Editor's Note: Khrystyna Kushnir is a consultant on micro, small and medium-sized enterprises with the Enterprise Analysis Unit of the World Bank Group.
At the G-20 summit in Pittsburgh last year, the assembled authorities agreed  to "scale up successful models of small and medium-sized enterprise (SME) financing." The G-20 assigned the IFC and other international organizations to launch a G-20 Financial Inclusion Experts Group and asked the private sector to come up with ideas through G-20 SME Finance Challenge . This increased attention to micro, small and medium-sized enterprises (MSMEs) begs the question -- what, exactly, should be considered an MSME?
With the issue of MSMEs playing out on an international level, it is tempting to try to find a universal MSME definition. A universal MSME definition would ease the design of loans, investments, grants and statistical research. One such effort is IFC’s SME Definition Deep-dive Analysis and Recommendations , although it's currently on hold because of internal restructuring.
As part of the G-20 follow-up work, IFC is currently working on a 2010 update of the Micro, Small, and Medium Enterprises: A Collection of Published Data . While recording the various definitions of MSME used in 120+ of the most populous world economies, I was struck by the wide range of approaches governments take to define what exactly an 'MSME' is in their economy.
For example, in China an MSME can be an enterprise  with 1 to 3000 employees; total assets from ¥ 40 to 400 million and business revenues from ¥10 to 300 million depending on the industry. Meanwhile, the EU considers an MSME an enterprise  with up to 250 employees and turnover of no more than €50 million or a total balance sheet of no more than € 43 million.
Hypothetically, the choice of MSME definition could depend on many factors, such as business culture; the size of the country’s population; industry; and the level of international economic integration. Or it could be the result of businesses lobbying for a particular definition, which would qualify their enterprise for governmental MSME support program. This can result in some very strange distinctions between firms. For example, in one case an enterprise manufacturing petroleum products  is considered an MSME if it employs less than 68 persons; but if a similar enterprise manufactures chemical products, then the threshold for it to be considered an MSME is raised to a maximum of 100 employees.
These issues make it difficult to adopt a universal MSME definition and raise the question of whether it makes sense to strive for what, in the end, might simply be a Procrustean bed . It might make more sense to measure MSMEs with a single ruler, e.g. annual sales or turnover and/or number of employees, but tailor the size breakdown to particular conditions in the country of operation.
In their paper Defining SMEs: A Less Imperfect Way of Defining Small and Medium Enterprises in Developing Countries , Tom Gibson and H. J. van der Vaartindeed suggest a less imperfect formula:
An SME is a formal enterprise with annual turnover, in U.S. dollar terms, of between 10 and 1000 times the mean per capita gross national income, at purchasing power parity, of the country in which it operates.
The Gibson and Vaart definition is attractive because annual turnover is a good tool to assess the contribution of MSMEs to GDP. It is also fulfills the criterion of working as a single ruler. But after extensive empirical research on MSME statistics and definitions in over 120 of the world's most populous economies, I can say that it is extremely difficult to obtain data on the annual turnover by MSMEs in developing countries. Also, the informal MSMEs in some developing countries outnumber formal MSMEs by 8 times. So the problem of one-size-fits-all remains even in the Gibson and Vaart definition.
From my research I have found that the number of employees was the most frequent characteristic used in the definitions of national governments and statistical agencies. The data on size breakdown was available for most of countries. The number of employees is also a fair characteristic to assess MSMEs’ contribution to GDP.
To sum up, I think there is both a technical and a conceptual problem with seeking a universal definition of MSMEs. Technically, lack of data is a big problem. But even if that could be overcome, economies have diverse structural, cultural and political reasons to adopt different definitions of MSMEs that would run counter to any universally agreed definition.