Being Self-Employee in Urban and Rural Areas: An Empirical Analysis

Self-employment trends vary significantly between European regions and across the observed years. The empirical results of an econometric exercise demonstrate that, in 2015, the coefficient related to rurality in Eastern European countries such as Albania, Bulgaria, and Montenegro, was found significant, suggesting that peripheral areas offered more opportunities for self-employment. In contrast, in Western European countries such as Belgium and the Netherlands, the coefficient was less relevant, indicating a smaller difference between various degrees of urbanization in terms of self-employment. The transition to remote work had a significant effect on self-employment, especially in the early years of observation (2010 and 2015). In countries with advanced digital infrastructures, such as Germany and Denmark, remote work facilitated self-employment, offering greater flexibility and job opportunities even in rural contexts. In Northern European countries such as Sweden and Denmark, public support proved positive in 2021, contributing to the resilience of self-employment during the pandemic. In contrast, in Mediterranean countries such as Italy, Portugal, and Spain, support was less significant, reflecting more restrictive public budgets and a lower capacity for intervention. The use of instrumental variables confirmed the significance of some of the relationships observed with standard regressions, while highlighting important differences. The results obtained with Instrumental Variables (IV) regressions indicate that work-related health problems are indeed correlated with residence in rural areas, especially in less developed countries. However, as mentioned, in smaller countries, or those with an efficient transport system, this correlation was less relevant. At an aggregate level, IV models showed results consistent with the baseline OLS models, confirming more general trends. In Eastern and Northeastern European countries, rurality coefficients were generally significant and positive. In Western and Southern European countries, the differences in the econometric outcomes of OLS and IV models were more pronounced, suggesting that reverse causality might have a greater impact on OLS estimates in these contexts.