Health Care

March 6, 2015
Jonathan Gruber of MIT, Organizer

Emily Oster, Brown University and NBER

Diabetes and Diet: Behavior Change and the Value of Health

Individuals with obesity appear to be reluctant to undertake dietary changes. Evaluating the reasons for this reluctance, as well as appropriate policy responses, is hampered by a lack of data on behavioral response to dietary advice. In this paper, Oster uses household scanner data to estimate food purchase response to a diagnosis of diabetes, a common complication of obesity. She infers diabetes diagnosis within the scanner data from purchases of glucose testing products. Households engage in statistically significant but small calorie reductions following diagnosis. The changes are sufficient to lose 6 to 11 pounds per year, but are only 10% to 20% of what would be suggested by a doctor. In the short term (1 month) changes by food type line up with doctor advice, but in the longer term only decreases in unhealthy food persist. Oster evaluates these changes in the context of a simple model of optimization under full information and find that individuals value the marginal 100 calories per day at between 0.2 and 1.0 life years. Analysis of heterogeneity suggests limited demographic heterogeneity but does identify some successful dieters. Those with large caloric reductions typically focus on a small number of food items. Oster compares the results to a policy of taxes or subsidies. A 10% tax on unhealthy foods would produce smaller changes than what is observed after diagnosis, but a 10% subsidy on healthy foods would have a much larger impact.


Matthew R. Grennan, University of Pennsylvania, and Ashley Swanson, University of Pennsylvania and NBER

Transparency and Negotiated Prices: The Value of Benchmarking in Hospital-Supplier Bargaining

This paper empirically analyzes the role of information in bargaining between hospitals and their suppliers. Hospital supplies account for a large percentage of both the level and growth of health care expenditures, and prices for the same input can vary dramatically across hospitals. This variation has prompted calls for increased transparency as a mechanism to lower prices, but whether such an intervention would be successful depends on the details of what information is provided and how market participants respond. Grennan and Swanson analyze a new data set including all purchase orders issued by over ten percent of U.S. hospitals over 2009-13. The empirical setting contains an intervention in which sample hospitals gained access to benchmarking data on other hospitals' negotiated prices. Using differences-in-differences identification strategies based on timing of hospitals' access to price information and on new product entry, the researchers find that access to information on purchasing by peer hospitals led to large reductions in prices, concentrated among hospitals earning that they were performing relatively poorly in contracting and for products purchased in relatively large volumes, and that the usefulness of price information is constrained by delays in renegotiation between hospitals and suppliers.


Zack Cooper and Stuart V. Craig, Yale University; Martin Gaynor, Carnegie Mellon University and NBER; and John Van Reenen, London School of Economics and NBER

Why is Health Care Spending on the Privately Insured in Grand Junction, Colorado So High? Prices, Competition, and Health Care Spending

Grand Junction, Colorado is located 247 miles southwest of Denver. With an economy fueled by tourism, the city is home to about 150,000 residents. The city also has hugely high health care spending for the privately insured. While the Grand Junction hospital referral region has one of the lowest levels of spending per Medicare beneficiary, its spending per beneficiary on the privately insured ranks in the highest quintile of the nation. This paper explores why Grand Junction and cities and regions like it have such high hospital spending for the privately insured. In particular, Cooper, Craig, Gaynor, and Van Reenen focus on analyzing the role that providers' prices play in driving health spending and seek to analyze the factors that impact providers' price levels.
This paper analyzes the variation in health care spending of Americans covered by employer-sponsored insurance using a new, unique national dataset from the Health Care Cost Institute that is composed of all insurance claims paid by Aetna, UnitedHealth, and Humana from 2007 through 2011. Crucially, this data includes the prices that these insurers negotiated for care with all providers. In what follows, the researchers document the contribution prices make to the spending variation of the privately insured and assess the extent to which negotiated provider prices vary within and across markets. Finally, the authors identify the key factors that are driving the observed variation in prices.
Ultimately, the researchers find weak correlation between Medicare spending per beneficiary and spending per privately insured beneficiary. Moreover, they find that provider prices explain the majority of hospital spending variation across the U.S. The authors also document extensive variation in provider prices within and across markets and illustrate that provider concentration is the strongest predictor of hospital prices.

Benjamin R. Handel, University of California, Berkeley and NBER; Jonathan T. Kolstad, University of Pennsylvania and NBER; Amitabh Chandra, Harvard University and NBER; and Zarek C. Brot-Goldberg, University of California, Berkeley

What Does a Deductible Do? The Impact of Cost-Sharing on Health Care Prices, Quantities, and Spending Dynamics

Measuring consumer responsiveness to medical care prices is a central issue in health economics and a key ingredient in the optimal design and regulation of health insurance markets. Handel, Kolstad, Chandra, and Brot-Goldberg study consumer price responsiveness to medical care prices leveraging a natural experiment that occurred at a large self-insured firm that forced all of its employees to switch from an insurance plan that provided free health care to a non-linear high deductible insurance plan with markedly higher cost sharing. The researchers find that the switch caused a 16.45% reduction in total yearly medical expenditures at the firm, leading to approximately $123 million less medical spending per year. The authors break down this total spending reduction into the three components of (i) provider price inflation (ii) consumer price shopping and (iii) quantity reductions and find that almost all of the overall drop in spending is due to quantity reductions. The researchers investigate the reduction in spending as a function of (i) consumer demographics (ii) consumer health status and (iii) types of medical services consumed. They also investigate how consumers respond to the complex structure of the non-linear high-deductible contract, and find that the majority of incremental spending reductions at any point in the calendar year come from consumers who are under the deductible. The researchers investigate whether consumers respond to (i) expected end-of-year marginal prices (ii) spot prices at the time of care and (iii) ex ante average prices and find that consumers respond to both expected end-of-year marginal prices and spot prices.


David Powell, RAND Corporation, and Seth A. Seabury, University of Southern California

Medical Care Spending and Labor Market Outcomes: Evidence from Workers' Compensation Reforms

There is considerable controversy over whether much of the spending on health care in the United States delivers enough value to justify the cost. Powell and Seabury contribute to this literature by studying the causal relationship between medical care spending and labor outcomes, exploiting a policy which directly impacted medical spending for reasons unrelated to health and using a unique data set which includes medical spending and labor earnings. The authors' focus on labor outcomes is motivated by their potential usefulness as measures of health, the importance of understanding the relationship between medical care and labor productivity, and the policy interest in improving labor outcomes for the population that they study - injured workers. The researchers exploit the 2003-04 California workers' compensation reforms which reduced medical care spending for injured workers with a disproportionate effect on workers incurring lower back injuries. The authors link administrative data on workers' compensation claims to pre-injury earnings, post-injury earnings, and earnings information for matched (uninjured) workers at the same pre-injury firm. They then test the effect of the reforms on labor force outcomes for workers who experienced the biggest drop in medical care costs. Adjusting for the severity of injury and selection into workers' compensation, the researchers find that workers with low back injuries experienced a 7.3% greater decline in medical care after the reforms, and that this led to an 8.3% drop in post-injury earnings relative to other injured workers. The authors estimate that this earnings decline is due both to an increase in injury duration and to lower earnings conditional on working. These results suggest jointly that medical care spending can impact health and that health affects labor outcomes.


Maria Polyakova, Stanford University

Regulation of Insurance with Adverse Selection and Switching Costs: Evidence from Medicare Part D

In this paper, Polyakova takes advantage of regulatory and pricing dynamics in Medicare Part D to empirically explore interactions among adverse selection, switching costs, and regulation. She first documents novel evidence of adverse selection and switching costs within Part D using detailed administrative data. Polyakova then estimates a contract choice and pricing model that quantifies the importance of switching costs for risk-sorting. Conceptually, how switching costs affect selection depends on evolution of contract space relative to initial conditions. In Part D, switching costs help sustain an adversely-selected equilibrium and mute the ability of ACA policies to improve risk allocation in this important marketplace.