Enterprise and Small Business Principles
Numbers of firms
If there are difficulties about defining the small enterprise, there are also methodological issues in measuring the number of smaller enterprises. This section looks at these issues but also shows that SMEs constitute nearly all of the enterprises and make a large contribution to overall employment in a given economy.
One reason for the lack of precision in terms of the actual number of enterprises in an economy is that there is no census of economic activity. Even if there were, there would still be the issue of informal economic activity. For instance, levels of illegal economic activity such as drug dealing are difficult to estimate (Fairlie, 2002). Equally, more legal forms of part-time or ad hoc activity such as homeworking or network marketing (individuals who are self-employed distributors selling to their family and friends (Pratt, 2000)) are also very difficult to pin down. Individuals who may be thought to be employees often find themselves ‘falsely’ self-employed because it allows for them or their ‘employer’ to reduce their tax liabilities (OECD, 2000).
In the absence, then, of a census of legal economic activities, much of what is known about SMEs is largely derived from two sources. First, governments throughout the developed world rely upon Labour Force Surveys (LFS) to estimate the number of selfemployed individuals and unincorporated businesses. Second, governments often make use of particular registers such as value added tax (VAT) or credit referencing companies (e. g. Experian, Dun and Bradstreet) to estimate the number of incorporated enterprises. There are a number of problems with both these data sources.
In terms of LFS data, individuals may choose, for whatever reason, to misrepresent their economic status. It is also likely that the LFS underestimates younger people because they are likely to be more geographically mobile than older individuals. The biggest problem, though, with LFS data is that it is a survey. As such, it is an incomplete appraisal of unincorporated business activity: for example, the enterprise population for the UK was revised upwards in 2003 because it was found that the UK LFS had underestimated the number of enterprises by more than 70,000 enterprises.
Registers of enterprises also have well known biases. Credit referencing data is likely to offer better coverage of larger enterprises, companies, enterprises in particular sectors (manufacturing and construction), and enterprises that seek external finance (Storey and Johnson, 1986, 1987). VAT registration data is just as problematic. Governments seeking to reduce regulatory burdens on smaller businesses may elect to substantially increase the VAT threshold. This occurred in the UK in 1991 when the government increased the VAT threshold from an annual turnover level of £25,400 to £35,000. Alternatively, if they are seeking to raise taxes they may decrease the threshold level. In either case, this has an impact on the robustness of VAT time-series data.
There are other problems with VAT data. Certain enterprises such as children’s clothes manufacturers are excluded from VAT registration because they are not liable to VAT. VAT registration is also not synonymous with firm ‘birth’. An enterprise may take a number of years to reach the point where their sales justify being registered for VAT. Similarly, VAT deregistration does not necessarily mean that the enterprise has ‘failed’. Enterprises may simply deregister because they have fallen below the turnover threshold.
The best estimate, therefore, of the enterprise population is often derived from a mix of survey information and registration information. In the UK, the enterprise population, as Table 2.3 shows, numbered just over 4 million in 2003. The vast bulk of these, some
2.9 million (70.1%) of enterprises, were made of enterprises that had no employees (the figures for employees (337,000 or 1.4%) represent partners of the business).
In terms of those enterprises with employees, the table shows that there were 1.23 million such enterprises. Again, the vast bulk of these were small enterprises either employing 1-4 employees (0.8 million) or 5-9 employees (0.22 million). Cumulatively, micro enterprises (0-9 employees) represent 95% of all enterprises, one-quarter of total employment and 22.3% of turnover. Table 2.3 also shows that the numbers of enterprises decline with employment size: 170,000 enterprises have between 10 and 49 employees whilst there are just 29,000 enterprises with between 50 and 249 employees. Nonetheless, the overall picture is that SMEs (< 250 employees) represent 99.8% of enterprises. Based upon this it is tempting to suggest that SMEs represent business activity. This neglects the fact that although large firms represent only 0.2% of enterprises, they contributed 52.4% of employment and 49.8% of sales in 2003.