| Author/Presenter |
Donata Bessey (University of Zurich) |
| Co-author |
Uschi Backes-Gellner (University of Zurich) |
| Title |
Marijuana Consumption, Educational Outcomes and Labor Market Success: Evidence from Switzerland |
| Abstract |
The aim of this paper is an empirical analysis of the impact of marijuana consumption on educational outcomes and labor market success. Existing research in the field focuses on consumption of one drug and relatively short-term consequences for outcomes. The innovations of our research project are that we use a theoretical foundation for the analysis, take into account different kinds of risky behaviors and a longer-term perspective on outcomes. The health-education gradient (positive relationship between more education and better health) is a very robust finding in the empirical literature. Previous research tried to establish causal relationships from one component of the human capital stock (i.e., health or education) to the other component. In a recent paper, Heckman (2007) proposed a synthesis of the two distinct literatures on health and education economics and developed a lifetime model of investment in human capital. An interesting feature of the model for our research question is the possibility to identify critical and sensitive periods in development. We want to test this feature and analyze during which periods of development drug consumption is especially harmful for educational attainment and labor market success of young adults. For the empirical test we use a unique Swiss data set, the Swiss Health Survey. It contains very rich information on individuals' health-related attitudes, behaviors, health status, and possible problems, but also on their educational attainment, professional status and personal income and many more. First estimation results (using ordered probit models) suggest that there are indeed critical periods with respect to alcohol and marijuana initiation for educational outcomes. In order to tackle endogeneity issues, we are constructing a candidate instrumental variable for marijuana consumption based on local availability of the drug. In a second research step, we plan to estimate the impact of consumption on employment status and hourly wages. |
| Web Link |
http://www.isu.uzh.ch/cms/emap/team/assistants/donatabessey/download/don |
|