SME route to development

Zaid Bakht

Photo: Shafiq Islam/ Drik News

Like elsewhere, small and medium enterprise (SME) is now the buzzword in the development parlance of Bangladesh. In the Poverty Reduction Strategy Paper (PRSP)-II, which is currently in force, development of SME has been identified as a critical element of the Strategic Block for pro-poor growth. SME is likely to constitute an important component of inclusive growth strategy underlying the upcoming Sixth Five-Year Plan (2011-2015).

Bangladesh Bank has attached high importance to channelling funds to this sub-sector for accelerating pro-poor growth.

The case for smaller enterprises is usually based on the conventional wisdom that enterprises with smaller scale of operation have certain desirable attributes. The smaller enterprises are expected to employ more labour relative to capital than their large-scale counterparts. They are likely to be more efficient as they face factor prices closer to their social opportunity costs and consequently use fewer resources per unit of output.

Being more labour-intensive and because of the likelihood that owners are less wealthy than the owners of large firms, the smaller enterprises will be more equitable in distributing the income that they generate. They are also more likely to operate in towns and villages, as well as in less developed regions, which would contribute to a wider spread of the benefits of development and partly alleviate the pressure of rapid urbanisation.

Given the relatively low skill and technology content and the geographically dispersed nature of smaller enterprises, the incidence of female employment is expected to be higher in them. This attribute would also render these enterprises more supportive of poverty reducing development strategies.

But which size group and activities actually enjoy the above-mentioned attributes remains the critical policy question. Empirical evidence on the performance of smaller enterprises around the world has been a mixed one. Depending on the stages of development, the relative role of the smaller enterprises within the overall economy has been found to vary widely across economies.

In the case of the developing economies the performance of the smaller enterprises with respect to the above-mentioned attributes has been found to be different in different industry categories. In some activities, the smaller enterprises performed better than their large-scale counterparts but in others the reverse was the situation.

Amongst the better performing smaller enterprises, the optimal size of the enterprise has also been found to vary across activities. While some micro and most small enterprises uphold these attributes in general, in many cases the size limit has to be extended to medium-sized units to include better performing enterprises.

Historically, the choice of the size group and activities has been decided in Bangladesh in an arbitrary manner. In all earlier Five Year Plans the emphasis was limited to small and cottage industries (SCI). Thus, for example, the Fifth Five Year Plan (1997-2005) mentioned SCI as an area where large scale employment opportunities exist and indicated the following inherent advantages of SCI over large and medium industries: lower capital investment, more jobs per unit of invested capital, lower capital-output ratio, lower infrastructure requirement, shorter start-up time, creation of micro entrepreneurial talent, lower consumption of energy, and less environmental risk. The cut-off size limit for small industry was, however, decided arbitrarily by different development agencies in different ways.

PRS P-I, initiated in 2005, replaced the expression small and cottage industry (SCI) by small and medium enterprises (SME), but did not provide any explanation for extending the coverage to include non-manufacturing activities and medium-sized enterprises and for taking cottage industries out of the purview.

PRSP-I argued the case for SME development in the same way as before, invoking the grounds of low capital intensity, spatial dispersal, opportunities for female employment etc. The cut-off size limit for small and medium enterprises has again been decided somewhat arbitrarily without any empirical verification to confirm that these size groups actually satisfy the desirable attributes of smaller enterprises.

One main difficulty in examining the performance of different size groups of enterprises is that the information base on economic performance of enterprises according to size is extremely weak. Bangladesh Bureau of Statistics (BBS) is entrusted with the responsibility of collecting this type of information. BBS collects, on an annual basis, detailed production data from a sample of manufacturing enterprises that have 10 or more workers under its Survey of Manufacturing Industries (SMI).

However, SMI data are not available by employment size group and the fixed assets data collected under SMI are problematic as they do not reflect current replacement costs. In the case of enterprises with less than 10 workers BBS conducts surveys very infrequently and the data also suffers from the same type of problem as SMI. As a result, BBS data cannot be used to assess the output size and performance of any particular size group of manufacturing enterprises.

The situation with respect to non-manufacturing enterprises is even worse. Other than complete census of these enterprises at long time intervals, BBS does not collect flow data on these non-manufacturing enterprises on a regular basis. Available BBS data, thus, does not permit assessment of the size of output or productivity in particular size group of manufacturing or non-manufacturing enterprises in Bangladesh. The estimates of the size of SME in Bangladesh and its contribution to GDP that appears in PRSP-II and other documents are, thus, guesstimates based on micro surveys.

Because of the lack of objective basis in determining the coverage of SME, policy measures aimed at promoting SMEs have often yielded distorted results. The extension of the purview of SME to include non-manufacturing activities has led to large number of trading enterprises of specified size to claim policy support earmarked for SMEs. Recent studies carried out at BIDS have shown that an overwhelming proportion of credit provided to SMEs under Bangladesh Bank's refinance scheme has gone to the trading sector.

Again, there is some on-going debate amongst the policy makers regarding whether micro enterprise (cottage industries and enterprises with less than 10 workers) should be included in the coverage, and it has been suggested that the expression should be changed to MSME to include micro enterprises along with small and medium enterprises.

In this context, the experience of other countries suggest that an across the board strategy of SME development is unlikely to yield desired result and instead the choice of the activity should be based on a positive, developmental attitude seeking to promote efficiency and growth. This would mean including activities that have high growth potential and leaving out those that are of residual nature with low value addition and low growth potential.

The pursuit of such a targeted strategy of SME development will require in-depth analysis of the performance of various size groups of enterprises in various manufacturing and non-manufacturing activities and determining their growth linkages. This in turn will require collecting detailed information on production activities of these enterprises across different size groups. The acute absence of such pertinent information has severely restricted the policy space for informed decision-making on SME development.

Zaid Bakht is Research Director, BIDS.