漫紫色ˇ笔迹。

版权声明：转载时请以超链接形式标明文章原始出处和作者信息及本声明
http://yanye.52blog.net/logs/2987203.html

Abstract

The paper examines the research on finance and the small and medium-sized enterprise (SME) sector in developing countries. The paper argues that theoretical insights into this field have largely been confined to studies undertaken in the US and UK. In developing countries' research on both the supply and demand for finance in relation to SMEs has been empirically-based and has been pre-occupied with gathering information on the characteristics of SMEs and lending institutions rather than on testing theoretical proportions that would improve our understanding of the relationship between finance and SMEs. A research agenda is outlined.

Key words: Finance, enterprises, liberalization, developing economies

Interest in the role of small and medium-sized enterprises (SME) in the development process continues to be in the forefront of policy debates in developing countries. The advantages claimed for SMEs are various, including: the encouragement of entrepreneurship; the greater likelihood that SMEs will utilize labor intensive technologies and thus have an immediate impact on employment generation; the fact that they can usually be established rapidly and put into operation to produce quick returns; the ability of SME development to encourage the process of both inter- and intra-regional decentralization; and the notion that they may become a countervailing force against the economic power of larger enterprises. More generally, the development of SMEs is seen as accelerating the achievement of wider economic and socio-economic objectives, including poverty alleviation.

Staley and Morse (1965,p. 318) identify a `developmental approach' to SME promotion which has as its objective the creation of "economically viable enterprises which can stand on their own feet without perpetual subsidy and can make a positive contribution to the growth of real income and therefore to better living levels." This approach emphasizes the importance of efficiency in new SMEs. Small producers must be encouraged to adopt new methods, to move into new lines of production and in the longer-run, wherever feasible, be encouraged to become medium- or even large-scale producers.

Unfortunately, the empirical evidence on the relation between efficiency and firm size is rather inconclusive. Little, Mazumdar and Page (1987) reported technical efficiency does not vary systematically with firm size. A collection of studies found efficiency was lower in smaller firms (Ho, 1980; Cortes, Berry & Ishaq, 1987; Goldar, 1988). Opposite conclusions can be drawn from the twelve country studies by Liedholm and Mead (1987).

More recent concerns associated with the growth and efficiency of smaller enterprises have also become prominent (Mazumdar, 1997). Using the case of Northern Italy, Piore and Sabel (1984) have argued that small enterprises are more efficient because they have adopted a flexible specialization approach. Correspondingly, there has been growing interest in whether this model has been or can be replicated in developing countries (Schmitz, 1989; Pederson, 1994; Schmitz & Musyck, 1994; Schmitz, 1995).

The role of finance has been viewed as a critical element for the development of small and medium-sized enterprises. Previous studies have highlighted the limited access to financial resources available to smaller enterprises compared to larger organizations and the consequences for their growth and development (Levy, 1993). Typically, smaller enterprises face higher transaction costs than larger enterprises in obtaining credit (Saito & Villanueva, 1981). Insufficient funding has been made available to finance working capital (Peel & Wilson, 1996). Poor management and accounting practices have hampered the ability of smaller enterprises to raise finances. Information asymmetries associated with lending to small scale borrowers have restricted the flow of finance to smaller enterprises. In spite of these claims however, some studies show a large number of small enterprises fail because of non-financial reasons (Liedholm, MacPherson & Chuta, 1994).

The purpose of this paper is to systematically examine research on the relation between finance and small and medium-sized enterprise development, and identify some of the gaps in our knowledge. While a considerable amount is known about the characteristics and behavior of small and medium-sized enterprises, this knowledge continues to be imperfect and a large number of questions remain unanswered in relation to finance and small enterprise development in developing countries. This paper discusses some of the issues raised by previous research and points to newer areas that can fruitfully be investigated.

Section two provides some general characteristics of previous research on small enterprise development and indicates the main areas of emphasis. The third section reviews some of the issues that emerge from the literature assessing the impact of financial sector reforms on small and medium-sized enterprises. A number of the underlying assumptions concerning the effects of reforms on both the demand and supply of finance are also discussed in this section. The fourth section examines the theoretical treatment of the relation between finance and small and medium-sized enterprise development and examines some of these perspectives in relation to low income countries. The concluding section prioritizes the areas where our knowledge of finance and SMEs in developing countries needs to be enhanced.

Economic research on small and medium-sized enterprises has a number of distinctive features. First, the bulk of research has predominantly been undertaken in the context of US and UK firms and has sometimes entailed comparisons with other European economies (Storey, 1995). As such, the theoretical work has assumed an institutional setting and made assumptions about the policy environment that is relevant to these economies. Second, much of the theoretical work on enterprises has related to larger firms, and in the context of smaller firms, research efforts have tended to focus on the larger small enterprises. Third, research has been divided between examinations of the macro environment within which small and medium-sized firms operate and studies on the internal workings of enterprises.

With respect to the macroeconomic environment, the work has involved: assessing the importance of aggregate demand and the role of macroeconomic policies for small enterprise development; the role played by the various formal sector and informal sector lending institutions that provide credit for the small scale sector and the importance of promotional policies in terms of providing managerial, technical and marketing information to small enterprises. With respect to internal factors, attention has concentrated on the choice of investment; employment; firm level performance and productivity; capital structure and the ownership and incentive structures for management.

SME Research on Developing Countries

There has been less work directly related to small and medium-sized enterprises in low income countries, although the geographical spread of this work has been wide, covering Africa, Asia and South America. It is also the case that the major proportion of this work has been empirically rather than theoretically focused, and it is reasonable to conclude that relatively little is known about the behavior of entrepreneurs in low income countries relative to those operating in higher income countries.

The literature on developing countries follows the same division as studies on the industrialized countries, by attempting to distinguish between the external and internal factors that affect small and medium-sized enterprise development. In terms of external factors, much of the earlier literature in the 1970s was concerned with the biases established against smaller enterprises through trade and industrial policies pursued in low income countries, and with the design of appropriate support institutions that would compensate for these so-called policy-induced biases. Schmitz (1982) identified a number of external factors that affected small enterprise development. These included lack of credit at reasonable cost, lack of working capital, poor infrastructure, and competition from larger and foreign firms. Policies of economic liberalization were introduced later, partly with the aim of reducing the bias in favor of larger enterprises. The 1980s also witnessed greater concern for the need to establish policies at the macro, meso and micro level regarding the development of smaller enterprises in low income countries (Stewart, 1990).

Earlier research on the internal workings of small and medium-sized enterprises was mainly concerned with the size of small firms and providing explanations for their growth. Staley and Morse (1965) examined the stages small firms pass through as an economy grows. They postulated several reasons why small firms in low income countries initially grow rapidly before their share of total industrial activity begins to decline. Rapid growth of small firms could be explained where (a) demand was rising as rural incomes were growing and where infrastructure costs still favored small firms locating near fragmented markets and (b) subcontracting and local assembly was common, as for example in varieties of machine-shop activities and where smaller firms produced a range of differentiated and innovative products serving small total markets. But as Anderson (1982) pointed out, these propositions had not been quantitatively tested by the early 1980s.

These early researchers were also preoccupied with investigating the extent to which small businesses create the foundation for larger company growth. As Anderson (1982,p. 923) reported:

    "firms practically always begin as very small entities, with
    low amounts of capital drawn from the savings of the owner
    or borrowings from friends and relatives; initial levels of
    employment are low, typically less than a dozen, though the
    figure varies with the nature of the business; the social
    and occupational backgrounds of the owners varies greatly;
    and the firms that expand into medium or large scale
    activities do so continually or in steps. Expansion can be
    very fast for some firms, though the growth rates appear as
    broadly distributed as their final sizes"

Anderson (1982,p. 926) concluded that the available empirical evidence suggested that a significant part of the growth of large-scale enterprises was rooted in the expansion of once small firms through the size distribution.

In low income countries, work directed towards the internal workings of enterprises has been hampered by the lack of basic data on the management and characteristics of smaller firms. Considerable effort has been expended on attempting to gather consistent and measurable information about small firms. Industrial censuses in a large number of low income countries have not been undertaken annually; they have concentrated on larger enterprises; they have only infrequently surveyed small enterprises, and have often been published with long delays. As a consequence, useful time series data for smaller enterprises from official sources are largely absent.

This has had implications for research efforts into small enterprises in low income countries in three important ways. First, a considerable amount of time has been spent on gathering baseline information on small firms. This has involved identifying universes and constructing samples, devising methods to deal with delinquent returns and editing the results in a consistent manner. Second, information collected tends to be more qualitative than quantitative because of the poor record keeping, as well as the lack of cross-referencing sources through formal channels that can be used to confirm data reliability. This tends to limit the use of the data in statistical analysis. Third, surveys are more often conducted on an ad hoc basis and at a point in time. Few compare different points in time and fewer still have attempted to use the same database for follow-up work.

As a result, time series work on the small scale sector is relatively scarce. The preoccupation with gathering baseline data and the restricted nature of the data that have eventually been collected has resulted in a preponderance of studies that describe and report on the characteristics and features of the small scale sector rather than test theoretical propositions about relationships and the expected behavior of the small firm sector. This is not to suggest that theorizing and testing of theories is completely absent in relation to work on small enterprises in low income countries. But in comparison with work in industrialized countries, or in relation to research on the behavior of larger foreign-owned enterprises in low income countries, it is demonstrably less evident.

In contrast to the earlier work, a distinctive feature of the current spate of empirical work undertaken in low income countries rests with its emphasis on attempting to identify the constraints facing the development of the small scale sector (Levy, 1993). Most surveys have sought to capture the range of factors that inhibit the growth and development of small firms. Since the early 1980s, various attempts have been made to classify constraints into those emanating from either the demand or supply side (Anderson, 1982; Little, Mazumdar & Page, 1987). On the supply side, again lack of access to investment and working capital were identified as prominent constraints. Others have included problems associated with the supply of raw materials, lack of skilled labor, insufficient training and poor infrastructure.

Three aspects relating to demand side constraints were repeatedly emphasized. The first of these is formidable competition from large and foreign enterprises causing instability in demand for the products of smaller firms and low rates of capacity utilization (Bruch & Hiemenz, 1984). Second, smaller firms often suffered from insufficient demand related to the low purchasing power of the main customers for many of their products (Harper, 1984). Third, smaller firms faced obvious disadvantages in export markets through their inability to gain marketing knowledge and supply goods on time and with sufficient quality. More recent research into factors likely to affect the export success of smaller firms has pointed to the importance of working in clusters (Schmitz, 1998). Collective efficiency is achieved by smaller firms producing differentiated products in such clusters.

A large proportion of the information on constraints has been collected from smaller firms through questionnaires asking owners and managers to give their views on either the kind of constraint they face, whether it be related to such factors as access to finance, poor managerial skills and lack of training opportunities and the high cost of inputs, or on the severity of the constraints, often ranking them on an ordinal scale. Few studies have concentrated on a particular constraint, so that finance has most often been identified as an inhibiting factor as part of a larger investigation into a wider range of variables. A recent exception has been the study by Cobham and Subramaniam (1998), which compared patterns of corporate finance in India with those found in industrialized countries in Europe. In India, they found sources of internal finance to be more significant among larger firms than smaller firms. The results in terms of the significance of financing as a constraint to development are mixed and it is difficult to draw conclusions about the subject. Interpretation is complicated because of the qualitative nature of the surveys and due to the fact that inquiries have almost exclusively been directed at firms that exist rather than following the histories of those that have eventually failed.

In summary, it cannot be denied that a considerable amount is known about the behavior of smaller firms in a range of areas relating to growth, efficiency, management, investment and employment. A smaller proportion of this work is theoretical in nature. The vast majority of studies, particularly those relating to low income countries, are empirical, and in general surveys have been used to generate basic information on smaller enterprises where official enumeration is lacking.

The dilemma facing researchers is how to maximize the use of existing surveys and forgo the need for newer inquiries that may waste resources and time by duplicating or replicating existing sources of information. What seems clear is that, in the past, there has been too distinct a separation between theoretical work that advances hypotheses about the small scale sector and empirical work that has not clearly sought to test hypotheses, but instead has been involved with describing the characteristics of small enterprises. In part, this can best be explained by the preoccupation with gathering original data that in some way has crowded out initiatives to apply the data to test theories. Alternatively, it may simply reflect data inadequacies once they had been collected. Whatever the reasons, it is apparent that work in relation to low income countries, where these data problems most evidently exist, has lacked the formalized hypotheses, data collection and testing approaches widely adopted in other branches of industrial studies.

In the 1990s, greater attention was given to assessing the impact on smaller enterprises of economic reforms introduced as part of World Bank structural adjustment programs (Cook, 1996). Overall views concerning the likely impact of these reforms on small enterprise development have varied. Whereas some claim structural adjustment has brought considerable benefits to small scale enterprises, others argue that wide-ranging constraints have frequently prevented such effects from reaching small scale enterprises. Empirical evidence in support of these claims is briefly examined in this section in relation to the financial reforms that have been implemented in low income countries over the past fifteen years.

The evidence reviewed is found in a growing but relatively small number of studies that directly attempt to measure the impact of economic liberalization and structural adjustment on the small scale industrial sector (Liedholm, 1990; Koppel, 1991; Steel & Webster, 1992; Boeh & Ocansey, 1995; Dawson, 1993; Steel, 1993, 1994; Vachani, 1994; Zake, 1994; Helmsing & Kolstee, 1993). Further evidence can be extracted from a larger number of studies that concentrate more specifically on constraints to small scale enterprise development. However, the ability to establish direct causal links is problematic owing to the paucity of time series data for measuring the impact of structural adjustment on small scale enterprise development, and to the limitations of evaluation methodologies.

Studies broadly fall into three categories: those attempting to determine the impact of measures designed to work through the market mechanism, such as interest and exchange rate policy; those that attempt to assess the overall impact of policy measures on small scale development, which generally include an array of market and non-market initiatives; and, those that review the policy environment. The former set of studies tends to rely on economic data drawn from official statistics, such as that supplied by the monetary authorities and census bureau at the establishment-level. Studies based on direct surveys, as a rule, tend to focus on whether or not the factors previously identified as constraining the development of small scale enterprises have been eased or removed. Finally, policy studies are generally broader in focus, reviewing the existing policy work and providing preions, often based upon what appears to work elsewhere. This brief review concentrates on the first category of studies, those that have attempted to evaluate the effects of specific policy changes, in this case dealing with the effects of financial reform on SME developments.

Financial Sector Liberalization

Financial liberalization is expected to result in the reallocation of domestic credit toward smaller enterprises, and the substitution of more expensive forms of credit for cheaper ones. Moreover, while nominal and real interest rises are anticipated, real returns are expected to outweigh this burden. It is also argued that the process of transferring from an administrative process of credit allocation to a market-based mechanism will not only improve the access to credit for smaller local enterprises, but will also lower the transaction costs associated with borrowing. Further, in cases where highly subsidized export credit schemes exist for larger enterprises, financial liberalization can be expected to remove this bias.

Questions have been raised in the developing country context over these predictions. Taylor (1988) argues that financial liberalization will not result in more funds for borrowing being available. As interest rates rise, funds available will be diverted out of the informal sector to the formal sector. An increased share of borrowing will take place in the formal financial sector but the total available funds between the two sectors will remain unchanged. The net result in a macro sense means there will be no new borrowing. This contradicts the McKinnon-Shaw hypothesis which argues that financial liberalization, by increasing interest rates, leads to higher savings, investment and growth. Unfortunately, the desired effects have not always materialized in the way that policy preions envisaged. As Steel (1994) highlights, high transaction costs and risks associated with small loans, a lack of collateral, and an historical orientation towards larger enterprises continue to restrict small scale enterprise access to formal credit.

The case of Ghana shows that despite financial sector reform, the strengthening of banking capabilities and the introduction of numerous financial instruments, such as the stock exchange, a venture capital company and business assistance funds, access to institutional credit for working capital and equipment continued to be a major constraint to small enterprise development (Steel & Webster, 1992). Even where demand for small scale enterprise products appeared strong, a lack of credit meant that many small enterprises did not have the capacity to respond and expand production. Interest rates of 30 per cent or more, high transactions costs and an administration and culture unfriendly to small scale enterprises contributed to the problem (Boeh & Ocansey, 1995). The Ghana study by Osei, et al., (1993) cites similar evidence; 95 per cent of the respondents depended solely on personal resources and loans from relatives and friends. Dawson's (1993) work in Ghana and Tanzania also confirms these findings; of the 672 small scale enterprises in the Ghana study, only two had received a bank loan. In Tanzania, the formal banking system was seen to be out of reach for almost all small enterprises. The World Bank reported that around 90 per cent of small enterprises surveyed indicated that access to credit was a major constraint to new investment (World Bank, 1994).

Kariuki's (1995) study of bank credit access in Kenya illustrates this point further. A survey of 89 small and medium-scale firms in manufacturing and service industries, combined with secondary information from commercial banks, found that from 1985 to 1990, the average real volume of credit for the sample firms fell, except for 1986, where there was a marginal increase of 1.5 per cent. Several deterrents to utilizing formal credit were identified. It was found that small scale borrowers were faced with higher nominal interest rates at higher inflation rates in the latter half of the 1980s. Moreover, the explicit transaction costs of borrowing were found to be high in relation to interest costs.

The cases of Bangladesh, Nepal and the Philippines appear to support these claims (Meier and Pilgrim, 1994). Despite specific programs aimed at small scale enterprises, only between 12 per cent and 33 per cent of those surveyed were found to have access to formal credit and, of those, the majority were from the larger end of the sector. Again, factors such as the relatively high cost of processing small loans, the need for high collateral, and bureaucratic procedures were seen to be restricting lending to small scale enterprises. The taxation policies that were also examined had little impact on small scale enterprises, particularly as many of those surveyed were found not to be paying taxes.

Similar evidence regarding the lack of importance given by small scale enterprises to tax policies is found in Southern Africa, including Niger, Botswana, Swaziland, Lesotho, Malawi, and Zimbabwe (Mead, 1994). Studies from these locations found little concern for government regulations, except from those enterprises in targeted locations and specific sectors such as food processing. Instead, the greatest concern for the majority of those surveyed was the lack of access to working capital, credit and finance.

Issues Raised in Research on Developing Countries

This brief review of empirical literature on finance and SME development in low income countries does raise a number of issues concerning the underlying assumptions in these types of investigations. These assumptions relate both to the theoretical and institutional context in which propositions about the nature of SME finance are being examined. Three issues would appear to be particularly relevant to the current analysis of SME finance in developing countries.

The first concerns the macroeconomic context in which financial reforms have been implemented. The second concerns the assumptions implicit in models of credit supply, in particular bank credit to SMEs in low income countries. The third is related to the relationship between corporate governance and enterprise finance that is assumed in studies of low income countries.

Macroeconomic Context. Much of the analysis of the impact of financial liberalization on SMEs has been done in the context of a simple interpretation of the process of financial repression and liberalization. The notion behind financial repression was developed by Goldsmith (1969), McKinnon (1973) and Shaw (1973). These authors concluded that there was a close relationship between financial development and economic growth. In their view, improved financial intermediation would increase the supply of credit and result in greater investment. Government intervention to control the pricing and allocation of credit, referred to as financial repression, would restrict financial development. This kind of analysis provided much of the theoretical underpinning for the introduction of financial reforms in low income countries in the 1980s and 1990s. It was argued that removing controls on interest rates and credit allocation would increase savings and improve the efficiency of investment.

However, both the experience of financial development and new theoretical insights linked to endogenous growth theory suggest financial markets play a different role in relation to growth than that portrayed by the so-called financial repression school. Financial repression in low income countries has led banks to underinvest in information capital (Caprio, 1994). Other economic reforms, such as trade liberalization, may have rendered uneconomic much of the information capital that is possessed by banks, which is, in any event, limited in low income countries (Griffith-Jones, 1998). As a result, banks are likely to lend relatively little, particularly to SMEs, which will ultimately reduce growth. Simply raising real interest rates as part of financial reform will not overcome this problem. In this situation, measures are required that reduce the cost of information capital to the financial sector. These include a wide range of measures to improve legal structures, audit processes, and accounting systems. These measures can be considered as pre-conditions for successful liberalization and have largely been bypassed by the repression school when designing financial reforms.

Further, credit rationing may continue to persist after financial liberalization to the detriment of lending to smaller enterprises. It has been argued that even in competitive credit markets, information asymmetries continue to keep up the costs of assessing the riskiness of investment projects, which are raised with interest rates, causing borrowers to seek projects with higher rates of return (Stiglitz and Weiss, 1981).

Supply of Finance. It is evident that banks play a key role in the financial system by pooling the liquidity risk of depositors and investing a large proportion of their funds in more liquid, but more productive projects (Griffith-Jones, 1998). In much of the financial literature, the principal-agent model has been used to rationalize the low level of bank lending to SMEs relative to larger enterprises. As developed, the application of principal-agent theory argues that banks have less perfect information on smaller firms than larger firms (costs of gathering this information are higher) and, as a consequence, lending to smaller firms is riskier. The observed outcome from this analysis is less lending to small firms relative to larger ones. In turn, lending institutions are likely to demand higher risk premiums.

In the developing country context, however, it may be argued that banks have better information to assess the riskiness of an investment than the small firm itself. This is because they are continually lending to small firms over extended periods of time and have acquired sufficient insights to be able to make sensible and sound judgments over lending decisions. Banks may have more experience about a small venture's survival prospects than they have information on larger firms, since the latter may be in a better position to conceal and manipulate information to their own advantage. Measuring the extent of information asymmetries is intrinsically difficult. Nevertheless, the low income country case may challenge the applicability of principal-agent analysis in terms of the conventional forms established in an industrialized country context, and lead to new insights in which lending institutions, rather than being seen as villains, are acting in the interests of small firm development. There is some empirical support for this argument from examining lending behavior of indigenous banks in Africa (Harvey, 1996).

In low income countries it is widely recognized that an imbalance exists between the demand and supply for finance, with the former exceeding the latter. An initial response to this situation was to adopt a supply-led approach by developing specialized development finance institutions that would provide funding for the fixed investment requirements of enterprises of all sizes. Financial liberalization and critical evaluations of these institutions have transferred attention to the development of capital markets as alternative sources of long-term investment financing. Hence, an active and broad-based capital market is expected to mitigate the acute shortage of term loans and equity financing (Nissanke, 1996).

In the context of low income countries, however, capital markets take a long time to develop and deepen and do not necessarily work in predicted ways (Singh and Weisse, 1998). Often, the lack of financial instruments and the number of participants restricts the capacity for financial deepening. In low income countries, many family-based indigenous enterprises are reluctant to change their corporate status. The development of capital markets is predicated on introducing efficient and effective mechanisms of regulation and supervision to avoid malpractices. If these are not in place, then the markets will fail to assess any intermediate risks appropriately, so that new capital does not emerge. As a consequence, bank finance is likely to remain the main source of finance for SMEs.

Corporate Governance and Enterprise Finance. It is clear that the Anglo-Saxon model of financing enterprises, with a heavy reliance on bond and equity markets, is less relevant to smaller enterprises and does not conform to patterns of SME financing in low-income countries where internally generated funds and bank finance are predominantly used. The relation between bank finance and SME development is frequently considered in terms of the premium paid for bank finance and the borrowers net worth (Gertler and Rose, 1994). Smaller enterprises pay higher premiums for bank finance where collateral is lacking and information to evaluate risks is imperfect. It is generally argued that borrowers' net worth measured by current assets and potential future earnings is positively correlated with economic development, while premiums attached to finance are negatively correlated to an enterprise's net worth. As a consequence, the problems of financing SMEs ought to decrease with economic development over time.

Clearly, the conditions required for easing the financing problems for SMEs in low income countries are not readily met. Economic growth is not assured and the banking system may be weak. Some elements of the supply problem are catered for through innovative forms of informal sector financing, but these are unlikely to be sufficient to meet the fixed and working capital requirements of growing SMEs. This is even the case in Asia, where considerable specialization in financial services between formal and informal sectors takes place (Biggs, 1991).

As already indicated, most of the theoretical work on small firm finance and the behavior of institutions that lend to small scale enterprises has been undertaken in the industrial countries (Chittenden, Hall & Hutchinson, 1996). A large proportion of this work has tended to concentrate on firms that, in terms of size, lie toward the upper end of the spectrum, where the range of ownership and financing options becomes greater.

In general, two areas of research have become prominent. First, there are studies that have attempted to examine the implications of different financial structures found in different sized firms. These studies are based, in part, on survey work that has attempted to catalog the range of finance sources available to smaller firms and to examine their implications for growth and investment. In firms where forms of formal equity holding have been employed, this work has been extended to incorporate a number of distributional issues concerning income flows to owners and managers and inside and outside shareholders (Myers, 1998). Much of this analysis has been set within the work of a principal-agent approach. The conditions under which each respective interest operates are examined with reference to the internal incentive systems that emerge in firms and to the external factors, such as the macroeconomic policy environment and the development of legal systems, that offer potential protection to outside investors in firms (La Porter, Lopez-de-Silanes, Shleifer & Vishny, 1998).

Second, there has been a concentration of theoretically based studies examining the behavior of various lending institutions as suppliers of finance to small and medium-sized enterprises. Typically, for small enterprises, these have involved models of lending behavior based on an agency work. Central to the hypotheses that have emerged from this body of research is the notion that information asymmetries lead to sub-optimal flows of finance available to smaller firms compared to larger firms. Imperfect information can lead to restricted flows of finance whether the problem lies within the firm, through poor record management, or in banks, through the relatively high costs associated with gathering information on smaller firms (Binks, Ennew and Reed, 1992). There is considerable debate over whether or not banks in low income countries have a comparative advantage in lending to smaller firms precisely because they may possess an accumulated knowledge concerning the riskiness of investing that places them in a position to make optimal rather than sub-optimal lending decisions. Building relationships with banks increases the information flow between lender and borrower (Berger & Udell, 1995).

The emphasis on the relative lack of theoretical work ought not to imply that the stock of knowledge gained about finance and smaller enterprises through empirical work is not valuable. Considerable insights have been gleaned from a wide range of empirical investigations (Hall, 1992; Kaplan & Zingales, 1995; Cosh & Hughes, 1996). In terms of work in the UK, some general conclusions have emerged and are summarized in Table 1. In order to enable comparisons, the likely situation in low income countries is presented on the right hand side of the table.

The general conclusions shown in the left hand column in Table I are drawn primarily from work in the UK, and have either been derived from hypotheses that have been subjected to empirical testing or have resulted from direct observation and measurement. Theories have been developed to explain them. Most of this kind of analysis continues to be undertaken in the context of the industrialized countries, which raises a number of issues concerning its relevance and applicability to the low income country case.

It was indicated earlier that the research into the relationship between enterprise finance and development is limited in the context of low income countries. An attempt is made in Table 2 to select a number of approaches relating to small enterprise finance and examine the elements that may contribute to the development of research in low income countries. Column 1 provides a summary of each theoretical perspective while columns 2-5 respectively outline the implications for the financing and capital structure of small firms, the implications for their growth, the hypotheses that can be tested, and finally the factors that need to be considered when applying the approaches to low income countries. The approaches summarized in Table 2 incorporate results from empirically-based studies (Bates, 1971); life cycle models (Weston & Brigham, 1981); pecking order approaches (Myers, 1984); principal agent models (Jensen & Meckling, 1976); transaction cost hypotheses and models of financial sector reform.

Simply examining the way in which the demand or supply of finance changes in response to policy changes may not be enough to indicate what kinds of enterprises will survive and contribute to economic growth. A useful perspective may be adopted by attempting to implant financial elements into an approach adopted by Dawson (1993) in an earlier study of Ghana and Tanzania. In this study, an attempt was made to identify a set of characteristics embodied in small enterprises that were either favored or disfavored by the adjustment process. Table 3 indicates the importance of technical and technological factors in determining the way in which small enterprises are likely to respond to structural adjustment. The relatively more technologically sophisticated enterprises appear to have been better able to upgrade their products and services to a level where they are able to develop linkages with the faster growing sectors of the economy. Further, they have been able to overcome scale constraints by finding new so-called niche markets more suited to their economies of flexibility and serving an import-substitution . In contrast, enterprises that were technologically weaker tended to remain dependent on low income groups for their main markets, where purchasing power may have been declining under adjustment.

What this type of analysis and the comparison between Ghana and Tanzania clearly pointed out, however, was not the importance of the cross-country differences in the response to adjustment processes, but variations between different types of enterprises within the countries. As pointed out above, some enterprises were better able to respond to the changes created in the policy environment than others. In this respect the creation of an environment favoring small scale enterprise development may not be sufficient to explain why some enterprises improve their performance and others fail. The response of small scale enterprises to structural adjustment reforms is more likely to be related to a wide range of factors, including internal enterprise characteristics, the level of development of the economy's infrastructure, and the institutional and political work that reinforces the sector's development.

This review has pointed to the weaknesses and gaps in our knowledge concerning the relation between finance and SME development. At least four strands of research can be identified that will contribute to a better understanding of the financing needs of SMEs and the ways to deliver financial services to them.

First, research is needed on the forms of finance used by small and medium-sized enterprises and made available by lending institutions and investors. In particular, a clear picture is required of the financing differences between firms of different sizes and the differences in financing in relation to types of ownership structures. Cross-country and regional differences may also exist in these respects. Second, research is required into the relation between different financial forms and firm-level performance. Existing research on `small size and performance has not isolated the importance of different forms of finance. Methods should be devised to examine the relationship between different financial structures of firms and a range of performance measures (including output, productivity, employment, and survival rates).

Third, research is required relating to the behavior of small and medium-sized firms with different forms of finance. We need to predict how different forms of finance will affect the allocation of profits between income (dividend flows), investment and consumption and their effect on other forms of expenditure relating to innovation, marketing and the development of human resources through training. In particular, the links need to be made between different forms of finance and the impact of small firm development on poverty alleviation. Fourth, research is required on the supply side of finance. This work should focus on formal and informal sector lending institutions and savers, and the macroeconomic environment, including economic policies, promotional policies, and the role played by private, international and non-governmental organizations.