What is the summary and analysis Credit Constraints & Productive Entrepreneurship in Africa ?

While most African countries have recently recorded relatively high growth, even during and in the aftermath of the global financial crisis, their productivity gap with more advanced economies is still substantial. A key reason is the continued prevalence of lowproductivity activities, including a large share of subsistence agriculture. As experiences from more developed regions have shown, the transition to activities with higher value added content requires the emergence of productive entrepreneurship, nurtured not only by market dynamics but also by effective and well-targeted policies. In particular, the development of productive entrepreneurship requires a business environment that facilitates access to credit for new high-potential enterprises. Entrepreneurs everywhere cite limited availability of finance as a major obstacle to their activities (Stein et al., 2010), but this constraint is especially binding in Africa. According to the World Bank's Enterprise Surveys, about 45 percent of firms in Sub-Saharan Africa (SSA) identify access to finance as a key obstacle to their business, relative to 13 percent in OECD countries. Access to credit and hence firm creation and growth in SSA is constrained by, among others, distortions in the financial markets, especially high collateral requirements and poorly designed and enforced property rights. This paper utilizes a theoretical framework where firm creation and growth hinge on matching the searching entrepreneurs with production technologies and access to credit. Limited credit due to the lack of collateral slows the creation and growth of new private firms, thus hampering also job creation. The model is suitable to many African countries where financial frictions stem from the limited collateral and a weak legal framework. It generates two empirically important results. First, access to investment capital arises as a binding constraint to entrepreneurship, and this constraint prevails even in the presence of excess liquidity in the banking sector. A key constraint to access to credit is the shortage of collateralizable assets. Second, legal rights and informational depth in credit markets are shown to support firm creation and growth, and hence private sector employment. The model is tested using data for a sample of 20 African countries over the period 2005- 09. We use as indicator for entrepreneurship a measure of 'new business density', proxied by the new business registration per one thousand people aged 15-64 (from the World Bank's Doing Business database). The results of the empirical analysis point to possible policy interventions that may help alleviate credit constraints and encourage entrepreneurship. The remainder of the paper proceeds as follows. After this introduction, the next section reviews the theoretical and empirical literature, with a focus on the determinants of entrepreneurship and the impact of institutional and policy reforms on entrepreneurship. The third section presents the theoretical model that underlies the empirical analysis. Section 4 contains the empirical analysis and the regression results. Section 5 summarizes the findings and discusses policy implications and possible policy interventions that may help address the constraints to entrepreneurship with a focus on the access to credit. 2 2. Overview of the literature 2.1 Theoretical literature Given the high persistence of unemployment, working poverty, underemployment and vulnerable employment in Africa, policy makers have put increasing emphasis on supporting productive entrepreneurship and small and medium enterprises (SMEs) to achieve high growth with job creation. This focus is driven by recognition that while 'necessity' (i.e., low productivity) entrepreneurship is abundant in Africa, the potential of an 'opportunity' or high productivity entrepreneurship has been mostly untapped. The unutilized potential impedes Africa's growth and employment, as the opportunity entrepreneurship is found to have a significant positive effect on development, while necessity entrepreneurship has almost none (Acs and Varga, 2005). This paper focuses on creation and growth of productive private firms. As Baumol (1990) underscored, while the extent of entrepreneurship across societies is mostly given, policies impact whether potential entrepreneurs enter into highly productive, less productive or even destructive activities. Among other objective, such policies strive to overcome both financial and non-financial (regulatory) constraints, which have impeded productive entrepreneurship and employment across developing countries. Brixiova (2010) developed a model of non-financial constraints to firm creation in Africa's low income countries, including skill shortages of potential workers and entrepreneurs. In this paper, we focus on financial constraints to productive entrepreneurship. In Africa, financial constraints to entrepreneurship are particularly severe because of unclear property rights and restrictions on using assets such as land as collateral. To reflect this fact, the framework presented below is a streamlined version of Brixiova and Kiyotaki (1997) who modelled how credit constraints slow down the creation and growth of private firms in transition economies in the early stages of transition, where the productive private sector was emerging and the financial sector underdeveloped. In turn, they build on research by Kiyotaki and Moore (1997) on the role of credit constraints in economies where a durable productive asset (land) acts as collateral.2 Iyigun and Rodrik (2005) developed a theoretical model where investment decisions and policy outcomes are subject to uncertainty and examine the growth effects of the interaction between institutional and policy reform and entrepreneurship. Their model showed that institutional reform has a negative growth effect in settings where entrepreneurial activity is vibrant but has a positive impact where entrepreneurial activity is weak. The authors concluded that institutional reform would be more successful where the level of entrepreneurship is weak. They then assessed the empirical relevance of their model using cross-sectional data and the ratio of self-employed to total non-agricultural employment as an indicator of entrepreneurship. They found that the relationship 2 A decrease in asset (land) prices then lowers the value of collateral and investment in the creditconstrained sector. Hart and Moore (1994) show formally how the possibility of default puts a limit on the amount of borrowing that entrepreneurs can undertake for even highly profitable projects. 3 between institutional reform and entrepreneurship is negative and statistically significant, and interpret this result as empirical evidence in support of their theoretical proposition. More recently, Aghion, Fally and Scarpetta (2007) examined the impact of private credit availability and stock market capitalization on firm entry and growth in advanced and emerging market economies. While their framework advances both theoretical and empirical literature, it is not directly applicable to most African countries where stock markets and more broadly non-bank financial institutions are either missing or still severely underdeveloped. 2.2 Empirical literature The empirical literature relevant to the issues we examine in this paper relates to two areas: (i) the empirical research on the determinants of entrepreneurshipin particular the role of credit and liquidityand (ii) empirical studies on policy and institutional reforms and their impact on entrepreneurship.3 The role of information sharing and lender and borrower legal rights for access to credit is key for this study. Some studies have found that information sharing facilitates access to credit (Jappelli and Pagano, 2002; Djankov et al., 2007; Brown et al., 2009).4 However, Negrin (2001) shows that wider information sharing led to less access to bank credit for small and medium-sized firms in Mexico. Regarding whether better legal rights ease access to credit, Demirg-Kunt and Maksimovic (1998) find that the proportion of firms using long-term external financing is larger in countries with efficient legal systems. Since long-term external finance is often necessary for long-term investment, credit constraints for entrepreneurs planning to undertake long-term projects may be greater in countries with inefficient legal systems.5 Qian and Strahan (2007) show that the effects of creditor rights on loans depend on borrowers' characteristics such as the size and tangibility of their assets. Specifically, foreign banks "appear especially sensitive to the legal and institutional environment, with their ownership declining relative to domestic banks as creditor protection falls" (p. 2803). This result is particularly relevant to African countries that rely highly on the presence of foreign banks. Using data from 12 transition countries, Haselmann et al. (2010) find that the supply of bank credit increased after a legal change and that changes in collateral law matter more than changes in bankruptcy law for increased bank lending. The second line of empirical research focuses on the effects of policy and institutional reforms on the creation and growth of entrepreneurial activity. However, most of the 3 For a recent review of literature on the links between entrepreneurship, reforms, and development (growth), see Baliamoune-Lutz (2010). More broadly, for studies on links between financial development and growth see, for example, King and Levine 1993, Rajan and Zingales, 1998; Baliamoune-Lutz and Ndikumana, 2007; Beck and Demirg-Kunt, 2008; and Baliamoune-Lutz, 2011. 4 See also the survey by Japppelli and Pagano (2000). 5 Using data from 49 countries and controlling for legal origin, La Porta et al. (1998) report "that commonlaw countries generally have the strongest, and French civil-law countries the weakest, legal protections of investors, with German- and Scandinavian-civil-law countries located in the middle." 4 studies have examined the effects of tax system reforms, and focused primarily on OECD countries and in many cases using micro-level (firm-level or individual) data. However, a limited number of studies have focused on other reforms besides taxation. Using establishment-level data from the manufacturing sector, Gaston and Werner (2002) explore the effects of financial liberalization on fixed investment in Mexico, focusing on the role of real estate as collateral. They find that financial constraints were reduced for the smallest firms, but not for larger ones, and given that banks relied on collateral in their lending decision, the importance of having real estate rose. Baliamoune-Lutz (2007) empirically tests the Iyigun-Rodrik (2005) model using data from developed and developing countries and exploring the role of institutional and policy reforms on contributions of entrepreneurship to development. She also finds that the effect of policy reforms (trade reforms, proxied by openness to trade) is negative when entrepreneurial activity is weak and positive when it is vibrant, while the effects of institutional reforms (measured by the International Country Risk Guide composite index) is positive when the level of entrepreneurship is low and negative when it is high. Finally, Baliamoune-Lutz (2010) focuses only on developing countries and uses panel data for the period 1990-2002 and Arellano -Bond GMM estimations to explore the impact of institutional and policy reforms on the growth effects of entrepreneurship. She obtains empirical evidence suggesting that the interplay of trade reforms and entrepreneurship has a negative impact on growth, whereas the interplay of financial sector reform and entrepreneurship has a non-linear positive effect on growth.6 The joint effect of institutional reforms and entrepreneurship on growth is not significant though. 3. A model of entrepreneurship This section presents a model where entrepreneurship, specifically firm creation and growth, is constrained due to the lack of access to finance, which in turn is limited by the lack of collateral. The model is a streamlined version of the model in Brixiova and Kiyotaki (1997), which was developed for transition economies at the early stages of transition, with emerging private sectors. In this paper we apply it to African countries, where in many instances formal and productive private sector is still underdeveloped (AfDB, 2011 forthcoming). As the World Bank Enterprise Surveys document, one of the key constraints faced by firms in Africa is the lack of access to credit (Appendix A, Figure A1). By emphasizing the lack of collateral and the weak application of the rule of law, the framework below is particularly suitable for most African countries where the financial sectors are dominated by banks and binding credit constraints co-exist with excess liquidity. Specifically, the weak implementation of the legal framework makes banks require high collateral (in 6 This seems to be consistent with Claessens and Perotti (2007: 748) who conclude that financial liberalization with the view to increase access "may in practice increase fragility and inequality, and lead to political backlash against reforms," in the absence of strong oversight institutions. 5 some cases over 100 percent of the loan), leaving the financial needs of many potential entrepreneurs unaddressed. We provide the details of the analytical framework below. The population is normalized to one and consists of infinitely-lived entrepreneurs - dynasties7 - and workers, with population shares and 1 , respectively. All agents are endowed with one unit of time every period and have risk neutral preferences in consumption of a single good, c. When starting their search, entrepreneurs are endowed with net worth 0 a0 > . The entrepreneur searches for a business opportunity; the search costs her ( ) / 2 2 d x = x units of the consumption good per unit of time; > 0 is the parameter of search efficiency. In turn, she finds the opportunity according to a Poisson process with the arrival rate of x . She produces output ( y) in the formal sector with labor (n), productivity (z) and capital (k) according to the production function: = 1 ( ) 1 1 y zk n (1) The output can be used either for investment or consumption. Capital is entrepreneurspecific, implying that once the entrepreneur invests in it, she is the only one who can use the accumulated stock until she retires and passes it on to her successor. Due to imperfections in legal frameworks and property rights in Africa, the outside value of the accumulated capital is lower than its worth to the entrepreneur, and in some cases significantly so. Put differently, in the case of default, lenders can recover only (1 ) portion of the accumulated capital, where reflects the weaknesses in the legal framework. The lenders limit the loan to the entrepreneur,b , to the recoverable value of her capital, that is: 8 b (1 )k (2) The entrepreneur finances capital ) (k from both borrowing ) (b and her own net worth (a), which she accumulates according to: a& = yn wn rb c (3) where w is the wage rate and r is the real interest rate on debt, which in equilibrium equals the rate of time preference (and time t is suppressed. Private firms are destroyed at exogenously given rate > r . The exiting entrepreneur consumes all her accumulated net worth except a0 which she passes on her successor, derives utility a a0 , and dies immediately after. The successor searches for another 7 That is when the entrepreneur retires, she passes accumulated capital on to her successor. 8 The value of depends on the specificity of the capital and strength of the legal framework, including the bankruptcy law. Renting the capital is not an option as the rental market is mostly absent in Africa. 6 business opportunity, with the initial net worth a0 . In addition to the formal sector, output can be produced in the informal sector according to = 1 1/(1 ) Yu Zu Nu , where Zu is productivity and Nu is the informal sector employment. Denoting mp to be the share of entrepreneurs operating private firms and n the number of workers per firm, the following labor market equilibrium condition for workers needs to hold: 1 (4) = Nu + mp n Letting mu be the share of entrepreneurs searching for business opportunities and mp the share of entrepreneurs running firms, the equilibrium conditions for entrepreneurs satisfy: = mu + mp (5) The change in the number of entrepreneurs searching for business opportunities, mu & , is given by the difference between inflows into the pool of searching entrepreneurs, mu = mp ( and the exits from it, ) u xm . From (5), mp mu & = & , with the initial value of number of private firms, mp0 , set to 0. u p u u u m& = m xm = ( m ) xm (6) Taking wages and interest rates as given, the equilibrium is characterized by (i) entrepreneur's choice of search effort, labor, capital, debt, and savings that maximize the expected discounted utility; (ii) worker's choice of allocation of labor and consumption; (iii) products and debt markets that clear and (iv) labor markets that satisfy (4)-(6). The equilibrium wage rate rises in the aggregate capital stock (K) , and is determined as follows: ( ) 1 w K zK w = (7) where K k di = i 0 , with i k being capital of an entrepreneur i . 7 With the constant return-to-scale production function and output price normalized to 1, profits are zero in equilibrium, that is 1/(1 )(zk) n w(K)n R(K)k 1 = + and the return on capital, ) R(K becomes:9 (1 )/ ( ) 1 ( ) R K = zw K (8) When K < K , the rate of return on capital is above the real interest rate on debt, R( and the entrepreneur borrows up to the credit limit for capital investment, i.e. K) > r the credit constraint is binding and b = (1 )k . The entire net worth is spent on the down-payment for capital: a = k . The return on net worth exceeds the real interest ) (r by the leverage ) (1/ times the difference between the return on capital that the entrepreneur owns and interest rate she pays for borrowing. Suppressing the time subscripts and denoting u J and J (a) as a present discounted value of an entrepreneur searching for a business opportunity and an entrepreneur running a private firm with net worth a , respectively, the corresponding Bellman equations are: [ ] + = u x u x J a J x rJ ( ) 2 max 0 2 (9) [ ] a a J rJ a a a J J a u & ( ) = 0 + ( ( )) + (10) where (9) states that the return from searching for a business opportunity equals the net expected return from running a business with the net worth a0 . According to (10), the return on running a firm consists of gains from accumulating net worth and expected utility of consumption at the time of exiting from the labor force and net expected gain from search.