Skills and social insurance: Evidence from the relative persistence of innovation during the financial crisis in Europe Andrea Filippetti 1, * and Frederick Guy 2 1 London School of Economics and Political Science, Department of Geography and Environment, National Research Council of Italy, Institute of Regional Studies (ISSIRFA), Houghton St, London WC2A 2AE, United Kingdom and 2 Center for Innovation Management Research, Birkbeck University of London, Malet St, London WC1E 7HX, United Kingdom *Corresponding author. Email: a.filippetti@lse.ac.uk Abstract We study private sector investments in innovation in the early days of the financial crisis (between mid 2008 and mid 2009), using a survey covering more than 5,000 firms across 21 European coun- tries. Our interest is in how the stock of skilled labour affects the persistence of investment in innov- ation during a macroeconomic downturn. We infer differences in skill from national levels of par- ticipation in vocational education and training (VET) programmes interacted with levels of employment protection (EP) and unemployment insurance (RR). These forms of insurance should lead VET students to undertake training for skills which are more risky as human capital invest- ments, but potentially more productive. We show that the strongest sustained investment in innov- ation is associated with a combination of high VET with either strong EP or strong RR. The result supports the view that the supply of skills makes an important contribution to innovation, and that social insurance can encourage socially beneficial risk-taking in educational choices. Key words: varieties of capitalism; labour market institutions; skills; innovation investment; financial crisis; EU labour market. 1. Introduction Social insurance may affect a country’s economic performance, and may do so through a number of different routes. In this paper we consider how income insurance may affect investment in skills, and how the combination of income insurance mechanisms and the skills of the workforce may affect the persistence of firms’ investments in innovation during an economic downturn. More generally, this con- tributes to our understanding of the institutional roots of national economic performance. Our understanding of the linkage between social insurance and investment in skills starts with Estevez-Abe et al. (2001). In their framework, two very different forms of social insurance (more spe- cifically, income insurance) appear essentially interchangeable: these are employment security (EP), and unemployment security (which we will designate by our empirical proxy for unemployment secur- ity, the short-term income replacement rate (RR)). At the national level, both EP and RR tend to be associated with high levels of par- ticipation in vocational education and training (VET). The explan- ation given by Estevez-Abe et al. (2001) for this association is that that VET produces skills which are more specific to particular indus- tries or firms (and, we would add, technologies) than those obtained in academic education, which produces skills that are more general, or transferable. In the face of uncertain demand, more limited trans- ferability of skills implies greater risk, and that income insurance makes people more willing to invest in such risky skills. We extend the argument used by Estevez-Abe et al. (2001) by not- ing that the choices made by VET students will be different when in- come insurance is strong: students can use VET to obtain skills for use in stable markets (e.g. hairdressing) or in markets for which future de- mand is far less certain (tradeable goods industries where jobs may be abruptly sent offshore, new technologies with uncertain futures). By encouraging riskier choices, insurance not only increases participation in VET but also changes the composition of skills produced by VET. Skilled labour plays a role in innovation by contributing to a firm’s ability to adopt new technologies that others have developed, to make incremental process improvements, to operate production systems which are flexible in the sense of being able to both vary and to make incremental improvements to the product and, in general, to arrive at novel solutions to problems (Borra ´s and Edquist 2015). Innovation behaviour varies across the business cycle: different national systems of innovation within Europe are associated with different responses to innovation following the onset of the financial V C The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com 1 Science and Public Policy, 2015, 1–13 doi: 10.1093/scipol/scv036 Article Science and Public Policy Advance Access published July 17, 2015