Although the EC definition appears to offer a reasonable basis for making international comparisons, from a conceptual standpoint, it is important to keep in mind those qualitative size-related characteristics which can influence organisational behaviour. Apart from independent ownership, which can result in a more limited internal resource base (particularly finance and management resources) than is the case in large firms, these characteristics include: a lesser ability to shape or control relationships with various aspects of the external environment (such as customers, suppliers, labour markets, sources of finance) compared with larger firms; and a coalescence of ownership and management, which is often associated with distinctive types of organisational behaviour. Whilst these are commonly recognised characteristics of SMEs, it is important to emphasise the heterogeneity of the SME sector, even within the context of a single national economy.

The Potential Roles of the SME Sector

Developing a SME sector is one of the issues facing Central and East European (CEECs) countries in their transformation from centrally planned into market based economies. The potential role of SMEs includes: generating employment and thus contributing to absorbing any labour surpluses which result from economic restructuring; contributing to the development of a diversified economic structure (including their role as suppliers to larger companies); contributing to the trade balance through export earnings or import substitution (see Smallbone, 1998); and, in some cases, as a source of innovative activity, thereby acting as a source of change in the market (Acs and Audretsch, 1993).

Whilst these are roles which SMEs can perform in any economy, there are additional aspects of the conditions pertaining during the transition period which suggest that SMEs may have additional roles to play. For example, in the Baltic States, the development of SMEs can contribute to economic adjustment from highly concentrated structures, that were overly focused on manufacturing industry based on mass production methods and relatively inflexible production processes, to more flexible production systems which include a wider range of consumer services. During the socialist period, manufacturing in the Baltic states was integrated into a technological chain, and locked into markets in the former Soviet Union, which resulted in a high level of military orientated production at the expense of consumer goods. As a consequence, there is a need to radically restructure the economic base in order to develop a more flexible system of production which is responsive to changes in consumer demands and the forces of market competition. In this context, sectoral restructuring is to a considerable extent dependent on the creation of many new enterprises.

Transforming a centrally planned, socialist society into a liberal, democratic market based system involves fundamental social change as well as economic restructuring, which SMEs can also contribute to. The creation of an entrepreneurial class is part of that process as the Lithuanian government has explicitly recognised in one of the stated aims of its programme for SME development: "the development of a new category of businessman in society" (Lithuanian Ministry of Economic Affairs, 1994). Alternatively, as Piasecki points out: "At an early stage of transformation, the development of the SME sector becomes one of the most cost effective instruments in the re-orientation of social awareness;...... without liberating social awareness, the emergence of the private sector and a market economy is impossible" (Piasecki, 1995).

SME development in transition economies can also contribute to the process of privatisation and/or the restitution of property. This may involve contributing to bringing privatised assets into productive use as an alternative to or following their liquidation, representing a means by which resources which would otherwise lie dormant can be brought into production, and so increase output at marginal opportunity cost. As the OECD (1996) noted: "Whilst the process of privatisation provides an essential contribution to the transformation of property rights, the micro-economic restructuring of former state enterprises can only be successful if entrepreneurship is fostered in these firms or their successors". In this context, many manufacturing SMEs in the Baltic States commenced trading using equipment and premises that were leased initially from state-owned companies (Smallbone et al, 1996).

However, a more negative view of the role of small businesses in the transformation process has been presented by Scase (1997). His core argument is that in assessing the role of small businesses in transition economies, it is necessary to make a distinction between 'entrepreneurship', which refers to a person's commitment to capital accumulation and business growth, and 'proprietorship', which refers to the ownership of property and other assets that may be used to realise profits but are not utilised for the longer term processes of capital accumulation. In fact, Scase emphasises that 'proprietors' are more likely to consume any surplus that is generated rather than re-invest it in their businesses. He argues that it is 'proprietors' rather than entrepreneurs which account for the majority of small businesses that are established in Eastern Europe, as individuals strive to protect themselves from the uncertainties characteristic of emerging market economies, by generating cash flows that can be consumed for raising general living standards, rather than for business expansion.

Whilst Scase is right to draw attention to the fact that only a minority of small business owners can be considered 'entrepreneurs' in the classical sense, as he recognises himself, this is not a characteristic that is confined to economies in transition (Scase & Goffee, 1982). One of the consistent themes emerging from the western literature on SME development is the need to recognise the variety of motives of those who start and run small businesses, which includes lifestyle reasons in many cases. As a consequence, not all SME owners are aiming to grow their businesses (Curran, 1986). Moreover, SME development in a transition context is more than an economic process. As Scase himself recognises, in transition countries the emergence of a stratum of small traders is part of the social transformation contributing to wider consumer choice and the emergence of a middle class. As far as economic development is concerned, the point to stress is that the role of SMEs varies at different stages of the transformation process as well as in different national contexts.

Structure of the Paper

The rest of the paper is divided into 3 main sections and a conclusion: section 2 considers the role of SMEs in employment generation in mature market economies; section 3 focuses specifically on their contribution to employment change in transition countries; section 4 is concerned with the local/regional or sub-national dimension to SME development, focusing on the implications for policy.


2. SMEs AND EMPLOYMENT GENERATION IN MATURE MARKET ECONOMIES

The Evidence

Ever since the publication of the Birch Report (Birch, 1979), it is the contribution of SMEs to employment generation which has been the main focus of attention of policy makers at the local, regional and national levels. More recent work undertaken in the USA has suggested that during the 1976-88 period, firms with less than 20 employees provided 37% of net employment creation in the US at a time when they provided 19% of total employment (US Small Business Administration, 1992). Although less dramatic than the earlier Birch findings, this study reinforced the original message with respect to the contribution of (very) small enterprises to new job creation.

Moreover, various studies in Britain and other developed economies, that include a variety of timespans and both recessionary and non-recessionary conditions, have confirmed that smaller firms have been making a disproportionate contribution to net employment creation (Commission of the European Communities, 1987; Storey and Johnson, 1987; Sengenberger et al, 1990). Although job losses, not surprisingly tend to increase during recessionary periods (such as 1989-91 in the UK), and the number of job gains decreased, SME performance has been shown to be relatively stable in comparison with larger firms over the economic cycle (EIM, 1994).

The increasing share of SMEs in total employment that occurred in many western countries during the 1970s and 1980s reflected growth in employment in SMEs at a time when many large enterprises were reducing jobs, as a result of a number of factors. The emergence of computer based technology in production, administration and information have reduced the role of economies of scale in many sectors. At the same time, rising living standards have contributed to more fragmented consumer markets, reflecting changing patterns of consumer expenditure and demand patterns. Moreover, as many large firms have downsized their activities in an attempt to reduce costs, this has contributed to new business opportunities (and thus employment creation) in SMEs. However, whilst the share of SMEs in total employment has increased since the 1970s in many western countries, it is very small enterprises which have most consistently created more jobs than large enterprises, in the aggregate at least (EIM, 1997). Part of this trend has been a growth in self employment which was a key feature of the labour markets in a majority of EU countries in the 1980s, reflecting a structural shift away from manufacturing and a combination of other socio-economic factors, such as changing tax regimes, increased contracting out of services and the development of more flexible manufacturing systems (EIM,1994).

An On-Going Debate

Whilst there is little doubt that, in aggregate terms, small firms have increased their share in employment and economic activity at a faster rate than large firms, the precise nature of their contribution has been the subject of much debate. A number of the issues which have potential implications for policy, are discussed below:

1) The evidence on the contribution of small firms to employment generation based on the Birch methodology has been challenged, which has led some authors to question the case for providing policy assistance on this basis. For example, in a recent review of the small firms/employment debate, Hughes concluded that: "by themselves, these studies offer no basis for the specific promotion of the SME sector by direct or indirect means" (Hughes, 1997).

2) A substantial issue that emerges from this methodological debate concerns the number of SMEs which are employment generators. On the one hand, some research evidence has suggested that overall job growth in the small business sector in the UK was due to increased employment in a large number of firms (Daly et al, 1991). On the other hand, there is a growing body of research evidence which suggests that relatively few SMEs actually create employment on a significant scale. For example, empirical studies in the UK suggest that 5-10% of firms surveyed over a time period will typically account for 40-50% of the total employment generated by a given cohort of surviving firms (Storey, 1994). In other words, such job generation that has occurred in the SME sector in recent years has been dominated by a few fast growing firms. The issue is a potentially important one for policy makers since it raises the question of the extent to which there is a case for targeting those firms which have job generating potential and if so, how this can be achieved (Storey & Johnson, 1987). At the same time, it should be noted that survey evidence from Poland and Estonia showed that the majority of manufacturing SMEs surveyed (400), increased jobs during the 1994-96 period, although in Latvia and Lithuania there was a much more varied and turbulent pattern. In other words, the net increase in employment was not a result of a few fast growing firms in the sample but was more broadly based. Interestingly, another difference compared with most western sample, was the disproportionate contribution to job generation made by firms in the 21-50 size band. However, it is important to stress that these were mainly established firms, since firms trading for less than one year were not included in the study.

3) This UK evidence raises the question of the characteristics of SMEs that are the main employment generators. In terms of age, the conventional wisdom is that young firms tend to create more jobs than older firms (Storey, 1994), since "most small businesses only grow in the first few years after start-up and then stabilise (Burns, 1989). Moreover, UK evidence shows that younger firms grow faster than other firms, with the smallest young firms tending to grow faster than the rest. At the same time, some employment created by new enterprises can be rather unstable because of higher failure rates that is typical of young businesses, with only 50% reported to survive until the end of year 5 (EIM, 1997). As Hughes (1997) has noted: "The world of the small business is exceptionally turbulent...... within this turbulent world, the very smallest firms are 6 times more likely to de-register (ie not survive) than the rest. In terms of size per se, although the very smallest firms grow faster than other firms, it appears that once a certain size threshold is passed, growth is not systematically related to size. In the UK for example, this threshold may be as low as 8 employees (Dunne and Hughes, 1994).

However, research undertaken at Middlesex University (based on a longitudinal study of manufacturing SMEs over an 11 year period) showed that if firms can survive the critical early years after start-up, they are capable of generating additional jobs over an extended period of time (North et al, 1994). One of the key findings of this longitudinal study was that job creation during the period 1979-90 was closely related to the growth of firms in output terms (measured in terms of the real value of sales turnover). Employment growth was particularly concentrated in the best performing high growth companies: 23% of firms accounted for 71% of the increased employment between 1979-90. Thus from an economic development point of view, it would appear that the best way of generating employment in the longer term is to focus on firms with the greatest growth potential. Whilst for entrepreneurs, growth objectives are more typically defined in terms of sales or profits than in terms of employment, policy makers need to recognise that employment growth is a consequence of significant output growth.

4) Another issue concerns the type of jobs which are created in small firms since it is not just the number of jobs that are important from an economic development perspective but also the quality of employment provided. In the 1980s, some development organisations in the UK (such as the Greater London Council) were reluctant to support small businesses because they considered that the jobs provided were of a poorer quality than those in larger companies. Indeed, such a view is supported by a review of the research evidence on the issue which concluded that "the evidence from both the UK and the USA suggests that, according to most measures, the job quality provided by small firms is lower than that in larger firms......wages are lower, training is less frequent and the evidence for a compensating higher level of job satisfaction is weak" (Atkinson & Storey, 1993).

However, as Johnson (1991) points out, lower wages in the small firm sector may, to some extent, be offset by other benefits of working in a small business, such as flexibility of hours, or more personalised relationships with the employer. Moreover, as other authors have emphasised, some of the apparent differences in the nature of employment between large and small firms may be associated with factors other than size. For example, service employment in general is characterised by high levels of part-time and casual working compared with manufacturing (Hakim, 1987) and one estimate has suggested that 90% of all small businesses in the UK (defined as employing 1-24 employees) are in the service sectors (Curran & Burrows, 1988).

Our own evidence from employment in manufacturing firms tends to show SMEs in a relatively favourable light in this respect (North et al, 1994). We found that the majority of manufacturing SMEs made only marginal use of peripheral sources of labour (such as part-time workers, homeworkers, and casual workers), and occupational wage levels were not consistently below national or regional averages. The conclusion is that many SMEs do provide good quality jobs, and with the right kind of support SMEs are capable of improving their working conditions and pay levels, especially when this goes hand in hand with business growth.

5) There has been much debate in the West about the extent to which the increase in employment in small firms represent additional jobs for the economy or simply a transfer of jobs from large to small firms. It has been argued that large firm fragmentation and externalisation strategies, have contributed to increased opportunities for small firms as subcontractors, franchisees or as suppliers of services which were previously provided in-house (Shutt & Whittington, 1987). For example, externalising selected manufacturing processes to small companies, or replacing in-house cleaning or catering staff with a contract to an outside supplier, may contribute to increased employment in small firms but this is at the expense of jobs in the contracting organisation.

Whilst there has undoubtedly been a shift towards more fragmented modes of production through the 1980s and 1990s in mature market economies, such as the UK, empirical evidence to support the "employment transfer" hypothesis is not conclusive (Keeble, 1990). Moreover, even if a transfer of employment is taking place, it can be argued that competitive economies increasingly need to be organised in ways which are different to those in the past, and a change in the nature of relationships between large and small firms may be a necessary condition if competitiveness is to be maintained. In other words, a long term, dynamic view of the restructuring issue and of the role of SMEs within it, needs to be taken rather than simply a short-term one.


3. SMEs, EMPLOYMENT GENERATION AND ECONOMIC DEVELOPMENT IN TRANSITION ECONOMIES

Empirical Evidence

The absence of reliable aggregate statistical data on SME development in CEECs make it difficult to produce reliable estimates of the employment they represent, although those data that are available suggest they are making a significant contribution. In Poland for example, official statistics show that the share of total employment in the private sector increased by 14% between 1990-95, accounting for 52% of total employment by the end of this period (Polish Ministry of Industry and Trade, 1995). However, it should be noted that the Polish Enterprise Register included 2.1m registered enterprises in 1995, of which, according to reliable informants, only between 1.2m and 1.5m were actually trading. A similar problem exists in the Baltic States and in other transition countries where business registration data typically overestimates the number of trading enterprises for a variety of reasons. Whilst data limitations typically make it impossible to accurately determine the share of economic activity or employment contributed by SMEs in transition economies, available estimates confirm the continuous expansion of the SME sector in the Baltic States during the transition period. According to business registers, at the beginning of 1995, about 55,000 businesses were registered in Estonia, 86,000 in Latvia and 120,000 in Lithuania; almost 90% of which were privately owned in all 3 countries (Smallbone et al, 1997). More generally, a number of empirical studies suggest that small private enterprises are a source of employment growth in transition economies although the extent to which this is so varies between countries.

For example, a recent study of manufacturing SME development in Poland and the Baltic States concluded that provided they can demonstrate their ability to survive and become established, manufacturing SMEs are capable of generating employment (Smallbone et al, 1996). The survey included indigenously owned private firms employing less than 100 and the employment change data referred specifically to the period between December 1993 and June 1995. In fact, the results showed considerable variation in SME performance (both in terms of employment and sales growth) between the four countries surveyed, reflecting differences in the pace of transformation. On the one hand, in Poland and Estonia there was a net increase in employment of 29% and 21% respectively during this 18 month period and the majority of surviving firms were able to increase employment. By contrast, in Latvia and Lithuania the pattern of employment growth was much more mixed and firms that were in existence throughout the 18 month study period actually experienced a net reduction in employment of 4.9% and 6.0% respectively. In fact, a similar comparison between the 4 countries emerged when we compared sales growth performance in 1994-5 and also export sales.

The explanation is that external conditions in Latvia and Lithuania were slower to stabilise than in either Poland or Estonia. Although the number of private enterprises grew rapidly, the external environment presented many threats as well as opportunities for entrepreneurs. In addition, there were still some privatised firms in Latvia and Lithuania where the process of restructuring was not complete, which had implications for employment. These findings emphasise the importance of external conditions in influencing the ability of SMEs to generate employment, suggesting that their contribution may vary at different stages in the transformation process.

Further empirical evidence at the micro level is provided from survey of 394 small firms in Bulgaria, undertaken in 1993 (Bartlett and Rangelova, 1997). Although the sampling frame included firms with up to 50 employees, the vast majority of surveyed firms were very small enterprises with less than 10 employees. The study showed that more than half the firms surveyed (59%) reported taking on new employees in the 6 months prior to the interviews, and two thirds claimed to be planning to expand employment during the subsequent year. Whilst recognising the inability of the private sector to compensate for the rising unemployment in Bulgaria that has resulted from the contraction of jobs in state firms, the authors emphasised the dynamic contribution of small private firms to the development of the Bulgarian economy and its important role in employment generation.

Variations in the Pace of SME Development

The development of SMEs during the transition period, and the associated employment that this creates, varies between countries according to a number of factors. These include: the starting point for transition, in terms of the recent tradition and experience with private enterprise affected the entrepreneurial orientation of the population during the transition period; the processes that have operated during the transition period (such as the form of privatisation used); the extent to which macro-economic conditions have stabilised; and the attitude and policy stance of government towards the development of private sector activity and market reform. Clearly, such factors can affect the nature and pace of SME development with implications for their ability to create employment.

Not surprisingly perhaps, in most CEECs, the early stages of transition to a market-based system saw a rapid increase in the number of small private enterprises as administrative reforms enabled latent entrepreneurship to be translated into capitalist enterprises. With respect to Poland, Piasecki and Rogut (1993) have characterised the development of the small business sector in Poland in terms of a 3 stage typology:

1) An initial phase of development of entrepreneurship (1981-88). This began under the centrally planned system following reforms in the early 1980s, which involved modification to the administrative and legal barriers to market entry thus paving the way for small business development. As a result, the number of private sector firms grew from 357,142 in 1981 to 572,451 in 1988. Nevertheless, they were still operating within the framework of the centrally planned economy and many of the entrepreneurs at this stage are said to have come from within communist party circles (Blawat & Dominiak, 1994).

2) A period of explosion of entrepreneurship (1989-91). Following two pieces of legislation introduced in 1989, the number of private firms is said to have increased sharply from 960,000 in 1990 to 1.7m by 1993 (Welter, 1995). This legislation abolished the legal and administrative barriers to private firms thus enabling them to operate on an equal basis with state owned companies. Other factors that contributed to the rapid growth in new business activity included the creation of a new socio-political climate following the elections in June 1989. In addition, there were economic factors which included the limited nature of competition in many markets, market opportunities created by shortages of certain products, and a latent consumer demand for a more varied range of products than had been available hitherto. In some cases, the collapse of state companies left market niches for small firms to exploit which extended beyond the state companies main sector of activity (Karawoska & Mrozinska, 1993).

3) A slowing down in the rate of development of the SME sector since 1991. Piasecki and Rogut interpret this slowing down as the "new" emerging market system beginning to regulate itself by which they mean the effects of increasing competition between firms on survival rates and a slowing down of start-ups as new market opportunities became more difficult to find. However, additional contributing factors included: a decrease in consumer purchasing power, an increasing tax burden and rising interest rates.

Although the details of the Piasecki and Rogut typology are specific to the Polish context, certain aspects are transferable to other countries. For example, in the Baltic States, where the existence of private enterprises was not tolerated during the Soviet period, the recent development of the SME sector has its origins in the co-operatives that were established after 1987 (as part of the initial privatisation of state owned industries). In Baltic countries, the explosion of entrepreneurship that was seen in Poland between 1989-91 occurred after 1991 when the reforms were introduced that enabled private enterprises to legally exist. During the initial stages of the transformation process, the main contribution of SMEs to employment was mainly a result of the new firm births that resulted from either de novo start-ups or as a result of the privatisation of former state-owned companies.

Distinctive Characteristics of SMEs under Transition

Another issue that needs to be considered when assessing the contribution of SME to employment generation in transition economies is the nature of the SMEs themselves which, particularly in the early stages of transition and in those countries where transformation is less advanced, have a number of distinctive characteristics in comparison with SMEs in mature market economies. These characteristics include: a high level of multiple ownership; high rates of non-survival; frequent changes in the core business activity, reflecting a lack of a clear product focus; an emphasis on trading and services rather than on manufacturing because of the need to generate a rapid return on capital employed; and a high level of informal sector activity. Most of these characteristics reflect a lack of political and economic stability in the external operating environment but there are implications for the nature and extent of the employment created. For example, recent survey results from the Ukraine, based on a survey of 350 small firms drawn from the manufacturing, construction and service sectors, showed a higher rate of growth in the number of temporary than permanent employees among surveyed firms since they were founded, reflecting their lower cost and the uncertainty caused by the unstable economic environment. Nevertheless, as our results from Poland and the Baltic States show, once the macro-economic and political environment stabilises, there is greater encouragement for entrepreneurs to develop their businesses in ways which are associated with creating employment for core or permanent employees (Smallbone et al, 1996; Smallbone, 1998).

As macro-economic and political conditions stabilise, there are other characteristics of SMEs in transition economies which appear much more positive from an economic development perspective. Various studies have shown that one of the distinctive characteristics of SME owners in CEECs is their higher education levels in comparison with their counterparts in the West (Blawat and Dominiak, 1994). A survey of 600 manufacturing SMEs in Poland and the Baltic States in 1995 showed that more than half the main owners were graduates, or equivalent: 60% in the Lithuaninan firms; 54% in Latvian firms; 49% in Polish firms; and 48% in Estonian firms (Smallbone et al, 1996). Although graduate education is not a guarantee of business success, it does tend to influence management behaviour, including a greater willingness on the part of SME managers to make use of external assistance in managing and developing the business, than in businesses led by managers with a non-graduate, 'craft' background. In the UK for example, a key issue affecting the impact and effectiveness of SME policy is the negative attitude of SME owners and managers with respect to the use of external assistance, whether this is offered by private consultants or business support organisations. In transition countries, such as Poland and the Baltic States, our evidence suggests that the problem in this respect is more a question of supply side deficiency than a lack of latent demand.

Differences in the attitudes of SME owners and managers are also reflected in the much higher propensity who appear to recognise their own management weaknesses than is typically found in western samples. For example, in our 1995 study more than three quarters of Baltic SMEs surveyed identified at least one area of management weakness (particularly in sales and management); in Estonia, the proportion rose to 89%. Encouragingly, the vast majority of firms where managers identified management weaknesses expressed interest in upgrading the firm's management skills. These results have implications for policies designed to enhance the contribution of SMEs to economic development. Clearly, access to an effective system of business support and management training can help to increase the (international) competitiveness of SMEs, as many governments in EU members states recognise.

Different Stages of Transformation

Finally, the nature and extent of the contribution of SMEs to economic development and employment generation is related to the external conditions in which they are operating, including the stage of transformation reached. In this respect, a distinction can be made between countries where reforms towards a market economy have been slow, or have not been properly installed, and countries which are more advanced with respect to market reforms. In the first group, the biggest obstacle for SMEs is the fundamental lack or inadequacy of those legal and financial institutions which are needed for a market economy, which often reflect a lack of real commitment and support for private enterprise on the part of the state. Private property rights are not properly guaranteed, laws relating to private enterprises (such as commercial codes and bankruptcy laws are still missing) or are highly inadequate. In addition, the financial system that is required for financing private economic activities is still missing for the most part. Belarus is a good example of this category but so too are Moldova, Ukraine and to some extent Russia. In such conditions, whilst SMEs may be generating employment, much of it may be relatively transient as external conditions do not encourage, or indeed facilitate, entrepreneurs to build businesses. In these conditions, the priority is to establish the institutional, legal and cultural conditions to facilitate the development of entrepreneurship.

In countries that are at a more advanced stage with respect to market reforms, the legal framework is no longer the most immediate problem for SMEs. In such cases, SME development is more impeded by a financial infrastructure which is still adjusting to the needs of the emerging private sector, distorted competition in some sectors, and enormous deficiencies regarding the business support infrastructure. Central European countries such as Poland, Hungary and the Czech Republic fall into this second category, as do the Baltic States with Estonia most clearly so. Whilst in these conditions, creating an enabling environment may remain the main priority, there may also be a case for direct support measures that are demand-driven. Some western governments use a concept of 'market failure' to justify this approach (Department of Trade and Industry, 1995). National governments in transition economies may need to accept this role if SMEs are to achieve their potential contribution to the transformation to market economies.


4. THE LOCAL/REGIONAL DIMENSION

Spatial Variations in SME Development

There is considerable evidence from mature market economies to suggest that significant spatial variations exist in the extent to which SMEs contribute to employment growth and economic development. In the UK for example, there are wide regional variations in new firm birth rates which between 1980 and 1990 were almost three times as high in London and the South East as they were in some of the peripheral regions (Keeble et al, 1993). Moreover, a cross-national comparison of new firm formation rates in a group of mature market economies, found similar regional variations within all 6 countries studied, with the most fertile regions having annual new firm birth rates between two and four times higher than the least fertile regions (Reynolds et al, 1994). Mason (1991) has explained these spatial variations in new firm formation rates in terms of differences in economic and occupational structures, entrepreneurial culture, and economic factors with respect both to demand and the supply of factors of production. Marked regional imbalances in SME development have also been shown to exist in Estonia, where in 1995 Greater Tallinn accounted for 54% of all registered businesses but only 29% of the total population (Smallbone et al, 1997). In terms of policy, concern with the gap between new business formation rates in Scotland and the average for the UK as a whole, led Scottish Enterprise (the agency with responsibility for promoting economic development in Scotland) to embark on a major programme to stimulate new enterprise development with the aim of closing the gap between Scotland and the rest of the UK in terms of new businesses created.

There is also considerable UK evidence to show that there are spatial variations in the ability of SMEs to create jobs, once they become established. For example, a number of studies have shown SMEs in rural areas to outperform their urban-based counterparts in terms of employment generation during the 1980s period (Keeble et al, 1992; University of Cambridge, 1992; North & Smallbone, 1995). Although the detailed explanations vary between authors, a common link is the role of the SMEs local operating environment in influencing its ability to grow and to create jobs (Vaessen & Keeble, 1995; North & Smallbone, 1996).

Since certain types of location appear more likely to stimulate new business activity than others and the ability of established SMEs to create employment also varies between types of location, there is a strong 'a priori' argument for SME support policies to be spatially differentiated; in other words, for an approach which is sensitive to the strengths and weaknesses of the local economy with respect to SME development.

Local/Regional Networks

Although the roles that SMEs perform in the national economy (see section 1) may also apply at the sub-national level, there are additional factors to be considered, although the weight given to them will tend to vary according to the theoretical perspective taken. For example, in the 'flexible specialisation' approach, it is suggested that an economy dominated by mass production is being replaced by more flexible production systems, which involves the decentralisation and vertical disintegration of large firms and the emergence of small firms in regional networks. Since the early 1980s, the SME sector has been attributed a key role in relation to endogenous local/regional economic development, based on the interdependence between SMEs. In this context, it has been argued that the competitiveness of local and regional economies depends to a considerable extent on the ability of SMEs within it to innovate (eg Cooke, 1996) which, it is suggested, is related to their inclination and ability to enter into interactive learning networks with other firms and various other external bodies and agencies (Morgan, 1997). Clearly, in such a view the local/regional dimension is central to the analysis with implications for policies designed to encourage and support network development.

The Delivery of Business Support

A local/regional dimension is also important in the delivery of support to SMEs. The ability of SMEs to survive and grow in increasingly competitive markets can be affected by the extent to which they are able to draw on appropriate resources from outside the firm when required. Since limited resources are one of the size-related disadvantages that SMEs face, they might be expected to turn to individuals and organisations outside the firm from time to time to assist with a variety of problems. In this respect, the aim should be to develop a business support infrastructure that is relevant to the needs of SMEs at different stages of their development. Business advice, information and counselling needs to be available at the pre-start and start-up stages, together with training to enable new entrants to have a basic understanding of business principles, accessed through local centres. However, once businesses become established, their support needs tend to become more specialised. Examples include information, advice and consultancy with respect to a range of issues including export intelligence, information and communications technology, and quality management. Since it may not be feasible to make all of the more specialised support available locally, access to these services may be facilitated through a single local entry point, or "one-stop shop" which has a signposting function for certain services. Such an approach has the potential advantage that it can help to integrate support programmes and service delivery at a local level, avoiding a proliferation and fragmentation of agencies which some countries (such as Britain) have experienced in past (Segal, Quince and Wicksteed, 1988). Business Links, introduced in the UK in the 1990s, represent an attempt to implement a strategy based on a "one-stop-shop", although until recently , the support for pre-start and start-up businesses has been largely outside their remit.

Although most transition countries have introduced some direct support measures to assist SME development, surveys show that the scope of this assistance is typically limited (eg OECD, 1996). In this respect, Welter (1997) explains the limited success of support centres for enterprise development in CEECs in terms of bureaucratic difficulties, problems with co-ordinating financial resources from a range of (donor) sources, and a modest response from SMEs themselves. Significantly however, Welter emphasises the need to adopt a decentralised approach to SME support and promotion, in which special emphasis is placed on the role of local and regional actors (Welter, 1997, p16). In this respect, she refers to the Hungarian local enterprise agency network as a good example of a programme involving local and regional actors.


5. CONCLUSIONS AND IMPLICATIONS FOR POLICY

Although the nature of roles assigned to SMEs in local economic development varies between different theoretical positions concerning local economic development, there is general agreement on two key points: firstly, that the development of a SME sector is an integral part of the transformation of a centrally planned into a market based economy; and secondly, that policies developed to promote and support SME development, need to have a local/regional dimension.

Certainly, in recent years in the UK and other mature market economies, SME development has been an important part of local economic development strategies, often combined with attempts to attract inward investment, whether from abroad or from other regions. Consensus has emerged among policy makers and practitioners around what has been described as a 'bootstraps' approach (Eisenshitz & Gough, 1993), in which local economic development institutions and initiatives emerge as the 'commonsense' of the age, although there has been academic debate about how local these strategies really are (Peck & Tickell, 1994). By contrast, more radical strategies based on the principles of flexible specialisation have emphasised the emergence of small firms in regional networks that resemble Marshallian industrial districts. In this type of approach, the emergence of local democratic structures is not simply seen as enhancing local democracy but also to act as a countervailing force in the face of the power of large firms, particularly TNCs. Despite these theoretical differences, which have implications for the type of policy required and the institutions to support it, there is a common acceptance of a role for SMEs in local/regional economic development which clearly presents particular challenges in the context of emerging market economies.

Another issue concerns the heterogeneity of the SME sector which results in a variety of roles for different types of firm, or groups of firm. In this context, the relative importance of policies designed to increase the rate of new business start-up and the survival of young firms on the one hand, and those which aim to encourage and support growth in existing firms on the other are likely to vary: firstly according to overall policy objectives (such as reducing unemployment v improving competitiveness), and secondly between different parts of the country, because of differences in the strengths and weaknesses of local economies. Nevertheless, at a national level resource scarcity, together with the differences that exist between SMEs in their contribution to economic development, would appear to justify a targeted approach. In view of the close connection between sales growth and employment growth, there is a strong case for linking policy support to a firm's growth orientation and performance. Moreover, in view of the importance of the generation of external income to national economic growth, there is a strong case for giving the highest priority to growth oriented firms that are seeking either to break into foreign markets, or increase their foreign sales.

As far as policy is concerned, it may be suggested that one of the lessons that can be learned from the experience in western Europe is the need to clarify what the objectives of policy are and on what basis intervention can be justified. This is more a question of learning from mistakes than from successes since in most EU countries, the approach to SME policy has been typically pragmatic. As a consequence, it is often necessary to infer policy objectives from the actions of policy makers (de Koning and Snijders, 1992). A more rational approach to policy development would seem to be:

1) Identify the reasons why SMEs may justify policies aimed to encourage and support them. From this it should be possible to establish a clear set of objectives for such policies.

2) Identify the types of policy and programmes needed to implement these objectives. This involves clarifying the size characteristics and constraints which SMEs need help in overcoming.

3) Identify the most effective delivery mechanisms for that policy in which the respective roles of national and local/regional government is clearly defined as well as those of the public and private sectors.