Initiation of marijuana use, human capital, and earnings among young adults


French, Michael T. (2016). Initiation of marijuana use, human capital, and earnings among young adults. 6th Biennial conference of the American Society of Health Economists. Philadelphia, PA.


The use, sale, and possession of marijuana (cannabis) in the United States is illegal under current federal law. At the present time (as of June 2015), four states (Colorado, Washington, Oregon, and Alaska) and the District of Columbia have fully legalized marijuana for both medical and recreational use. Eighteen states have legalized marijuana for medical use only and a number of states have also decriminalized the possession of small amounts of marijuana. For example, Delaware recently passed legislation that decriminalizes the private use of up to an ounce of marijuana, replacing penalties with a civil fine. With legalization of marijuana at the forefront of local, state, and federal policy discussions, there is great scientific and policy interest in determining whether early initiation of marijuana use, particularly during an individual’s formative teenage years, can have long-lasting effects on human capital accumulation and lifetime earnings. While states have been contemplating and debating the legalization of marijuana for recreational and/or medical use, parallel efforts by government officials are underway to educate the public on the potential adverse effects of irresponsible use of drugs (including marijuana and tobacco). These campaigns have apparently been successful because tobacco and other drug use have been on a steady decline in the U.S., even in states where recreational use of marijuana is currently legal. In the present study, we use data from a unique longitudinal survey to assess the human capital and labor market consequences of early initiation of marijuana use. The data are obtained from several waves of Add Health, a school-based, longitudinal study of adolescent health-related behaviors and their consequences in young adulthood. The Add Health data have many desirable features pertinent to our study, the most notable being an extensive list of background characteristics and official records of high school grades. High school transcripts were requested for the full sample and abstracted for approximately 80 percent of Wave 3 respondents. Detailed educational attainment and labor market information was obtained from survey respondents in Wave 4. Extensive information about marijuana use in grade school, high school, and later was collected as well as the use of other drugs. Using various econometric models, we estimate the effects of early initiation of marijuana use (i.e., grade school or high school) on four main outcomes—educational attainment, high school grade point average (GPA), employment, and earnings as a young adult. Utilizing confidential data on key genetic markers, we control for possible endogeneity and/or selection bias of marijuana use. Results show that early initiation of marijuana use is negatively related to all four of our outcomes, but the estimated effects are diminished considerably in augmented models and when controlling for possible endogeneity of marijuana use. These findings have important implications for high school administrators, employers, and government policymakers.


Reference Type

Conference proceeding

Book Title

6th Biennial conference of the American Society of Health Economists


French, Michael T.

Year Published


City of Publication

Philadelphia, PA

Reference ID