Job satisfaction and flexicurity over the business cycle : evidence from Swiss individual-level data

The effects on employment of the recent economic crisis have become evident and persistent in many OECD countries, exacerbating on the one hand the demand for more flexibility by the firms; on the other the need to ensure workers security. 'Flexicurity', an institutional frame implementing a progressive flexibility of the labor market and at the same time guaranteeing its stability, has been defined as a successful model at the EU level before the crisis. The objective of this paper is therefore to empirically assess the effect of a micro-level measure of flexicurity on workers' job satisfaction using individual-level data from the longitudinal Swiss Household Panel (SHP), related to temporary and permanent employees in Switzerland for the period 2005 to 2011. Switzerland provides a particularly appropriate market to examine the potential effects of flexicurity type arrangements due to the relatively high incidence of flexible employment contracts. To this end, the sample of workers is disaggregated into different groups according to their employment contract and their perceived job security; we therefore analyse whether workers who are heterogeneous in terms of both the type of labour contract and their perceived security do also differ with regard to life satisfaction and specific aspects of the job satisfaction. Usually satisfaction variables, given the ordinal nature of the dependent variable, are analysed using ordered probit models. Van Praag and Ferrer-i-Carbonell (2006), however, developed a procedure, called Probit (OLS) or POLS, that consists in transforming an ordinal dependent variable in a "pseudo" continuous one, and allows the application of a linear model. With longitudinal data the POLS method permits for the inclusion of individual level fixed effects. Our methodology at first estimates a linear fixed effects model, thus controlling for unobserved time-invariant characteristics, while in a second step we implement a two-stage model to control for endogeneity. We also extend our analysis and seek to identify how insecurity affects temporary workers compared to permanent workers examining the impact of regional labour market conditions: we examine the effects on relative satisfaction and perceptions of security on workers in the seven Swiss statistical regions. First results show that job stability and perceived security are not necessary associated, and that job satisfaction is relatively low, mainly when perceived job security is low. The relationship between wellbeing and job security also varies according to regional employment conditions. This indicates that the duration of the contract may be less important if the worker perceives that he is not at risk of unemployment; in this regard, from a policy point of view, a greater 'flexicurity' can be obtained either directly from employer, adopting changes in work organization, or indirectly by policy makers through an appropriate mix of active labour market policies.