did, however, indicate that the female entrepreneurs were significantly less likely to apply for credit than male entrepreneurs because of the fear of rejection. These results are consistent with the theory of discouraged borrowers (Jappelli, 1990; Kon & Storey, 2001) and the findings of empirical studies such as Cole & Mehran (2008), Roper & Scott (2009) and Mijid (2009).

4.4 Availability of credit

The results shown in Tables 2 and 3 indicate that there were no significant differences in the availability of debt to male and female entrepreneurs, and gender was not a significant predictor of financial leverage. This suggests that women do have equal access to sources of debt when they apply. The results are consistent with the findings of previous empirical studies on gender and finance such as Coleman & Cohn (1999) and Cole & Mehran (2008). The results indicate that our sample did not suffer from gender discrimination by banks when they applied for credit approval (the proxy for credit availability).

5. Conclusions and recommendations

The empirical findings of this study were that the female-owned SMEs were significantly smaller than the male-owned SMEs in terms of number of employees, more likely to be in retail business and more likely to be organised as sole proprietorships. Female owners were also likely to be less educated and also less likely to have collateral. The results indicate that there were gender differences in some but not all of the firm and entrepreneurial factors. It was also found that gender differences did exist in the SMEs' demand for debt finance, in line with the theory of discouragement (Jappelli, 1990; Kon & Storey, 2001). No gender discrimination was found in the availability of debt finance to SMEs. The results suggest that significant factors discouraging female entrepreneurs from applying for credit were lack of collateral, lack of business registration (a firm characteristic) and weak educational background (an entrepreneurial characteristic). Freel et al. (2010:1) note that if the extent of discouragement is indeed large or significantly larger than rejection, then addressing the causes of discouragement becomes very important.

On the basis of this study's findings, the authors recommend that the firm and entrepreneurial characteristics of female-owned SMEs be explored, with the aim of reducing discouragement by female entrepreneurs. Our recommendations focus on three main areas: female entrepreneurs, commercial banks, and government agencies and NGOs.

Female entrepreneurs: Without equity collateral, it is virtually impossible to get the required funding from commercial banks. Women entrepreneurs therefore need to have personal assets to be used as collateral. This could reduce their level of discouragement and improve their demand for credit. They also need to join trade associations and make contacts. Successful women entrepreneurs should help new women entrepreneurs to develop their skills through a mentorship system. Female SME owners also need to take greater responsibility for their own learning: they need to develop a positive attitude towards entrepreneurship and training to improve entrepreneurial education, competency and computer literacy. To get the required funding from commercial banks, they need first to become investment-ready. Women entrepreneurs should attend training organised by commercial banks, government agencies and NGOs to improve their entrepreneurial skills and loan application process.

Commercial banks: Commercial banks need to recognise women entrepreneurs as a distinct market opportunity and seek to understand their particular needs. The authors found that the websites of the four major South African commercial banks offer products and assistance on how to start, run and finance a business, including lending guidelines directed at all entrepreneurs (male and female). Awareness of the products that can benefit women entrepreneurs specifically could be promoted by commercial banks through print and electronic media, trade associations, government agencies and women-based NGOs. In addition, commercial banks should collect and report gender-disaggregated data on their loan clients. Commercial banks in South Africa could introduce a ‘competency-as-collateral’ scheme, as is done by some banks in India. Such a scheme integrates measures of management competency with the conventional credit risk assessment criteria. Female SMEs owners without collateral but who are educated and possess the required management competencies could be eligible for credit facilities. In addition, commercial banks in South Africa (through the creation of special banking units for female entrepreneurs) could adopt the model popularised by the Grameen Bank in Bangladesh, which provides micro-finance for women entrepreneurs, especially those without collateral. Financing institutions should disaggregate their portfolios and targets and put in place strategies that will help them to better understand and serve the women's market.

Government agencies and NGOs: By creating organisations such as the National Economic Empowerment Fund and agencies such as Khula Enterprises for loan guarantees, that can help female-owned SMEs without adequate collateral, the South African Government has focused on female entrepreneurship as part of the strategic plan for the country's economic growth. However, the existence of these organisations and their products must be publicised, to ensure that female SME owners are aware of their existence. Government agencies such as Statistics South Africa should collect gender disaggregated data on entrepreneurship. Government agencies and NGOs should organise gender mainstreaming workshops where government agencies, commercial banks, NGOs and female entrepreneurs can come together to discuss the financial challenges faced by women entrepreneurs. Organisations such as the South African Women Entrepreneurs Network (an arm of the Department of Trade and Industry), which is specifically dedicated to networking by female entrepreneurs, should have branches in all the major towns in South Africa.

Further studies could focus on the impact of gender on other sources of finance such as angels, venture capital, government and its associated agencies, and the stock market.