Conclusion
While there are numerous stylized facts on firm entry, exit, survival and growth (Geroski, 1995), the literature has not resolved yet the long debate on the relationship between business cycles and entrepreneurship. Most studies support the opinion that such a relationship is complex, and the subject still warrants further research (Acs and Storey, 2004; Audretsch and Pena-Legazkue, ˜ 2011; Fritsch, 2008). Our findings suggest that economic context matters for the process, rather than the action, of entrepreneurship. Specifically, a recession-driven shakeout generates a decline in entrepreneurship, and our mediation tests reflect that lower entrepreneurial activity is to some extent explained by a reduced opportunity perception of individuals during recessionary periods. Rather than a full direct effect of the economic shakeout on entrepreneurial activity, consistent with cognitive and planned behaviour theories, we empirically demonstrated that this relationship is partially mediated through a lower and more pessimistic opportunity perception, which leads to a drop in entrepreneurial action. The paper is not without limitations. Only one proxy has been used to describe the recession shakeout effect (e.g., changes in unemployment rate of regions as in the study by Fritsch et al., 2015), and the study is limited to the case of Spain during a precise economic recessionary period (2007---2010). Although the results can be seen as solid and valid enough as several additional tests for robustness were conducted (e.g., findings were similar when other proxies of the current crisis were considered, such as changes in GDP per capita), we suggest that further studies should be undertaken in other geographical contexts to authenticate our entrepreneurial process view.
