The incidence of self-employment in Spain, excluding the agricultural sector, was close to 15% in 2017 (because of the specific nature of self-employment in agriculture, this study does not consider that sector). This share was above the European average (11.9%) and had remained barely unchanged during the previous 10 years.

​​​​However, this apparent stability may hide very different situations, as self-employment encompasses highly heterogeneous groups of workers. This research analysed two categories of self-employed workers: firstly, the economically dependent self-employed, encompassing workers who depended on one single client for work in the previous year. Secondly, those who declared that they were involuntarily self-employed, either out of necessity or because a former employer requested it. Finally, a closer look taken at the interaction of both groups.

Identifying these different types of self-employed workers is very important. Although most legal systems distinguish between wage employment and self-employment, there is a grey area between them. In recent years there has been a shift towards greater outsourcing, which has led to an increase in the number of “false” self-employed. The main reason is that it costs less for employers to hire self-employed workers than to take on employees. Thus, it seems relevant to identify these self-employed workers separately, because their characteristics and the effects of economic conditions and labour-market institutions are expected to be different than for voluntary and/or economically non-dependent self-employed workers. This could be relevant for the design of specific economic policies aimed at each group of workers.

Looking at the reasons for becoming self-employed, the transition into self-employment is a voluntary choice for most self-employed workers in Europe, as only 16.9% of the self-employed declare their status as involuntary. For Spain, the percentage is 21.7%: this is above the European average, although below the incidence observed in most Eastern European countries. Regarding economically dependent self-employment, only 3.6% of total self-employed workers in Spain claim that they rely on a single client, below the European average (8.4%) although at a similar level to that of countries such as France and Portugal.

Regarding regional differences within Spain, involuntary self-employed workers represent over 30% of total self-employment in Galicia and around 25% in other regions such as Asturias, Andalusia, and Cantabria. In contrast, the incidence is 15.7% in La Rioja and below 20% in other regions such as Aragon, the Balearic Islands, and Castilla-La Mancha. Similarly, the dispersion of dependent self-employment is high across regions and while 7% of total self-employed workers are dependent in Madrid, the percentage is particularly low in the Basque Country (1.1%).

Data from the 2017 European Union Labour Force Survey (EU-LFS) allows analysis of which individuals show a higher probability of becoming involuntary and dependent self-employed workers. The incidence of involuntary self-employment is particularly high for workers aged under 30 years, with a lower level of education and working in the construction sector. Larger differences are observed between occupational categories, as the incidence of involuntary self-employment reaches 51% among elementary occupations whereas in the highest-skilled occupations (managers) the incidence is just 12.2%.

Regarding dependent self-employment, males, immigrants, and younger workers are more likely to become dependent self-employees. As in the case of involuntary self-employment, the incidence is higher in the construction sector and in low-skilled occupations (drivers and machine operators, and elementary occupations). Women exhibit a somewhat higher incidence of involuntary self-employment, although differences by gender are more notable in the case of dependent self-employment, where men are twice as likely to be dependent workers. A similar pattern is observed with respect to the differences between immigrants and natives. There are hardly any differences in involuntary self-employment, but the share of dependent self-employment among immigrants almost doubles the incidence among nationals.

To analyse in greater detail the effect of GDP and other regional variables, such as the production structure and a set of institutional variables, different econometric models have been estimated. The main purpose of this research is to understand whether or not those variables have a different effect across the different types of self-employed workers once the demographic and employment characteristics of the workers have been taken into account.

Although simple data correlations show that self-employment is a more likely choice when the economic situation worsens, the econometric model shows that this is only the case for the economically non-dependent and the voluntarily self-employed. The rest of self-employed workers do not appear to be significantly affected by GDP.

Regarding the structure of the production system, the basic model suggests that a bigger manufacturing sector predicts a higher incidence of self-employment. But when we differentiate by types of self-employed workers, the results show that this effect is mostly driven by those most willing to enter into self-employment: the non-dependent and voluntarily self-employed.

Similarly, the results indicate that the share of regional employment in the public sector is positively associated with higher rates of self-employment. Nonetheless, the evidence shows that the larger the public sector, the smaller the incidence of dependent and involuntary self-employment (those who might be less willing to initiate an entrepreneurial activity).

Finally, two more variables are included to help understand the situation in the regional labour markets. Firstly, the incidence of minimum wages as an approximation to regional wage distribution. There is a positive relationship between low-wage regions and the share of dependent self-employees, especially if these are involuntary. Secondly, a similar result is found when the involuntary part-time rate is analysed; it yields a higher incidence of the most vulnerable group of self-employed workers (dependent and involuntary) in those regions with less favourable labour market conditions.

The results of this research are useful for gaining a better understanding of the characteristics and the behaviour of self-employed workers, a highly heterogeneous group which includes some workers faced with precarious labour market conditions. In particular, it seems that different labour market conditions and institutions have a different effect on the self-employed, depending on their reason for becoming self-employed and on their degree of economic dependence. Therefore, specific economic policies should be designed for each group of workers.

The 2017 EU-LFS ad-hoc module on self-employed workers was available for researchers, for the first time, in July 2019. There are no previous ad-hoc modules devoted to this topic. This makes this study especially relevant, since we are exploiting unique and detailed information about the self-employed. However, at the same time, a lack of longitudinal information prevents us from analysing the dynamics of involuntary and dependent self-employment (i.e. duration and transition issues).

CARRASCO, R. (1999): «Transitions to and from self‐employment in Spain: an empirical analysis», Oxford Bulletin of Economics and Statistics, 61(3), 315-341.

CENTENO, M. (2000): «Is self-employment a response to labour market rigidity?», Economic Bulletin, 37-44.

ROMÁN, C., CONGREGADO, E., and MILLÁN, J. M. (2011): «Dependent self-employment as a way to evade employment protection legislation», Small Business Economics, 37(3), 363-392.

THÖRNQUIST, A. (2015): «False self-employment and other precarious forms of employment in the ‘grey area’ of the labour market», International Journal of Comparative Labour Law and Industrial Relations, 31(4), 411-429.

TORRINI, R. (2005): «Cross-country differences in self-employment rates: The role of institutions», Labour Economics, 12(5), 661-683.