A e s e l fi fi © ( K C P 1 B ( h Available online at www.sciencedirect.com ScienceDirect Review of Development Finance 7 (2017) 18–28 Financing the growth of SMEs in Africa: What are the contraints to SME financing within ECOWAS? Peter Quartey a,∗, Ebo Turkson a, Joshua Y. Abor b, Abdul Malik Iddrisu a a Department of Economics, University of Ghana, Ghana b Business School, University of Ghana, Ghana bstract This study attempts to provide some understanding about SMEs’ access to finance within the West African sub-region with particular interest in stablishing whether there are similarities and/or differences in the determinants of SMEs access to finance across countries in SSA. For robustness ake, we developed both subjective and objective measures of access to finance. Using data from World Bank’s Enterprise Survey data set, we xamine the determinants of access to finance both at the sub-regional level and at the country-level. We found that, generally, at the sub-regional evel, access to finance is strongly determined by factors such as firm size, ownership, strength of legal rights, and depth of credit information, rm’s export orientation and the experience of the top manager. However, we found important differences in the correlates of firms’ access to nance at the country level. The findings of this study therefore have important implications for policy. 2017 Africagrowth Institute. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license http://creativecommons.org/licenses/by-nc-nd/4.0/).eywords: Finance; SMEs; Growth; Constraints; ECOWAS; Africa ontents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2. Overview of SME financing in SSA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3. Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 4. Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 4.1. Data and summary statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 5. Analysis and discussion of results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 5.1. Sub-regional analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 5.2. Country level analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 5.3. Robustness checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 6. Conclusions and policy implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Appendix A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 1. Introduction∗ Corresponding author. a eer review under responsibility of Africagrowth Institute. ( v c i 879-9337 © 2017 Africagrowth Institute. Production and hosting by Elsevier t .V. This is an open access article under the CC BY-NC-ND license c http://creativecommons.org/licenses/by-nc-nd/4.0/). p ttp://dx.doi.org/10.1016/j.rdf.2017.03.001 pA great deal of the development studies literature has focused ttention on the difficulties small and medium-sized enterprises SMEs) face in their daily operations. Generally, there is the iew that SMEs decry their level of access to finance and the ost involved in obtaining such funds where available. Obstacles ncluding those created by commercial or equity banks, institu- ional imperfections and SMEs themselves have been the main hallenges. While a lot has been done to address this reoccurring henomenon (see Abor and Quartey, 2010; Aryeetey, 1998), the roblem is still pervasive, requiring further scrutiny given the velop s w a s U i e t i t a i e a t e t S o S i a r r b o e l s S 1 s i r s c i o t a a p i i t p i a E m w 9 l c S s t s a a e c h t s p o A d t m i m o d 2 t A f ( i o a t E w Generally, within SSA, official schemes (where finance is v provided to SMEs by government and/or other international bod- f ies) and informal sources of finance remain the main alternatives N a a 1 This measure illustrates the distance of an economy to the “frontier”. The frontier represents the best performing economy observed with regards to the p ease of accessing credit across all economies and years included since 2005. An f t a P. Quartey et al. / Review of De prawling fields of economic resources waiting to be discovered nd tapped by SMEs. The need to finance the development and growth of SMEs n Sub-Saharan Africa (SSA) economies has been of concern o many policy makers for two main reasons. First, for SSA o be able to compete effectively in the increasingly global- zed environment, its micro and small enterprises should grow nd transform into thresholds where they will be able to adapt fficient production techniques. Indeed, the SME sector within SA has been referred to as the ‘Missing Middle’ in the context f financial inclusion or access to financial (including bank- ng) services. SMEs are by their nature unable to provide the equired collateral that large firms have in obtaining formal anking sector loans and at the same time too large to ben- fit from micro-finance loans and forms of financial support chemes (Abor and Quartey, 2010; Quartey, 2002; Aryeetey, 998). Secondly, because the small scale end of the manufactur- ng sector is a cornerstone to employment creation, micro, and mall enterprises within that sector should be seen to be growing nto thresholds that assure employment generation (by the adop- ion of labour intensive production), increased incomes/earnings nd, ultimately, poverty reduction and economic prosperity. This s achievable once the needed finance is available for investment o propel growth and employment. Against this background, this paper seeks to examine SME ccess to financing within the West African sub-region. The otivation for this study is to add to the growing literature on hat determines SMEs access to finance in SSA, in the fol- owing ways: first, in spite of Africa’s specificities regarding ME financing, not many studies have been undertaken to ascer- ain why SMEs within SSA are considered riskier to lend to nd therefore more financially constrained. Besides, to a large xtent, most of the available studies, in Africa and elsewhere, ave employed subjective measures of access to finance. This tudy departs sharply from existing studies by employing an bjective measure of firms’ access to finance to investigate the eterminants of SMEs’ access to finance in SSA. We believe this easure would yield more meaningful results than the subjective easures. . Overview of SME financing in SSA The operations of small and medium-scale enterprises SMEs) occupy an admirable position in the economic landscape f most economies in the world, especially for developing coun- ries. It is estimated that more than 95% of enterprises across the orld are SMEs and they account for approximately 60% of pri- ate sector employment (Ayyagari et al., 2011). SMEs account or a greater share of businesses in South Africa, Ghana, and igeria (Abor and Quartey, 2010; Gbandi and Amissah, 2014) nd their contribution to GDP, in Ghana for example, stood at bout 49% in 2012 (PWC, 2013). The SME sector in West Africa has been exploding over the ast three decades mainly as a result of the few formal avenues e or pursuing interest-bearing investment options. Most coun- r ries within the sub-region have shallow stock markets, whilst o t the same time interest rates have not been able to catch up dment Finance 7 (2017) 18–28 19 ith rising inflation. Under these circumstances entrepreneur- hip has been the attraction for investing excess money holdings. nfortunately the SME sector in West Africa is a mixture of self- mployment outlets and dynamic enterprises that are involved n an array of activities mainly concentrated in urban areas. In ddition, they lack access to the financial resources needed to xpand, grow and transform into higher size thresholds. Ques- ions have been posed as to what might be the main reasons for his. According to Collier (2009) the lack of access to finance by MEs in Africa is unfortunately as a result of two high risk char- cteristics. First, the provision of finance for Africa is generally ated as riskier than for other regions. Second, the provision f finance for small firms is globally rated as riskier than for arge firms. The ensuing sections discuss some stylized facts on MEs within ECOWAS regarding access to finance and explore ome of the alternatives that have been deployed within the sub- egion by various countries to relieve SMEs of their financing onstraints. The World Bank’s distance to 1 the frontier index f 2014 shows that with the exception of Ghana and Nigeria, ll other West African countries were at best 60 percentage oints away from New Zealand (the best performing country n the world). While Ghana and Nigeria were about 35 and 40 ercentage points away from the frontier respectively, Ghana’s ndex was higher than the average for OECD high income and urope and Central Asian countries (World Bank WDI, 2015). The lack of access to credit by SMEs (who make up over 0% of the private sector) in the developing world is clearly onfirmed from data on bank and domestic credit to the private ector in West African countries, and among income groups and ub-regions of the world (as shown in Table 1) between 2000 nd 2014. With the exception of high income countries, banks ontribute over 90% of domestic credit to the private sector over he period 2000–2014. The proportion of domestic credit to the rivate sector contributed by banks was even higher for West frican countries than other countries. Noticeably in almost all he countries within the sub-region there has been an increase n both domestic and bank credit to the private sector (as % f GDP) over the period, with Cape Verde’s ratio more than oubling and Guinea remaining stable at very low levels of less han 6% of GDP. These stylized facts on credit to SMEs in West frica necessitated the need for a study of this sort to unpack the actors that explain why SMEs within West Africa have difficulty n accessing finance. To do this, we begin by providing anecdotal ccount of the various alternatives to SME financing within the COWAS sub-region.conomy’s distance to frontier is indicated on a scale from 0 to 100, where 0 epresents the lowest performance and 100 the frontier. The measure on the ease f getting credit is based on the sum of the strength of legal rights index and the epth of credit information index. 20 P. Quartey et al. / Review of Development Finance 7 (2017) 18–28 Table 1 Bank and domestic credit to the private sector (ECOWAS, income and country groups). Country/region/group Bank credit to private sector (% of GDP) Domestic credit to private sector (% of GDP) 2000 2005 2010 2014 2000 2005 2010 2014 Benin 11.6 16.3 23.3 25.1 11.6 16.3 23.3 25.1 Burkina Faso 11.6 16.4 17.3 28.0 11.6 16.5 17.3 28.0 Cape Verde 38.2 37.7 61.0 62.4 40.1 39.1 61.9 62.9 Cote d’Ivoire 14.9 13.0 16.5 20.3 15.1 13.2 16.6 20.3 Gambia, The 6.5 9.5 15.2 13.1 6.7 9.6 15.4 13.5 Ghana 13.8 15.4 14.6 15.8 14.0 15.5 15.3 16.8 Guinea 3.9 4.4 5.7 .. 4.0 5.7 5.7 .. Guinea-Bissau 4.6 1.1 6.1 12.1 4.6 1.1 6.1 12.1 Liberia 3.1 6.3 14.4 19.5 3.3 6.7 14.8 19.6 Nigeria 12.3 13.2 15.4 14.5 12.4 13.2 15.4 14.6 Senegal 18.6 23.2 25.6 33.3 18.7 23.2 25.7 33.3 Sierra Leone 2.0 3.3 7.8 4.8 2.1 3.4 7.8 4.8 Togo 16.0 17.5 22.8 34.1 16.0 17.5 22.8 34.1 Niger 4.8 6.8 12.3 14.2 4.8 6.8 12.3 14.2 Mali 16.4 17.2 18.0 24.4 16.5 17.2 18.0 24.4 Europe & Central Asia .. 96.0 106.3 93.6 .. 96.0 106.6 93.8 Latin America & Caribbean 24.7 24.0 36.7 46.7 25.7 25.5 40.0 52.0 Middle East & North Africa 38.6 42.3 50.7 45.1 38.8 42.8 51.0 45.2 S S t S T a i t i B i a t i t t a t a f t t t a g fi s fi 3 8 t K m E o s t a A n l l q t A c 2 S i t t i s 1 a t I a 2 ub-Saharan Africa 33.0 33.8 32.4 29.1 ource: World Bank WDI (2015). o banks and non-bank financial institutions financing of SMEs. he official schemes are often set up with the primary motive of ncreasing the flow of finance and credit to SMEs to enable them ncrease their operational capacities, productivity and compet- tiveness in the local and international markets. With regards o informal sources of finance, many SMEs in SSA (especially hose in Nigeria and Ghana) have made use of sources such s owner’s savings, money lenders, friends and relatives, credit nd savings associations, susu, etc. to finance their growth and ransformation. See Appendix Table A1 for the key players in he SME finance industry in Ghana. For most economies in SSA, there are comprehensive SME nance programmes that have been in place since the 1960s. . Literature As already noted, the development of SMEs is often under- ined by a number of factors, including inadequate finance, lack f managerial skills, equipment and technology, poor access o capital markets, among others (Steel and Webster, 1991; ryeetey et al., 1994; Gockel and Akoena, 2002). A cursory ook at the literature on SME development reveals that inade- uacy of funds significantly constrains SME development (see, ryeetey, 1994; World Bank, 1994; Parker et al., 1995; Arthur, 003; Mensah, 2004; Deakins et al., 2008; Okpara, 2011). A survey of the extant literature in Africa particularly reveals he existence of a finance gap in the SME sector (see, for nstance, Sowa et al., 1992, for Ghana, and Daniels and Ngwira, 993, for Malawi). Aryeetey et al. (1994), for example, reported hat 38% of the SMEs surveyed mention credit as a constraint. 2 t is also established that most SME loan applications in Africa o re not granted (see Osei et al., 1993; Dawson, 1993; and Bani, H 003). In this vein, Aryeetey (1998) observed that only half of57.3 62.1 55.4 29.2 MEs’ applications for formal finance such as bank loans have ny chance of success, and about two-thirds of loan applica- ions by microenterprises were likely to be unsuccessful while igsten et al. (2000) observed that about 90% of small firms re denied credit from the formal financial sector due to their nability to fulfill conditions such as collateral security. Consis- ent with this, Osei et al.’s (1993) showed that about 95% of he SMEs depend solely on personal resources and loans from riends and relatives. Using data on firms in six African coun- ries, Bigsten et al. (2003) found that among those firms which pplied for a loan, small firms had substantially worse chances of etting a loan from banks. Berg and Fuchs (2013) found that the hare of SME lending in the overall loan portfolios of banks in ve Sub-Saharan African countries varies between 5 and 20%. Hansen et al. (2012) showed that about 39.6%, 18.3% and .5% of small firms in Ghana, Kenya and South Africa respec- ively cited access to finance as a barrier to SME growth. untchev et al. (2012) used data from the World Bank’s nterprise Survey which covered 13,685 companies across 38 ub-Saharan countries to examine SMEs access to finance. The uthors observed a strong correlation between the size of a busi- ess and their access to credit, where smaller businesses are more ikely to be credit ‘constrained’ and thus depicting the difficul- ies faced by small business entrepreneurs in securing loans from ommercial sources. On the main sources of external finance for MEs, Kuntchev et al. (2012) found that of the small businesses n sub-Saharan Africa that obtained external financing, 6.3% ook the form of equity, 48.5% was formal external debt, 17.4% emi-formal financing and 27.8% informal financing. Gbandi nd Amissah (2014) maintained that finance contributes about 5% to the success of SMEs in Nigeria and that more than 70% f funds available to SMEs are from the informal finance sector. aselip et al. (2014) investigated the financing of energy SMEs velop i fi t p e i s e t i t A a p e fi c i c a d t t p t 4 n i fi S m S b v r t S t h m g u a a o f t a c t A i i e fi t s r E i a t b n ( e s b p t I a t i c l b e c t t ( a 1 a i p e y fi a fi C a fi S P. Quartey et al. / Review of De n Ghana and Senegal. They observed that the lack of access o affordable finance is the predominant, persistent, barrier to stablishing and scaling up a commercially viable energy SME ector. Beck and Cull’s (2014) study concluded that compared o other regions, enterprises in SSA are less likely to have a loan han those in other developing regions of the world. However, bor and Biekpe (2006) established that, most SME operators ossess no knowledge of, and do not make use of the various nancing initiatives available to the SME sector in Ghana. There s also the perception that most of the schemes are difficult to ccess and hence, alternative sources of financing available to he SME sector remain largely untapped — a result that can be artly attributed to the stringent eligibility criteria for accessing hese funds (Abor and Biekpe, 2006). With regards to the drivers of SMEs financial constraints, a umber of studies have investigated the causes of SSA SMEs’ nability to acquire adequate external finance (Buatsi, 2002; acerdoti, 2005; Ghandi and Amissah, 2014). For instance, acerdoti (2005) noted that the inability of SMEs in SSA to pro- ide adequate financial statements and quality collateral reduce heir chances of accessing finance from formal financial institu- ions. Also, the absence of credible credit reference bureaus in ost countries in SSA and its attendant effect of interest rates ndermine the chances of SMEs gaining access to finance (Bass nd Schrooten, 2005). Using a sample data of over 10,000 firms rom 80 countries, Beck et al. (2003) found that the size, age nd ownership of firms are the factors that explain and predict he obstacles that SMEs face. However, at the country level, nstitutional development is the only country characteristic that xplains the cross-country variations in the financial difficul- ies faced by firms (Beck et al., 2003). Silva and Carreira (2010) evealed that the size of a firm and its cash flow are highly signif- cant in influencing the firm’s financial constraints whereas the ge of the firm had no significant impact. A cross-country study y Schiffer and Weder (2001) revealed that the size of a firm is egatively related to the risk it poses to the lender. Beck et al. 2006), illustrate that larger, older firms and foreign-owned firms njoy increased access to finance. In addition, Beck et al. (2004) how that, in terms of access to external finance, small firms enefit disproportionally from higher levels of property rights rotection (see also Demirgüç-Kunt and Maksimovic, 1998). nvestigating creditor protection, Love and Mylenko (2003) find hat the presence of private credit registries in a country is asso- iated with lower financing constraints and a higher share of ank financing. Important empirical insights are also provided by the litera- ure studying finance as a barrier to firm entry. Bertrand et al. 2007) suggest that the banking reform in France during the 980s influenced product market competitiveness by increas- ng entry and exit of firms and lowering industry concentration, specially in bank-dependent industries. Guiso et al. (2004) anal- se variations in financial development across Italian provinces nd find that financial development enhances entrepreneurship. etorelli and Strahan (2004) show that increased competition w mong banks in the United States helped the creation of new a rms due to enhanced access to finance. Similarly, Black and m trahan (2002) employ US data and find that that entry of new Ament Finance 7 (2017) 18–28 21 rms increased following deregulation. Overall, these studies rovide evidence that financial development – through bank- ng competition – increases credit availability and enhances ntry and efficiency in the corporate sector, thus confirming the mportance of access to finance. Thus, from the above discussions, access to credit remains challenge in the respective African countries reviewed. How- ver, unpacking the reasons why SMEs are unable to access redit over time and comparatively across selected African ountries will provide enormous policy lessons for sustainable evelopment and poverty reduction and this forms the focus of he next section. . Methodology Within the financing constraints literature firm’s access to nance has normally been modeled based on financial state- ents. For SMEs in developing countries this is impossible ecause financial data is limited. This is so because SMEs are not equired to file detailed financial reports because most of these MEs do not raise equity or debt from public markets. What as been readily available and has been mostly used are aggre- ate measures of financial development. The problem with such ggregate measures is that they do not provide the distribution f financing among such firms. SMEs will have differing access o finance and therefore the use of aggregate measures of finan- ial depth masks the heterogeneity of firms’ access to finance. s noted by Claessens and Tzioumis (2006) ‘the only way to nvestigate firms’ problems accessing finance is through tailored rm-level surveys directly addressing the issue of financing con- traints’ (pg. 6). This motivated our choice of the World Bank nterprises survey data for our study. In order to achieve our research objective, we estimate func- ionally the following: Firm access to crediti,k = β0 + β1firm characteristicsi,k + β2country characteristicsk + εi,k (1) Eq. (1) shows a functional relationship between firm’s access o credit (on one hand) and, firm characteristics and country char- cteristics (on the other). In this equation, the firm characteristics s represented in a vector form comprising of firm’s attributes ike age of the firm, size of the firm, gender of top manager, and xperience of top manager as reported by firm i in country k. The ountry characteristics variable also represents a set of dummies hat controls for country-specific factors that are unobservable nd may influence a firm’s access to credit. The dependent vari- ble (firm access to credit) represents both specific and general roblems faced by firm i in country k. In line with the objective to investigate the determinants of rm’s access to finance, we have used four measures of access to nance from the enterprise survey data. For robustness checks, e have developed both subjective and objective measures of ccess to finance. Access to finance 1, 2 and 3 are subjective easures whilst access to finance 4 is an objective measure. ccess to finance 1 is a binary variable, which equals one if a 22 P. Quartey et al. / Review of Development Finance 7 (2017) 18–28 s size S fi A i l t r O ( a d o fi f a a m fi 1 4 t o a f m w c r fi i s 7 r 5 a c o t c o c r b r i 4 t d m c n a F h 1 7 Chart 1. Firm ource: World Bank (2015). rm indicates that access to finance is an obstacle and zero if it s not an obstacle. Access to finance 2 is derived from the top hree obstacles that the firms face (ranked in order of severity). ur measure of access to finance 2 is a relative measure taking value of 1 if the firm indicates access to finance as the third bstacle, 2 if it is the second and 3 if it is the number one obstacle aced by the firm. A firm that does not indicate access to finance s an obstacle has a zero value. Access to finance 3 is a relative easure of the severity of access to finance as an obstacle to the rm. The severity index takes a value zero if it is not an obstacle; if minor, 2 if moderate, 3 if major and 4 if very severe. These hree measures would be used for the robustness checks analysis. With regards to access to finance 4, we attempt to obtain an bjective measure of a firm’s access to finance which we derived rom the shares of internal and external financial resources of orking capital. It takes a value of zero if the firm uses internal esources to finance at least 75% of its working capital, 1 if nternal resources contribute between at least 50% but less than 5% of working capital, 2 if between at least 25% and below 0% and 4 if the internal resources are less than 25% of working apital. According to the pecking order theory, because the cost f financing increases with asymmetric information, when it omes to methods of raising financial resources for working apital, firms prefer financing that comes from internal funds, efore resorting to external finance in the form of debt, and ssuing new equity in that 2 order . Accordingly, the model containing finance 1 as the depen- ent variable is estimated using the binary probit model while the odels containing finance 2 and 3 are estimated using the multi- omial probit model. The last model which captures finance c 4 s the dependent is estimated using the ordered probit C model. s t 2 Source: Boundless (2014). “Pecking Order Consideration.” Boundless V inance. Boundless, 30 Dec. 2014. Retrieved 16 Apr. 2015 from ttps://www.boundless.com/finance/textbooks/192/capital-structure- m 3/capital-structure-considerations-102/pecking-order-consideration-440- i 966/. o distribution. lso, the estimations are conducted both at the sub-regional evel and at the country-level. Finally, we would discuss the esults from the estimations that made use of access to finance 4 which is an objective measure of access to finance) as the depen- ent variable. The other three subjective measures (access to nance 1, 2 and 3) will be used for robustness checks/sensitivity nalysis. .1. Data and summary statistics The Enterprise Surveys are conducted by the World Bank nd its partners across all geographic regions and cover small, edium, and large companies. The data is obtained from several ross-country surveys across several years with an ‘access to nance’ component. It also contains some questions regarding ources of funds for investments and collateral requirements that eveal substantial variation of financing practices across firms nd countries. Chart 1 shows the size distribution of firms in each country hat constitute our sample. Generally there is a preponderance f small firms in all the countries within the sub region and elatively fewer large firms. Cape Verde, Burkina Faso and Togo eported the highest percentage of medium and large firms of 9.3%, 43.4% and 42.6% respectively, while Guinea reported he least of 12.1%. Our study is based at the sub-regional level all the 15 member ountries of ECOWAS and at the country level on 6 member ountries including Ghana, Mali, Senegal, Gambia, Guinea, and ote D’Ivoire. The choice of these six countries is based on data uitability and the 2014 ranking of “getting credit” distance to he frontier index of the countries in West Africa in 2014. Strikingly, unlike most countries within the sub-region Cape erde has the least percentage of firms and were mostly icro/small-sized firms (50.7%), with Guinea Bissau account- ng for the most of about 88%. Table 2 reports summary statistics n a number of indicators relating to access to credit by sam- P. Quartey et al. / Review of Development Finance 7 (2017) 18–28 23 Table 2 Summary statistics on measures of access to credit within ECOWAS. Country Collateral as % Loans req. % of firms that % that use Bank % access to finance of loan collateral (%) Bank for WC a major obstacle B- Fasso 173.9 90.0 98.0 34.7 72.0 Benin 293.4 89.3 96.7 39.4 60.8 Cape Verde 157.4 93.2 92.2 45.5 39.2 Cote D’Ivoire 59.1 74.5 70.8 8.2 70.3 Gambia 192.2 86.2 72.4 14.4 40.8 Ghana 178.3 71.6 90.3 20.7 63.8 Guinea 95.0 53.8 53.4 2.7 60.1 Guinea Bissau 63.8 80.0 60.7 1.3 73.6 Liberia 53.3 73.3 66.4 17.3 35.4 Mali 176.1 63.5 79.5 13.6 55.8 Niger 200.9 83.1 96.6 35.9 51.0 Nigeria 129.0 80.2 – 4.9 56.0 Senegal 122.0 88.2 83.6 9.1 48.2 S T p b r t 5 b l fi t K s a s fi t a s d 5 fi i t fi t c b i o m l r 1 n e t c i fi c t a a t l n 5 t e a f F 2 b r ierra Leone 51.2 89.7 ogo 265.9 80.0 led firms operating within the various countries within the sub egion. . Analysis and discussion of results In this section, we discuss the determinants of firms’ access o finance using our objective measure of access to finance at the ub-regional level as well as at the country level. For robustness ake, we juxtapose these findings with additional estimations hat employ some subjective measures of access to finance. .1. Sub-regional analysis The results shown in Table A2 column 3 relate to access to nance 4 after controlling for sector of activity and country spe- ific effects. Generally, access to finance is significantly driven y factors such as firm age, firm size, ownership, experience f the top manager (at 10% level of significance), strength of egal rights, depth of credit information, firm performance (at 0% level of significance) and sector of activity. Table 3 shows stimates of the marginal effects after controlling for country haracteristics. We find that firm’s age significantly increases rm’s access to finance from external sources and this buttresses he fact that older firms are less likely to have difficulties in ccessing finance compared to newer firms. Specifically, an increase in the age of a firm improves the ikelihood that the firm would devote less than 75%3 of its inter- al resources for working capital, indicating improved access o external sources of financing working capital. Similarly, it is videnced that larger firms are significantly more likely to gain ccess to external finance relative to smaller firms in West Africa. or instance, relative to smaller firms, larger firms are about U 4.0%, 17.0% and 16.0% significantly more likely to devote a fi e 3 This statement should be interpreted with reference to the base group. The i ase group is 0, representing the case when a firm uses at least t 75% of its internal esources for working capital. w68.7 28.7 37.3 96.1 21.1 52.3 etween 50–75%, 25–50%, and less than 25% respectively of heir internal resources for working capital. This shows that size matters for access to external finance y SMEs in West Africa. This is in line with the findings that arger firms are more likely to gain access to finance than smaller rms (see Beck et al., 2006; Buatsi, 2002; Bigsten et al., 2003; untchev et al., 2012). Large firms do not face challenges in ccessing external finance because, unlike small firms, large rms have adequate collateral, able to prove creditworthiness, dequate credit history and a developed bank-borrower relation- hip necessary for securing external funding. Further, relative to omestic ownership, foreign ownership significantly improves rm’s ability to obtain external financing with foreign firms hav- ng about 4.0% more likelihood to use between 50 and 75% of heir internal resources as working capital relative to 75% of heir internal resources. Additionally, improvement in the strength of legal rights ncreases firm’s access to external financing, while improve- ent in the depth of credit information leads to about 0.4% eduction in the likelihood of firms to use 50–75% of their inter- al resources for working capital relative to using at least 75 of heir internal resources for working capital (see, Table 3). This mplies that, improvement in the depth of credit information onstrains firm’s ability to access external sources of financing nd so firms are forced to rely heavily on their internal resources o finance working capital. .2. Country level analysis In order to unmask important differences in the influential actors across countries, country specific analysis is conducted. sing access to finance 4 as a dependent variable, we find that ccess to finance is significantly influenced by factors such as rm age, firm size, gender of top manager, ownership, experi- nce of top manager, formality, and firm performance at least n one of the estimations as shown in Table A2. Specifically, he determinants of access to finance varies across countries: hereas in Ghana, access to finance is significantly influenced 24 P. Quartey et al. / Review of Development Finance 7 (2017) 18–28 Table 3 Pooled ordered logit results (marginal effects): access to finance 4 — ECOWAS. Variables Dependent variable: access to finance 4 Outcome (1) Outcome (2) Outcome (3) Firm age 0.003** 0.002** 0.002** (0.000) (0.000) (0.000) Firm size 0.024*** 0.017*** 0.016*** (0.005) (0.004) (0.003) Gender of top mgr. (fe) 0.016 0.012 0.012 (0.017) (0.014) (0.014) Foreign ownership *** *** *** 0.040 0.033 0.034 (0.009) (0.009) (0.010) Experience of top manager 0.000 0.000 0.000 (0.000) (0.000) (0.000) Formality status 0.023 0.014 0.013 (0.083) (0.047) (0.041) Exporter 0.002 0.002 0.002 (0.010) (0.007) (0.007) Strength of legal rights index 0.013*** 0.009*** 0.008*** (0.002) (0.001) (0.001) Depth of credit information index −0.004* −0.002* −0.002* (0.002) (0.001) (0.001) Firm performance −0.000 −0.000 −0.000 (0.000) (0.000) (0.000) Sector (service) −0.006 −0.004 −0.004 (0.006) (0.004) (0.004) Control for country effect Yes Yes Yes Observations 5065 5065 5065 D esourc b b a s p m m a c e p t c fi f c t c 2 b t v t a i t u l a m T fi fi f r s o a t i fi u f b r r t t l c fi ependent variable: access to finance 4 = 0 if firm uses at least 75% internal r elow 25%. Robust standard errors in parentheses. y factors such as firm size and formality, access to finance is ignificantly driven by; firm size, gender of top manager, and for- ality in Mali; ownership, experience of top manager, formality nd firm performance in Senegal; and firm age, ownership, and xperience of manager in Gambia. These variations in the impor- ance of the various covariates in explaining firms’ access to nance across the countries reflects, in particular, country spe- ific constraints in firms’ access to finance. Also, consistent with the work of Beck et al., 2006; Buatsi, 002; Bigsten et al., 2003; Kuntchev et al., 2012, we observe hat the older firms are significantly more likely to gain access o finance compared to newer firms in Gambia. For instance, an ncrease in the age of a firm improves the likelihood of the firm to se less than 25% of its internal resources for working capital by bout 0.5% relative to using at least 75% of its internal resources. his may mean that older firms have wider networks including nancial institutions such as banks and suppliers, making it easy or them to access credit for their operations. Also, firm size is ignificantly positively associated with the probability of having ccess to finance in Ghana and Mali. In Mali, for example, an ncrease in the size of a firm raises the probability of that firm sing less than 25% of its internal resources for working capital y a little over 3% relative to using at least 75% of its internal esources for working capital (see Table 4). The importance of he age and size variables in our model of access to finance is onsistent with the results obtained by Beck et al. (2003). f Strikingly, gender of top manager significantly determines r rm’s access to finance but only in Mali. This result is strikinges for working capital; = 1 if between 50–75%; = 2 if between 25–50%; = 3 if s it contradicts the conventional notion that women dispro- ortionately face barriers to financial access compared to their ale counterparts, thus requiring us to dig into what peculiar haracteristics of the Malian economy might help explain this henomenon. We explain this conundrum as follows: first, this ould be due to government policy measures to tailor its support or SMEs more in favor of women-owned businesses in order o meet the Millennium Development Goals (MDGs); second, ompared to men, Malian women are largely engaged in small usinesses that do not require huge investments, thus providing ery little incentive for them to solicit for external finance such s borrowing from banks and other financial institution. Given hese, it would be realized that female top managers are less ikely to face difficulties in accessing finance compared to their ale counterparts. In addition, relative to domestic-owned firms, foreign-owned rms are more likely to use less than 75% of their internal esources for working capital compared to using at least 75% f their internal resources for working capital in Senegal. Fur- her, an increase in the experience of a top manager increases a rm’s probability of using less than 75% of its internal resources or working capital relative to using at least 75% of its internal esources for working capital in Senegal. In particular, we find hat, an increase in the experience of top manager raises the like- ihood of a firm to use less than 25% of its internal resources or working capital relative to using at least 75% of its internal esources for working capital in Senegal by about 0.2% (Table 4). P. Quartey et al. / Review of Development Finance 7 (2017) 18–28 25 Table 4 Ordered logit country level regressions (marginal effects). Regressors Dependent variable: access to finance 4 Ghana Mali Senegal Gambia Outcome (1) Firm age 0.000 0.001 −0.001 −0.002 (0.001) (0.002) (0.001) (0.001) Firm size 0.044*** 0.115*** 0.008 −0.016 (0.014) (0.030) (0.023) (0.019) Gender of top mgr. (fe) −0.022 ** 0.172 (0.035) (0.068) Foreign ownership *** −0.017 0.030 0.082 −0.057 (0.034) (0.067) (0.016) (0.041) Experience of top manager 0.000 −0.001 0.003* 0.002 (0.001) (0.002) (0.001) (0.002) Formality status −0.067*** 0.196*** −0.038 (0.009) (0.020) (0.098) Firm performance 0.000 0.001 * 0.002 −0.001 (0.000) (0.001) (0.001) (0.001) Outcome (2) Age 0.000 0.000 * −0.000 0.005 (0.001) (0.000) (0.001) (0.003) Firm size 0.028*** 0.015** 0.005 0.043 (0.009) (0.006) (0.013) (0.044) Gender of top mgr. (fe) −0.013 0.035 (0.019) (0.026) Foreign ownership −0.010 0.004 0.077** 0.087** (0.019) (0.010) (0.033) (0.042) Experience of top manager 0.000 −0.000 0.001* −0.005 (0.001) (0.000) (0.001) (0.003) Formality status −0.101*** 0.021*** −0.109*** (0.018) (0.006) (0.023) Firm performance 0.000 0.000 0.001* 0.002 (0.000) (0.000) (0.001) (0.001) Outcome (3) Age 0.000 0.000 −0.001 0.005* (0.001) (0.001) (0.001) (0.003) Firm size 0.021*** 0.033*** 0.006 0.045 (0.007) (0.010) (0.017) (0.045) Gender of top mgr. (fe) −0.010 0.086 0.124* (0.014) (0.067) (0.070) Foreign ownership −0.008 0.009 0.139* (0.014) (0.022) (0.082) Experience of top manager 0.000 −0.000 0.002* −0.005* (0.001) (0.001) (0.001) (0.003) Formality status −0.102*** 0.041*** −0.334 (0.022) (0.009) (0.288) Firm performance 0.000 0.000 0.002* 0.002 (0.000) (0.000) (0.001) (0.001) Controlled for export status and sector of activity Yes Yes Yes Yes Observations 894 509 407 177 Dependent variable: access to finance 4 = 0 if firm uses at least 75% internal resources for WC; = 1 if between 50–75%; = 2 if between 25–50%; = 3 if below 25%. Robust standard errors in parentheses. *** p < 0.01. ** p < 0.05. r l fi w a r fi s I s * p < 0.1. Relative to non-formal firms, formal firms are significantly ess likely to use less than 75% of their internal resources for orking capital as against using at least 75% of their internal esources for working capital in Ghana, indicating a relative carcity of external financing options available to formal firms. n n contrast, however, it is realized that, in Mali, formal firms are r ignificantly more likely to devote less than 75% of their internal eesources as working capital compared to informal firms. This nding lends credence to the fact that formality is often seen as n important criterion to accessing external finance since formal rms are to a large extent able to meet loan requirements than on-formal firms. Firm performance is significantly positively elated to access to external finance but only in Senegal. For xample, an improvement in the performance of a firm raises 2 velop ( T o e 7 fi s 1 s 5 a t t f e p s s s e l a r a s i l s e o p o c i w o d t v a e f t l t m t w a i g t 6 d r s t a i b f c c fi i c w a c f s f s a o W s i w S e t e Appendix A s a c fl v c 6 P. Quartey et al. / Review of De about 0.2%) the likelihood of the firm to devote less than 25% f its internal resources for working capital compared to using 5% and above of its internal resources for working capital as hown in Table 4. .3. Robustness checks As a robustness check we estimated three different models of he logistic regressions to predict the likelihood of the regressors xplaining a firm’s access to finance. The three subjective mea- ures used are access to finance 1, 2 and 3. We did this both for the ub-regional level and the country level. 4 The results obtained , argely confirm what we found in our core model. At the sub- egional level, therefore, access to finance is driven by factors uch as larger firms, firms owned by foreigners and strength of egal rights while variables such as age, firm size, ownership, xperience of the top manager, depth of credit information, firm erformance and sector of activity attained statistical signifi- ance, at least in one of the equations, at the country level. Thus, hereas the sub-regional level analysis obscures some important ifferences in terms of the influence of the various explanatory ariables on access to finance, the county level analysis delin- ates these differential effects. For instance, it is only in Ghana hat the depth of credit information is an important factor linked o firm’s access to finance. This suggests that as lenders become ore aware of the credit history of potential borrowers, firms ith bad credit history find it difficult to obtain external finance, n particular, bank loans. . Conclusions and policy implications The overall objective of this study is to examine SMEs access o finance within the West African sub-region with particular nterest in establishing whether there are similarities and/or dif- erences in the determinants of SMEs’ access to finance across ountries in SSA. Specifically, the study pursued the follow- ng research goals: (1) explore alternatives to SME financing ithin the ECOWAS sub-region; (2) predict SME access to redit using firm specific characteristics and other controls for elected countries in the ECOWAS sub-region. In order to achieve the first research objective, we conducted n extensive review of financing options available to SMEs in est Africa. In most West African countries, long-term financ- ng in terms of equity capital is virtually non-existent for the ME sector. This makes debt financing the dominant channel hrough which SMEs can access funds. Unfortunately, how- ver, the SME sector in most countries within the ECOWAS ub-region faces serious constraints in accessing formal finance nd this problem is connected to factors such as SMEs’ lack of ollateral, difficulties in providing creditworthiness, small cash ows, inadequate credit history, high risk premiums, underde- eloped bank-borrower relationships and high transaction costs. 4 Due to space limitations, we do not report these estimations. However, they an be made available upon request.ment Finance 7 (2017) 18–28 o fill the finance gap of the SME sector, most economies, specially those in Africa have developed comprehensive SME nance programmes, many of which can be traced back to the 960s. Governments, over the years, have used interest rate sub- idies, directed lending, guarantee funds and a variety of other pproaches to fuel the financing needs of the SME sector but all o no avail. In line with the second objective of the study, we have used our measures of access to finance from the World Bank’s Enter- rise Survey data set. For robustness sake, we developed both ubjective and objective measures of access to finance. Also, the xamination of the determinants of access to finance is estimated t the sub-regional level as well as at the country-level. Gener- lly, at the sub-regional level, it is clear that access to finance s strongly determined by factors such as firm size, ownership, trength of legal rights, depth of credit information, firm’s export rientation and the experience of the top manager. Further, it is bserved that the use of different measures of access to finance, n particular, the use of subjective vis a vis objective measures f access to finance did not matter much for the importance of he correlates of access to finance in SSA. The results obtained t the country level, however, bring to bear some important dif- erences in what explains firms’ access to finance at the country evel. The findings of this study therefore have important implica- ions for policy: First, since formality strongly influences SMEs’ ccess to credit, the study suggests that SMEs in SSA must take iant steps towards formalisation in order to increase their poten- ial for accessing formal credits. Second, given that firm size and epth of credit information are key to accessing credit, the study ecommends that SMEs should join Business Associations and eek group credit schemes. Related to the above, these associ- tions should also promote credit information among potential orrowers as a way of reducing information asymmetry in the redit market. This would go a long way to boost creditor con- dence as bad borrowers will be clearly identified and refused redit. Third, the findings that ownership matters for the prob- bility of accessing credit imply that African SMEs can learn rom their foreign counterparts in order to position themselves or institutional borrowing. Finally, past accounts of government upported credit schemes to the SMEs sector indicates a seri- us lack of efficiency and transparency in the operation of such chemes. As a remedy, we maintain that future governments as ell as international donor interventions must be administered fficiently and transparently. P. Quartey et al. / Review of Development Finance 7 (2017) 18–28 27 Table A1 Pooled ordered logit results: ECOWAS sub-region. Regressors Dependent variable: Access to finance 3 (severity index) Access to finance 4 (pecking order) Firm age −0.00685** 0.000604* (0.00324) (0.000327) Firm size −0.418*** 0.187*** (0.0505) (0.0535) Gender of top manager (female = 1) −0.0190 0.0569 (0.130) (0.190) Ownership (foreign = 1) −0.514*** 0.487*** (0.0979) (0.120) Experience of top manager 0.00438 0.00642* (0.00369) (0.00351) Formality status −0.743 0.0172 (0.493) (0.745) Exporter 0.0320 0.0467 (0.0999) (0.103) Strength of legal rights *** index 0.0130 −0.286 (0.0559) (0.0840) Depth of credit information index −0.0121 0.0130 (0.0198) (0.0222) Firm performance −6.16e-05 0.00227* (0.00107) (0.00138) Sector of activity (services) −0.228*** −0.121** (0.0656) (0.0589) Controlled for country effects Yes Yes Observations 3428 5065 Dependent variable: access to finance 3 (severity of access to finance as an obstacle (1–4; 0 = no obstacle); access to finance 4 = 0 if firm uses at least 75% internal resources for working capital; = 1 if between 50–75%; = 2 if between 25–50%; = 3 if below 25%. Robust standard errors in parentheses. *** p < 0.01. ** p < 0.05. * p < 0.1. Table A2 Pooled ordered logit results: (access to finance 4) — country specific. Regressors Dependent variable: access to finance 4 Ghana Mali Senegal Gambia Firm age 0.000116 0.0114 −0.00724 ** 0.0407 (0.00917) (0.0157) (0.0142) (0.0207) Firm size *** *** 0.380 0.887 0.0881 0.360 (0.113) (0.223) (0.241) (0.359) Gender of top mgr. (F = 1) −0.188 1.290** (0.288) (0.632) Ownership ** ** (foreign = 1) −0.147 0.224 1.244 0.862 (0.278) (0.487) (0.497) (0.432) Experience of top mgr. 0.00191 −0.0110 0.0281* −0.0387* (0.00950) (0.0172) (0.0151) (0.0220) Formality status *** *** * −1.123 −12.69 −2.141 (0.174) (0.612) (1.207) Exporter 0.0235 0.0628 0.447 −0.448 (0.184) (0.316) (0.308) (0.477) Firm performance 0.000819 0.00394 0.0261** 0.0143 (0.00276) (0.00441) (0.0132) (0.0109) Sector of activity (services) −0.142 −0.0190 0.153 0.167 (0.146) (0.214) (0.218) (0.430) Observations 894 506 407 118 Dependent variable: access to finance 4 = 0 if firm uses at least 75% internal resources for working capital; = 1 if between 50–75%; = 2 if between 25–50%; = 3 if below 25% Robust standard errors in parentheses. *** p < 0.01. ** p < 0.05. * p < 0.1. 2 velop R D A A D A D A G G A A G H A H B K B B L B M B O B O B B P P B Q S B S B S /. S B C S C World Bank (1994). Industrial Structures and the Development of Small and Medium Enterprise Linkages: Examples from East Asia, Meyanathan S. (Ed.), EDI Series, World Bank. C World Bank (2015). World Development Indicators. D 8 P. Quartey et al. / Review of De eferences bor, J., Quartey, P., 2010. Issues in SME development in Ghana and South Africa. Int. Res. J. Finance Econ. 39, 218–228. bor, J., Biekpe, N., 2006. Small business financing initiatives in Ghana. Prob- lems Perspect. Manag. 4 (3), 69–77. rthur, P., 2003. The implications of state policy for micro-enterprise devel- opment. In: Tettey, Wisdom, Puplampu, Korbla, Berman, Bruce (Eds.) (2003). Critical Perspectives on Politics and Socio-Economic Development in Ghana. Leiden, 153–175. ryeetey, E., 1994. Supply and Demand for Finance of Small Enterprises in Ghana. World Bank Discussion Paper 251. The World Bank, Washington, D.C. ryeetey, E., 1998. Informal Finance for Private Sector Development in Africa. Economic Research Papers No. 41. The African Development Bank, Abid- jan. ryeetey, E., Baah-Nuakoh, A., Duggleby, T., Hettige, H., Steel, W.F., 1994. Supply and Demand for Finance of Small Scale Enterprises in Ghana. Dis- cussion Paper No. 251. World Bank, Washington, DC. yyagari, M., Demirgüç-Kunt, A., Maksimovic, V., 2011. Small vs. Young Firms Across the World — Contribution to Employment, Job Creation, and Growth. Policy Research Working Paper 5631 (The World Bank Develop- ment Research Group). ani, R.J., 2003. Micro Enterprise Development in Ghana. Accra. eck, T., Cull, R., 2014. Small- and Medium-Sized Enterprise Finance in Africa. Africa Growth Initiative, Working Paper 16/July 2014. eck, T., Demirgüç-Kunt, A., Maksimovic, V., 2003. “Financial and Legal Institutions and Firm Size” World Bank mimeo. eck, T., Demirgüç-Kunt, A., Maksimovic, V., 2004. Financing Patterns Around the World: Are Small Firms Different? World Bank working paper. eck, T., Demirguc-Kunt, A., Laeven, L., Maksimovic, V., 2006. The determi- nants of financing obstacles. J. Int. Money Finance 25, 932–952. erg, G., Fuchs M., (2013). Bank Financing of SMEs in Five sub-Saharan African Countries: The Role of Competition, Innovation, and the Govern- ment. Policy Research Working Paper/WPS6563, the World Bank. ertrand, M., Schoar, A., Thesmar, D., 2007. Banking deregulation and industry structure: evidence from the French banking reforms of 1985. J. Finance 62, 597–628. igsten, A., Collier, P., Dercon, S., Fafchamps, M., Guthier, B., Gunning, W., et al., 2000. Credit Constraints in Manufacturing Enterprises in Africa. Work- ing Paper WPS/2000. Centre for the study of African Economies, Oxford University, Oxford. igsten, A., Collier, P., Dercon, S., Fatchamps, M., Gauthier, B., Gunning, W.J., Oduro, A., Oostendrop, R., Patillo, C., Soderbom, M., Teal, F., Zeufack, A., 2003. Credit constraints in manufacturing enterprises in Africa. J. African Econ. 12 (1), 104–125. lack, S.E., Strahan, P.E., 2002. Entrepreneurship and bank credit availability. J. Finance 57 (6), 2807–2833. oundless, 2014. “Pecking Order Consideration.” Boundless Finance. Boundless, 30 Dec. 2014. Retrieved 16 Apr. 2015 from https://www.boundless.com/finance/textbooks/192/capital-structure-13/ capital-structure-considerations-102/pecking-order-consideration-440-7966 uatsi, S.N., 2002. Financing non-traditional exporters in Ghana. J. Bus. Ind. Mark. 17 (6), 501–522. etorelli, N., Strahan, P.E., 2004. Finance as a Barrier to Entry: Bank Competi- tion and Industry Structure in Local Us Markets. NBER Working Paper No. 10832. laessens, S., Tzioumis K., 2006. Measuring firms’ access to finance. Paper prepared for Conference: Access to Finance: Building Inclusive Financial Systems, organized by the Brooking Institution and the World Bank in Washington, D.C., May 30-31, 2006. ollier, P., 2009. Rethinking Finance for Africa’s Small Firms in “SME Financ- ing in SSA,” Proparco Magazine on Private Sector and Development, Issue 1 May 2009.aniels, L., Ngwira, A., 1993. Results of a Nation-wide Survey on Micro, Small and Medium Enterprises in Malawi. GEMINI Technical Report No 53. PACT Publications, New York.ment Finance 7 (2017) 18–28 awson, J., 1993. Impact of structural adjustment on the small enterprise sec- tor: a comparison of the Ghanaian and Tanzanian experiences. In: Helmsing, A., Kolstee, T. (Eds.), Small Enterprise and Changing Policies: Structural Adjustment Financial Policy and Assistance Programmes in Africa. IT Pub- lications, London, pp. 71–90. eakins, D., North, D., Baldock, R., Whittam, G., 2008. SMEs’ Access to Finance: Is There Still a Debt Finance Gap? Institute for Small Business & Entrepreneurship. 5–7 November 2008 - Belfast, N. Ireland, 1–19. emirgüç-Kunt, A., Maksimovic, V., 1998. Law, finance and firm growth. J. Finance 53, 2107–2131. bandi, E.C., Amissah, G., 2014. Financing options for small and medium enterprises (SMEs) in Nigeria. Eur. Sci. J. 10, 1. ockel, A.G., Akoena, S. K., 2002. Financial Intermediation for the Poor: Credit Demand by Micro, Small and Medium Scale Enterprises in Ghana. A Fur- ther Assignment for Financial Sector Policy? IFLIP Research Paper 02-6, International Labour Organisation. uiso, L., Sapienza, P., Zingales, L., 2004. Does local financial development matter? Quarter. J. Econ. 119 (3), 929–969. ansen, A., Kimeria, C., Ndirangu, B., Oshry, N., Wendle, J., 2012. Assess- ing Credit Guarantee Schemes for SME Finance in Africa: Evidence from Ghana, Kenya, South Africa and Tanzania. AFD Working Paper 123. aselip, J., Desgain, D., Mackenzie, G., 2014. Financing energy SMEs in Ghana and Senegal: outcomes, barriers and prospects. Energy Policy 65, 369–376. untchev, V., Ramalho, R., Rodríguez-Meza, R., Yang, J.S., 2012. What Have We Learned from the Enterprise Surveys Regarding Access to Finance by SMEs? Version: February 14, 2012. ove, I., Mylenko, N., (2003). Credit Reporting and Financing Constraints. World Bank Policy Research Paper No. 3142. ensah, S., 2004. A Review of SME Financing Schemes in Ghana. UNIDO Regional Workshop of Financing SMEs, Accra. kpara, John O., 2011. Factors constraining the growth and survival of SMEs in Nigeria: implications for poverty alleviation. Manag. Res. Rev. 34 (2), 156–171. sei, B., Baah-Nuakoh, A., Tutu, K., Sowa, N., 1993. Impact of structural adjustment on small-scale enterprises in Ghana. In: Helmsing, A., Kolstee, T. (Eds.), Small Enterprise and Changing Policies: Structural Adjustment, Financial Policy and Assistance Programmes in Africa. IT Publications, London, pp. 53–70. arker, R., Riopelle, R. Steel, W., 1995. Small Enterprises Adjusting to Liber- alization in Five African Countries. World Bank Discussion Paper, No 271, African Technical Department Series, The World Bank, Washington DC. WC (2013). Ghana Banking Survey. uartey, P., 2002. Financing small and medium-sized enterprises in Ghana. J. African Bus. 4, 37–56. acerdoti, E., 2005. Access to Bank Credit in Sub-saharan Africa: Key Issues and Reform Strategies. IMF Working Paper, Vol., pp. 1–39, 2005. chiffer, M., Weder, B., 2001. Firm Size and the Business Environment: World- wide Survey Results. IFC Discussion paper 43. ilva, F., Carreira, C., 2010. Are Financial Constraints a Threat to Innovation Activity? Evidence from Portuguese Firms. Faculdade de Economia/GEMF, Universidade de Coimbra. owa, N.K., Baah-Nuakoh, A., Tutu, K.A., Osei, B., 1992. Small Enterprise and Adjustment, The Impact of Ghana’s Economic Recovery Programme on Small-Scale Industrial Enterprises. Research Reports, Overseas Devel- opment Institute, 111 Westminster BridgeRoad, London SE1 7JD. teel, W.F., Webster, L.M., 1991. Small Enterprises in Ghana: Responses to Adjustment Industry. Series Paper, No. 33, the World Bank Industry and Energy Department, Washington DC.