Entrepreneurial dynamics of the self-employed and of firms: a comparison of determinants using Spanish data

Some of the large scale initiatives to address the overall problem of having accurate and comparable indicators of entrepreneurship include: the COMPENDIA project (van Stel 2005); the Global Entrepreneurship Monitor (GEM) conducted by Babson College and The London Business School since the year 2000, and the OECD Entrepreneurship Indicators Programme (EIP), begun in 2006. Importantly, in early 2007, EUROSTAT, the statistical agency of the European Commission, became a partner in this activity and the EIP is now a joint OECD-EUROSTAT Programme: (http://​www.​oecd.​org/​document/​58/​0,3343,en_​2649_​44392116_​44441658_​1_​1_​1_​1,00.​html).

The list of personal characteristics and social factors that supposedly affect the decision to exploit an opportunity and potentially end up being self-employed includes: the degree of optimism, self-efficacy, internal locus of control, tolerance for ambiguity, need for achievement, family background, socio-cultural environment, etc. (Shane and Venkataraman 2000).

Empirical research (Evans and Jovanovic 1989; Evans and Leighton 1989; Carrasco 1999; Dawson et al. 2009; Millán et al. 2009) has looked at individual occupational choices using individual survey data (the transition from occupations, and the transition from unemployed to employee or self-employed). The transitions from occupations, from employee to self-employed, and among them from self-employed without employees to employers, or vice versa, are justified by individuals learning through experience (Jovanovic 1982). This research tests for the different factors that, according to the theory, affect occupational choice, especially wealth constraints, but they do not explain the equilibrium number of self-employed in the economy.

If a firm is represented by a production technology (production plant) and a scale of capacity at which average production cost is minimized, then the number of firms would be determined by the ratio of market demand over the efficient scale of production. However, if the boundaries of the firm are determined by the scope of the authority and control of the owner-entrepreneur then an entrepreneur may control several production plants and several firms. In this paper, the scope of authority of the entrepreneur is represented by the average span of control (employees per employer) and differences in span of control are explained as the result of occupational choices. Our view of the firm in this paper is as a legal entity that can participate in transactions with other economic agents and can hold ownership rights over non-human assets; this view is consistent with what we observe in the official statistics.

Firms may also look for larger size as a way of achieving market power and monopoly profits.

Holmstrom (1999) makes an explicit distinction between his theory of the firm as a sub-economy and the theory of the firm from transaction costs (Williamson 1985) and property rights (Grossman and Hart 1980) theories of the firm. Both, transaction costs and property rights, highlight the relevance of asset ownership, the efficiency of transactions depending on uncertainty, and asset specificity; but they do not explain the reasons why ownership of non-human assets is placed in legal entities (firms) or in persons, such as the entrepreneurs or the workers.

The Economically Active Population Survey is a sampling-based investigation performed quarterly, considering the population living in family dwellings in the Spanish national territory. Its main goal is to reveal the characteristics of that population with regard to the labour market. It focuses on providing data on the main population categories related to the labour market (employed, unemployed, active population, inactive population) and obtaining classifications of these categories, depending on certain characteristics. The official published statistics include time series data for the country as a whole and for each AACC. Desegregated data by region and industry must be requested from the Statistical Office and is supplied only for levels of desegregation that assure statistical representation.

The Central Companies Directory consists of an electronically organised set of information, with data on identification, localisation, territorial distribution and classification by size and economic activity of the units, both Company (legal entity) and Production unit, obtained from administrative sources and statistical files, complemented with other information coming from the current statistical operations of the INE. The economic scope of the Central Companies Directory covers information relating to the entire national territory.

Appendix 1 contains a map of the political organization of Spain and of the location of the regions considered in the study. We also provide data on productivity, income per capita, and unemployment rate of each region, to have an idea of the heterogeneity of economic conditions among the regions, even though they belong to the same country.

The DIRCE’s definition for an enterprise is “an organization under the direction of an individual, a legal entity or a combination of both types of personalities, created for the production of one or several kinds of goods or services in one or several locations”. The enterprises as legal entities include self-employed individuals who have a licence to act in the market place even though they do not hire workers.

According to the EPA, the number of self-employed in Primary industries was 559,000 in 1995 and 376,700 in 2006.

The average span of control of 16 employees per employer (2006’s average) is relatively high for effective direction and supervision. However, note that entrepreneurs will hire professional managers to perform this function and the total number of occupied individuals includes employees working for the public sector who are under a different employment and management regime. When we take into account the number of people that the EPA includes in the group of managers (374,100), the average span of control goes down to 11.5 in 2006. Excluding public employees (2,882,200), the average decreases to 9.5 in 2006.

This high number is conditioned by the fact that this industry includes public employees. If they are excluded, the average span of control decreases to 24.4 in 2006.