Using self-employment as proxy for entrepreneurship: some empirical caveats

Self-employment is the most frequently used measure of entrepreneurship. However, its definition varies between countries, which makes comparisons difficult. We present an analysis of Swedish self-employment data and show that even within one country, the depicted development differs greatly depending on the source used. Unlike previous claims in cross-country studies, we find that there is no basis for categorising Sweden as having increased its self-employment rate more than others. This demonstrates a need to carefully specify the characteristics of the data, and their advantages and disadvantages, before drawing conclusions about the frequency of entrepreneurship in different countries.