2. Can SMEs deliver on their promises?
From the critique developed above, it seems that this article has a straight answer to this question: no! If SMEs approaches are ill defined and analytically inadequate, and if the arguments for SMEs are too many and hence not clear, it would be difficult to see how SME approaches could deliver on their promises. Furthermore, it would be impossible to assess the results in any meaningful way, because SMEs are expected to deliver almost on anything. However, this is not a fully adequate critique of the lack of power of the SME approaches to deliver on their expectations because it would miss some important information.
One of the major problems of having one (ill defined) tool of policy as a panacea is that the likelihood of conflicting and empirically unproven objectives emerging in the process is very high indeed. Conflicting objectives in SME approaches are many - for example, SMEs are more often defined by the small number of workers employed, but are expected to address the issue of massive unemployment; or are defined as being localised and closer to local markets but, at the same time, as being fully flexible to changing market and competitive conditions (which would have to address the issue of globalization). The very (most common) definition of SMEs by the (small) number of workers employed raises the possibility that these firms are actually capital intensive. This could come into conflict with the goals of employment promotion, exploration of factor endowment led comparative advantages, promotion of national entrepreneurship and adaptability to local market conditions and demands.
Capital intensity would also raise the question of labour skills and training, which could undermine the goal of factor market flexibility.
Many small firms in the same industry are more likely to generate unproductive rent-seeking than if rents are allocated from the start on the basis of clear economic
indicators.5 An example of an agro-industry in Mozambique may help to understand the issue.
The large cashew processing factories were all closed in the late 1990s as a direct result of the liberalisation of exports of raw cashew. It is often argued that these firms were technological and scale inefficient, such that they were seeking rents through protection against exports of raw cashew. Liberalisation finished rent seeking and opened the doors for smaller firms, more efficient technologically and in terms of scale, to emerge. One such smaller firm, which initially benefited from significant business support, was successfully started in the North of Mozambique and has been developing well. This firm processes less than 4% of locally available cashew nuts and thus has the possibility of selecting high quality cashew. Its owner even argues that exports of raw cashew are good for his business because it is mostly bad quality cashew nuts (the largest proportion of cashew nuts produced in Mozambique) that are exported raw. Hence, he needs not seek protection against exports of raw cashew. Because of the success of the firm, its owner is doubling the capacity of the existing factory and plans to start a new one. Initial investment is cheap because of the labour intensity of the project but mostly because of highly subsidised business support that is given - hence, local business people are starting to invest in this processing industry, and about five new factories are underway. Additionally, there are many smaller factories that are being built.
The success of all of them depends, amongst other factors such as technology and management, on the possibility to select and buy cashew nuts of high quality. Thus, supply of good quality raw material is crucial for the success of the processing plants. Cashew nuts have been collected mostly by small holders, and there is no systematic policy to change this. The government and NGOs see support to small holders as a way to address rural poverty and a series of other development goals, but there is no evidence that small holders can supply quality raw materials to the factories in the foreseeable future. The processing capacity that is being installed will absorb more than one third of current supply of cashew nuts - this will create extreme pressures for quality cashew and will bring back the pressure to reduce exports of raw cashew.
Without changing the supply conditions of raw materials, the processing industry will fail. Some industrialists have anticipated this problem and are investing heavily of plantations of cashew trees, thus reducing and eventually breaking the links with local suppliers, and avoiding the uncertainty of competition for raw materials. Only some investors will be able to follow this road, and they will tend to grow and dominate the industry. Very small processing plants will only function as satellites of the new larger ones through subcontracting and will not be able to compete in the raw material market. Thus, the new larger firms will either seek rents in the form of protection, which will not be helpful in the long run, or will acquire rents through vertical integration of the industry. Most likely, there might be an interim period in which the two forms of rents are sought at the same time.
The very small processing firms will, most likely, pay an additional cost: they will compete for available cashew in the market, will sell to larger firms only the cashew of a given standard of quality, and will have to compete for access to the lowest layer of low quality market for the remaining cashew nuts. They will operate as a selection mechanism of the larger firms.
In brief the firms that are now starting as small and medium will grow large if they vertically integrate and use the very small firms as a selection mechanism. The very small firms will either disappear or remain in business as long as they are sub-contracted by the new larger firms.
This example shows three important points. First, rent-seeking is not necessarily avoided by firms being smaller (or larger). Actually, a clear strategy to support vertical integration and successful entrepreneurs in the industry in exchange for aggressive and successful penetration of external markets, may accelerate employment creation, export growth and building of industrial efficiency, and minimise rent-seeking. Without a market for rents there is no rent-seeking. Second, industrial growth often involves some sort of fallacy of composition: what is true to one small/medium firm my not be true to many small/medium firms. Thus, if the scale of operations does not increase - either through internal vertical integration or some other form of industrial association, network of partnership - industrial inefficiencies will develop and, with it, rent-seeking may increase. Third, business support has to be directed at growth of firms and operations, and continuous industrial development (including the development of networks, regulation of subcontracts and promotion of vertical integration), or firms will not survive and industrial restructuring will yield a high social cost through bankruptcies and unemployment.
Thus, the issue is not whether small and medium firms are supported as such, but whether they are capable of growing and restructure entire industries in the process.
More generally, the issue is not whether firms start large or small, and whether each industrial operation should be organised on large or small scale. The literature on Taiwan and South Korea, for example, has shown that industrialization and technological development can be equally successful irrespectively of economic strategies favouring larger (South Korea) or smaller (Taiwan) corporations to develop. Of course, scale is not only a technical issue. It is also political, social and economic, and also depends on how interest groups interrelate between them and within the state. However, what matters most is that the organization of industries and economies is based on sufficiently large scale to generate dynamic growth - be it through a highly internalised and vertically integrated Chaebol (like in South Korea), or highly structured industries and industrial associations
(more like in Taiwan).6
Anyway, in industrialising and industrialised economies, firms tend to grow. In the era of globalisation (born with capitalism), the growth of firms tend to include an international component. It does not really matter whether plant A belonging to firm Z is small. Is firm Z small? Is firm Z operating as a
small firm?7 Is it not that plant A, belonging to firm Z, benefits from firm ZТ‘s production, trade and financial network and
Empirical studies about the SME versus large firm debate are often inadequate. Many suffer from selection bias, which results from a practical problem of sampling - SMEs chosen are frequently those that have managed to survive, rather than a random selection that would include the ones that disappear and may not even be known, by the researcher, to have existed. Others face identification problems - what is a SME? Is the plant/firm behaving like a small firm or only being defined as a small firm?
Many studies are confused by the lack of clarity and precision that surrounds the SME issue - groups of firms are defined by the number of workers employed and then their performance is compared irrespectively of technology, management, competitive conditions, business cycle and specificities of industries. One particular element in such confusing results is the very definition of performance indicators - these indicators are often very questionable and may even differ significantly from the goals that form the support basis in which the arguments for SMEs are set. For example, the performance of firms maybe compared with respect to returns on capital and other financial ratios (which are not part of the arguments for SMEs), rather than with respect to employment, democratic management, market flexibility and so on (which are part of the arguments for SMEs). Another particular element in such confusion is that performance of the firms is often compared without consideration to the fact that in order to exist, not to speak of being efficient, SMEs may actually depend strongly on large firms - for example, through the supplier network, subcontracting, access to finance or trade related supplier/customer credit, etc. Thus, the relative efficiency of SMEs may depend on a network of industrial linkages with other and larger firms.
Some studies define a group of firms as small and medium because of their initial conditions - for example, the number of workers employed when the firm started - and then continue to analyse the performance of the firm through time not considering that the initial conditions may have changed and firms that are successful may be those that have grown to become large, or have associated with others and thus are no longer small.
Finally, most studies compare SMEs and large firms on a one-to-one basis - for example, how much pollution a SME creates compared with a larger firm; how much demand pressure on raw materials, fuels, spares, equipment, etc; how much competition for labour and upward pressure on real wages and skills, etc. These comparisons are highly inadequate as to achieve the same economic results (for example, level of exports and production, level of employment, etc.), a lot more SMEs are needed than large firms. The question, then, is what happens when all of those SMEs are put together? For a given level of output achieved through one large firm, what is the relative impact of all the necessary SMEs on pollution, demand for raw materials, spares, fuels and equipment, labour and skills, etc?
Related to these issues, when the number of individual SMEs becomes large, existent technologies, skills and industrial structures may become obsolete. This may lead, simultaneously, to concentration of capital and productive capacities (or an end to the smallness of firms and reduction of the number of firms), as well as to a fall on marginal employment as demand for new technologies and reduction of the number of firm change competitive conditions and the social and economic dynamics of capital accumulation.
Thus, the existing analytical framework for the evaluation of the relative advantage (or disadvantage) of different types of firms defined by scale seems to be inadequate, because the debate, as such, is not focused on the fundamental issues, the dynamics of firms and industries in nowadays capitalist economies (developed or underdeveloped). A closer examination of the issues may actually show that SMEs are not what they seem to be, SME development goals may be conflicting and confusing, and that empirical studies may be unsound.
More fundamentally, though, whether firms are large or small, their performance cannot be adequately assessed independently of social, economic and technological pressures, and industrial dynamics and structures of which firms form an active part - and which influence and are influenced by firms through networks, partnerships, and internal (to the firm) industrial and global vertical integration.
See, for example, Chang 1996 and Khan and Jomo (eds) 2000 about differences in institutional setting, social conditions and rent-seeking in Asia.
See, for example, Amsden 1989, Chang 1996, Wade 1990.
Using mainstream economic terminology, one would define "вЂ¦operating as a small firmвЂ¦" a firm that takes as given the competitive conditions in the industry. Firms that internalize key complementary
industrial operations, vertically integrate, associate and/or develop different sorts of networks and partnerships, are firms that change the competitive conditions and define them (or try to). Irrespectively of any subjective definition of scale, these firms do not operate as small firms, and thus cannot be said to be small firms. A practical example may help to clarify this point: would anybody argue that a bottling plant of Coca-Cola, anywhere in the world, is a small firm (even if the size of the plant is small according to the number of workers employed or any other arbitrary definition)?
See, for example, Fine and Murfin 1984, Kozul-Wright and Rowthorn (eds) 1998.