Health Economics Program Meeting

April 20, 2012
Michael Grossman of City University of New York's Graduate Center and Theodore J. Joyce of Baruch College, Organizers

D. Mark Anderson, Montana State University; Daniel I. Rees, University of Colorado Denver; and Benjamin Hansen, University of Oregon

Medical Marijuana Laws, Traffic Fatalities, and Alcohol Consumption

To date, 16 states have passed medical marijuana laws, yet very little is known about their effects. Using state-level data, Anderson, Rees, and Hansen examine the relationship between medical marijuana laws and a variety of outcomes. Legalization of medical marijuana is associated with increased use of marijuana among adults, but not among minors. In addition, legalization is associated with nearly a 9 percent decrease in traffic fatalities, most likely as a result of its impact on alcohol consumption by young adults. These estimates provide strong evidence that marijuana and alcohol are substitutes.


Tinna Laufey Asgeirsdottir and Porhildur Olafsdottir, University of Iceland; Hope Corman and Kelly Noonan, Rider University and NBER; and Nancy E. Reichman, University of Medicine and Dentistry of New Jersey

Are Recessions Good for Your Health Behaviors? Impacts of the Financial Crisis in Iceland

Asgeirsdottir, Olafsdottir, Corman, Noonan, and Reichman exploit the October 2008 financial crisis in Iceland-a severe and unexpected macroeconomic shock that can be pinpointed to a single day-to identify the effects of a macroeconomic downturn on a range of individual health behaviors. They use data collected in 2007 (during the boom) and 2009 (during the bust) from "Heilsa og líðan," a longitudinal health and lifestyle survey carried out by the Public Health Institute of Iceland that includes pre- and post- reports of the same health behaviors as well as other relevant variables. They investigate the effects of the crisis on a range of health-compromising behaviors (smoking, heavy drinking, "junk food" consumption, and indoor tanning) and health-promoting behaviors (dental visits; consumption of fruits, vegetables, and other health-promoting foods; use of dietary supplements; and getting the recommended amount of sleep). Because they observe information on health behaviors as well as key hypothesized mechanisms (work hours, real income, and emotional distress) on the same individuals over time, they are able to investigate mechanisms underlying changes in health behaviors. Thus they are able to disentangle-at least to some extent-the general effects of the economic downturn overall from the effects of individuals' changes in circumstances that resulted from the shock. They find that the crisis reduced all of the health compromising behaviors examined and most of the health promoting behaviors. The exceptions are that the crisis increased consumption of fish oil and getting recommended amounts of sleep. To some extent, the effects operated through changes in real household income: the authors infer that a substantial portion of the effects operated through the large prices changes that resulted from the crisis.


John Cawley, Cornell University and NBER; David Frisvold, Emory University; and Chad Meyerhoefer, Lehigh University

The Impact of Physical Education on Obesity among Elementary School Children

In response to the dramatic rise in childhood obesity over the past several decades, the Centers for Disease Control and Prevention (CDC) and the American Academy of Pediatrics (AAP), among other organizations, have advocated an increase in the time that elementary school children spend in physical education (PE) classes. However, little is known about the effect of PE on child weight. Cawley, Frisvold, and Meyerhoefer measure those effects by instrumenting for child PE time with state policies, using data from the Early Childhood Longitudinal Study, Kindergarten Cohort (ECLS-K) for 1998-2004. Their results indicate that PE lowers BMI z-score and reduces the probability of obesity among fifth graders (in particular, boys), but the instrument used here is not powerful enough to reliably estimate the effects for younger children. This study represents some of the first evidence of a causal effect of PE on youth obesity, and thus offers at least some support for the assumptions behind the CDC and AAP recommendations. The authors find no evidence that increased PE time crowds out time in academic courses or has spillovers to achievement test scores.

Charles J. Courtemanche, University of Louisville and NBER; Garth Heutel, University of North Carolina, Greensboro and NBER; and Patrick McAlvanah, Federal Trade Commission

Impatience, Incentives, and Obesity (NBER Working Paper No. 17483)

Courtemanche, Heutel, and McAlvanah explore the relationship between time preferences, economic incentives, and body mass index (BMI). They first provide robust evidence that greater impatience increases BMI and the likelihood of obesity. Next, they show that cheaper food leads to the largest weight gains among those exhibiting the most impatience, potentially helping to explain why increases in BMI have been concentrated amongst the right tail of the distribution. Finally, they show that both the (present-bias) and (long-run patience) parameters of a quasi-hyperbolic discounting specification predict BMI, suggesting that obesity is partly attributable to both rational intertemporal tradeoffs and time inconsistency.


John Cawley, Cornell University and NBER; David Frisvold, Emory University; and Chad Meyerhoefer, Lehigh University

The Impact of Physical Education on Obesity among Elementary School Children

In response to the dramatic rise in childhood obesity over the past several decades, the Centers for Disease Control and Prevention (CDC) and the American Academy of Pediatrics (AAP), among other organizations, have advocated an increase in the time that elementary school children spend in physical education (PE) classes. However, little is known about the effect of PE on child weight. Cawley, Frisvold, and Meyerhoefer measure those effects by instrumenting for child PE time with state policies, using data from the Early Childhood Longitudinal Study, Kindergarten Cohort (ECLS-K) for 1998-2004. Their results indicate that PE lowers BMI z-score and reduces the probability of obesity among fifth graders (in particular, boys), but the instrument used here is not powerful enough to reliably estimate the effects for younger children. This study represents some of the first evidence of a causal effect of PE on youth obesity, and thus offers at least some support for the assumptions behind the CDC and AAP recommendations. The authors find no evidence that increased PE time crowds out time in academic courses or has spillovers to achievement test scores.


Daniel Millimet, Southern Methodist University, and Rusty Tchernis, Georgia State University and NBER

Estimation of Treatment Effects without an Exclusion Restriction: with an Application to the Analysis of the School Breakfast Program (NBER Working Paper No. 15539)

The increase in childhood obesity has garnered the attention of many in policy making circles. Consequently, school nutrition programs such as the School Breakfast Program (SBP) have come under scrutiny. The identifi…cation of the causal effects of such programs, however, is difficult because of non-random selection into the program and the lack of exclusion restrictions. Millimet and Tchernis propose two new estimators aimed at addressing this situation. They compare the new estimators to existing approaches using simulated data. Thsyhow that while correlations might suggest that SBP causes childhood obesity, SBP is likely to reduce childhood obesity once selection is addressed.


Geoffrey Joyce and Dana Goldman, University of Southern California and NBER, and Julie Zissimopoulos, University of Southern California

Digesting the Doughnut Hole

The standard Medicare Part D benefit has a unique non-linear design: coinsurance rates are a function of total drug expenditures. Beneficiaries face a deductible, followed by a coinsurance rate of 25 percent, until their total drug expenditures reach the initial coverage limit, which was $2,930 in 2012. Beyond that threshold, they pay 100 percent of their costs (the so-called "doughnut hole") until their annual out-of-pocket expenditures reach the catastrophic limit, which was $4,700 in 2012. Above that threshold, beneficiaries pay just 5 percent of their drug costs for the remainder of the year. Using Medicare claims data (Parts A, B, D) for 2006 through 2008, Joyce, Zissimopoulos, and Goldman study Medicare beneficiaries ages 65 and older with diabetes, comparing changes in prescription drug use, health care utilization, and mortality over the coverage phases for two types of beneficiaries: those facing differential prices over the course of the spending distribution (non-LIS) and enrollees receiving full low-income subsidies (LIS) who have access to medications at little or no cost throughout the year. The authors find that, relative to LIS beneficiaries, those facing the coverage gap adjust their spending behavior when prices rise and over time. Among those reaching the doughnut hole in 2007, 24 percent of the non-LIS group did not reach the coverage gap in 2008, compared to just 11 percent of LIS beneficiaries, who reduce their use of medications by 6 to 7 percent after reaching the coverage gap. Difference-in-difference results reveal that a higher percentage of non-LIS beneficiaries stop therapy altogether and are more likely to switch to generics or lower cost brands after reaching the coverage gap. Despite these changes in prescription drug use, though, there is no evidence that the coverage gap affected health, as measured by changes in medical care spending and use of inpatient and outpatient services.