Health care.

The NBER'S Program on Health Care met in Cambridge on May 8. Kate Bundorf, NBER and Stanford University, and Mark Pauly, NBER and University of Pennsylvania, organized the meeting. These papers were discussed:

Mark Duggan, University of Maryland and NBER, and Fiona Scott Morton, Yale University

Helen Levy, University of Michigan and NBER, and David Weir, University of Michigan, " The Impact of Medicare Part D on Drug Utilization and Out-of-Pocket Spending: Evidence from the Health and Retirement Study"

Jay Bhattacharya, Stanford University and NBER, and Mikko Packalen, University of Waterloo, "The Other Ex-Ante Moral Hazard in Health"(NBER Working Paper No. 13863)

Liran Einav, Stanford University; Amy Finkelstein, MIT and NBER, and Mark R. Cullen, Yale University, "Using Price Variation to Estimate Welfare in Insurance Markets"

Guy David, University of Pennsylvania, and Tanguy Brachet, The Children's Hospital of Philadelphia, "Human Capital Accumulation and Forgetting in Emergency Medical Services"

Tom Chang, MIT, and Mireille Jacobson, University of California, Irvine, "Ownership Status and Response to Cost Shocks: Evidence from California's Seismic Retrofit Mandate"

The federal government began providing insurance coverage for Medicare recipients' prescription drug expenditures on January 1, 2006 through a program known as Medicare Part D. Rather than setting pharmaceutical prices itself, the government contracted with private insurance plans to provide this coverage. Enrollment in Part D was voluntary, with each Medicare recipient allowed to choose from one of the private insurers with a contract to offer coverage in her geographic region. Duggan and Scott Morton evaluate the effect of this program on the price and use of pharmaceutical treatments. Using data on product-specific prices and quantities sold in the United States, the researchers find that Part D substantially lowered the average price and increased the total use of prescription drugs by Medicare recipients. The results further suggest that the magnitude of these average effects varies across drugs, as economic theory predicts.

Levy and Weir use data from the 2004 and 2006 waves of the Health and Retirement Study to estimate the extent of adverse selection into Part D--that is, whether beneficiaries with high existing demand for prescription drugs disproportionately choose to enroll in the program--and the impact of Part D on medication use and out-of-pocket spending. They compare changes in use and spending for those who gained new Part D coverage to changes for those who were consistently covered by employer-sponsored insurance or Medicare HMOs, and for those who had no drug coverage in 2004 or 2006. Their results suggest that there was substantial selection into Part D: among Medicare beneficiaries with no drug coverage in 2004, those with high use and/or spending in 2004 were most likely to be enrolled in Part D in 2006. On average, the use of prescription drugs (number of prescriptions taken) did not change dramatically in response to Part D. Monthly out-of-pocket drug spending for previously uninsured, newly enrolled beneficiaries decreased, though. The median decrease was $30, compared to median baseline spending of $100 per month. In contrast, median out-of-pocket spending for those consistently covered by employer-sponsored insurance, Medicaid, Medicare HMOs, or privately purchased prescription drug insurance did not change between 2004 and 2006. These results are consistent with respondents' subjective perception that Part D did not change their use of drugs but did reduce their out-of-pocket spending. Somewhat surprisingly, Medicare Part D does not seem to have reduced the extent of cost-related non-compliance among those who previously had no drug coverage.

It is well known that public or pooled insurance coverage can induce a form of ex-ante moral hazard: people will make inefficiently low investments in self-protective activities. However, Bhattacharya and Packalen point out another ex-ante moral hazard that arises through an induced innovation externality: this mechanism will cause people to devote an inefficiently high level of self-protection. As an example, they analyze the innovation induced by the obesity epidemic. Obesity is associated with an increase in the incidence of many diseases. The induced-innovation hypothesis says that an increase in the incidence of a disease will increase technological innovation specific to that disease. The empirical economics literature has produced substantial evidence in favor of this hypothesis. Bhattacharya and Packalen estimate the associations between obesity and disease incidence; they then show that if these associations are causal, and the pharmaceutical reward system is optimal, then the magnitude of the induced innovation externality of obesity roughly coincides with the Medicare-induced health insurance externality of obesity. The current Medicare subsidy for obesity thus appears to be approximately optimal. They also show that the pattern of diseases for obese and normal weight individuals are similar enough that the induced innovation externality of obesity on normal weight individuals is positive as well.

Einav and his co-authors show how standard consumer and producer theory can be applied to estimating welfare in insurance markets with selection. Their key observation is that the same variation in prices needed to trace out the demand curve in any applied welfare analysis also can be used to trace out how costs vary as market participants endogenously respond to the price of insurance. With estimates of both the demand and cost curves, welfare analysis is straightforward. Moreover, because endogenous costs are the distinguishing feature of selection models, the analysis of the cost curve also provides a direct test for the existence of selection. The researchers discuss the data required to implement this approach and then apply it using individual-level data from a large private employer in the United States on the health insurance options, choices, and medical expenditures of its employees and their dependents. They detect adverse selection in this market and estimate that its efficiency cost, if these choices occurred in a free market setting, would be about 0.2 percent and 2 percent of the surplus that could be generated from efficient pricing. They estimate that the social cost of the subsidy needed to achieve the efficient outcome is about an order of magnitude higher than the social welfare gain from correcting the market failure.

David and Brachet examine the underlying channels through which organizational forgetting is achieved and their magnitude. They develop a framework for studying the relative contributions of labor turnover and skill decay to organizational forgetting. They then apply the framework to the performance of paramedics by using the universe of trauma-related ambulance runs in Mississippi betgween 1991 and 2005. Because paramedics are dispatched based on proximity and not on reputation, the relationship between experience and performance plausibly operates through learning-by-doing rather than through selection on unobservable quality. The authors conduct both individual and firm-level analyses to separate skill decay from labor turnover effects. They find that greater individual experience is robustly related to improved performance at the trauma scene. Their estimates suggest that about a third of human capital accumlated in a given quarter is eroded by skill decay. They find that turnover increases organizational forgetting by about 25 percent

Chang and Jacobson use a unified theoretical framework to model three popular theories of not-for-profit hospital behavior: 1) "for-profits in disguise"; 2) social welfare maximizers; and 3) perquisite maximizers. The authors develop testable implications of a hospital's response to a fixed-cost shock under each of these theories. They then examine the effect of a recent unfunded mandate in California that requires hospitals to retrofit or rebuild in order to comply with modern seismic safety standards. Because the majority of hospitals in the state were built between 1940 and 1970, well before a sophisticated understanding of seismic safety, a hospital's compliance cost is plausibly exogenously predetermined by its underlying geologic risk. The authors present evidence that within counties seismic risk is uncorrelated with a host of hospital characteristics, including ownership type. They show that hospitals with higher seismic risk experience larger increases in the category of spending that should be affected by retrofitting. Further, hospitals facing higher compliance costs are more likely to shut down, irrespective of ownership type. In contrast, private not-for-profits alone increase their mix of profitable services, such as neonatal intensive care days, obstetrics discharges, and MRI minutes. Government hospitals respond by decreasing the provision of charity care. As expected, for-profit hospitals do not change their service mix in response to this shock. These results are most consistent with the theory of not-for-profit hospitals as perquisite maximizers; it allows the authors to reject two of the leading theories of not-for-profit hospital behavior--"for-profits in disguise" and "pure altruism."