Essays on the Labor Supply Dynamics of Older Workers

The European Union laid out the Lisbon agenda in order to become "the most competitive and dynamic knowledge-based economy in the world, capable of sustainable economic growth with more and better jobs and greater social cohesion". The fulfillment of this ambition should not discard the introduction of policies enhancing the economic participation of the elderly. In fact, lower fertility rates and increasing life expectancy have triggered the aging process of European population and are deeply altering its demographic structure. An aging society may exert larger pressure on the financial sustainability of pay-as-you-go pension systems owing to the decreasing age-profile of employment rates. On the one hand, the revenues coming from workers contributions to the scheme are expected to fall because a smaller fraction of individuals will be at work. On the other hand, the overall pension burden is expected to increase because a rising proportion of the population will be eligible for pension benefits. In view of these considerations, European Union fosters Social Security reforms which assure long-run financial equilibrium and fully take into account the ongoing demographic modifications. At the same time, such institutional changes should be accompanied by public programs aimed at extending the working life of the elderly. Indeed, stimulating the labor market participation of older workers is an ulterior strategy to alleviate the financial burden of Social Security because many of them who no longer find profitable the permanence in the labor market are likely to apply for retirement benefits. The development of policies aimed at achieving these purposes requires the support of analyses examining the labor supply dynamics of older workers. Throughout the thesis we will focus on transitions out of employment by considering elderly individuals at work in a given time period and describing their probability of becoming not employed in the future. In line with Blau (1994), we opt to estimate the dynamics of labor market position rather than imposing some arbitrary definition of retirement because the exit from the labor force may occur following alternative paths, such as becoming labor pensioners at the end of a full time job or starting an unemployment spell due to firm-downsizings that turn out to be a bridge towards ad-hoc early retirement arrangements. In order to preserve such heterogeneity, in this study workers are allowed to leave employment for whatever reason and to perform any possible employment trajectory. Chapter 1 studies the impact of Information and Communication Technology (ICT) skills on the probability of ceasing from work. The point at issue is of particular interest because the Lisbon agreements place the improvement and the diffusion of ICT knowledge among the priorities of the Union. Our work draws data from the waves 2000-2004 of the Bank of Italy Survey on Household's Income and Wealth. Remarkably, the questionnaire of the 2000 survey provides unique information to measure ICT knowledge at the individual level. In fact, respondents are asked to report the utilization of a PC at work and to rank their computer literacy according to a predefined scale. This distinction is useful to disentangle the effects of actual individual investments in ICT training from those due to implicit job requirements reflecting firm organization. The econometric specifications adopted allow for the potential endogeneity of technological knowledge in a labor supply framework. In fact, older individuals who plan to work longer may be more willing to enhance their skills as they face a longer period over which training costs can be recouped. Improvements in ICT skills are estimated to produce a positive effect on employment chances only for males with high education. This pattern corroborates the hypothesis of complementarity between educational attainments and technological knowledge (Weinberg, 2004). The literature provides a huge amount of evidence suggesting a positive relationship between individuals health and their own employment chances. Instead, few studies consider the role of the health conditions of other household members in the determination of the opportunity cost of keeping on working. Chapter 2 explores this research area by investigating how the labor supply dynamics of married workers is influenced by their own health conditions and by those of their cohabiting partners. Exploiting the European Community Household Panel (1995-2001), the overall psychophysical well-being of individuals is described by means of alternative indicators based on either self-assessments or more objective health indexes. Partners employment conditions are included among the explanatory factors because of their close relationship with health and the well-documented coordination between the labor market positions of couple members (Michaud, 2003). We also control for different sources of heterogeneity that may arise in our set-up and prevent us to estimate consistently the causal effects of interest. Our results confirm that healthier individuals are more prone to keep on working and, on the contrary, suggest that those living with a spouse in poor health are characterized by a higher propensity towards leaving employment. Finally, Chapter 3 deals with the study of labor market transitions when the use of panel datasets involves a number of major complications, for instance due to severe attrition, small sample size, short time-span covered or lack of relevant variables. Following the approach proposed in Güell and Hu (2006), we estimate the individual hazard rate of stopping working only resorting to repeated independent cross-sections representative of the same population. Although workers are not observed over time, we are still able to evaluate their own likelihood of becoming not employed by combining the labor market outcomes experimented in different time periods by individuals of the same birth-cohort. Drawing data from the ISTAT survey Aspetti della Vita Quotidiana (1993-2003), we recover the evolution of the risk of leaving employment in a period characterized by several pension system reforms. We propose evidence that while the Amato (1995) and Prodi (1997) reforms passed, the cohorts of older workers at time show the highest hazard rate of leaving employment. On the contrary, their counterparts in 2001 and 2002 are more likely to prolong their working life. These findings suggest that, in the short run, repeated institutional changes designed to lower the Social Security generosity actually induce an increase in the exit rate from the labor force.