Frontiers in Health Policy Research
The NBER's fifth annual conference on "Frontiers in Health Policy Research," organized by Alan Garber, took place on June 7 in Bethesda, Maryland. The program was:
Meltzer and Chung use data from California in 1983 and 1993 on hospital charges and cost-to-charge ratios to examine the effects of competition on costs. They consider both high and low cost admissions and their costs before and after the establishment of the Medicare Prospective Payment System (PPS). Comparing persons above and below age 65, they find that competition is associated with increased costs in both age groups before PPS, but decreased costs afterwards. This is especially true among those over age 65 with the highest costs. The authors conclude that the combination of competition and prospective payment systems may result in incentives to selectively reduce spending among the most expensive patients. This implies the need to carefully monitor outcomes for the sickest patients under prospective payment systems in competitive environments.
Many health maintenance organizations (HMOs) have exited the market for Medicare managed care; since 1998 the number of participating plans has fallen from 346 to 174. Cawley, Chernew, and McLaughlin study how the equilibrium number of HMOs participating in Medicare managed care markets varies with the Health Care Financing Administration (HCFA) capitation payment. They have data from virtually every county in the continental United States, plus the District of Columbia, from 1993-2001. The authors find that in 2001, only 12.3 percent of counties in the 48 contiguous states received a HCFA payment greater than what the authors estimate is necessary to support a single HMO in the Medicare managed market. HCFA particularly appears to underestimate the payment necessary to HMOs in rural and sparsely populated areas. In order to support a single HMO in half of the U.S. counties in the year 2001, HCFA would have to pay $799.24 per average enrollee per month in the median county, the authors estimate. The high payments necessary to support HMO participation in rural counties suggest that in many U.S. counties it may not be possible to achieve both of the goals of the Medicare managed care program: to offer an attractive alternative to fee-for-service Medicare and to save money.
What impact does Medicare have on the health of the United States population? Lichtenberg's goal is to obtain precise estimates of medical utilization and outcomes, by age, for those close to age 65. Using information obtained from medical providers (hospitals and doctors) pooled over a number of years, he finds that utilization of ambulatory care and, to a much smaller extent, inpatient care, increases suddenly and significantly at age 65. This is presumably because of Medicare eligibility. The number of physician visits in which at least one drug is prescribed also jumps up at age 65. Reaching age 65 also has a strong positive impact on the consumption of hospital services, but most of this impact appears to be the result of postponement of hospitalization in the prior two years. Does this increase in utilization lead to an improvement in outcomes - a reduction in morbidity and mortality - relative to what one would expect given the trends in outcomes prior to age 65? Yes, Lichtenberg finds. The Medicare-induced increase in health care utilization leads to a reduction in days spent in bed of about 13 percent and to slower growth in the probability of death after age 65. Physician visits also appear to have a negative effect on the male death rate. Data on age-specific death probabilities every 10 years back to 1900 - both before and after Medicare was enacted - provide an alternative way to test for the effect of Medicare on longevity. The data support the hypothesis that Medicare has increased the survival rate of the elderly by about 13 percent.
What options are available in the design of refundable tax credits for health insurance purchases? What are the effects of different options, and how should they be evaluated? Pauley and Herring compare three alternative types of tax credit schemes: fixed dollar credits toward the purchase of a benchmark basic plan; proportional credits toward the purchase of a benchmark plan; and fixed dollar credits which can be used for the purchase of any plan with a premium at least as large as the credit. They show that is inappropriate to evaluate alternative credit designs primarily on their effect on the number of uninsured persons. Credits claimed by people who were already insured can have very beneficial equity and efficiency effects. Using two new modeling techniques, the authors show that credits will have very little effect until they exceed a certain threshold. Above that level, credits which cover on average about half of the premium of a benchmark policy may substantially reduce the number of uninsured, especially if they take the fixed-dollar form. These new estimates illustrate the tradeoffs in program design. The remaining key questions are what value consumers place on avoiding charity or bad debt care and what value policymakers place on increasing the level of coverage of different types of formerly uninsured persons.
In their second paper, Pauly and Herring discuss various options for using refundable tax credits to reduce the number of uninsured persons. The effect of tax credits on the number of uninsured depends on the form of the credit scheme adopted. Moreover, since large subsidies for private insurance directed to low-income persons have never been implemented, there is considerable uncertainty about the effects of various tax credit proposals. The authors find that small credits will do little to reduce the number of uninsured but that credits covering about half of the premium for a benchmark policy might have a significant effect, especially if they take a fixed-dollar form and can be used for policies with few restrictions. Finally, Pauly and Herring discuss the issues surrounding the "costs" of these credits schemes, and the policy issues raised by the uncertainty of the effects.
Sloan reviews recent empirical evidence on the effects of hospital ownership conversions on the quality of care and the provision of public goods, such as uncompensated care. He presents new results on these topics based on hospital discharge data from the Healthcare Cost and Utilization Project's (HCUP) Nationwide Inpatient Sample. His analysis reveals that conversions from government or private non-profit to for-profit ownership had no effect on in-hospital mortality, but rates of pneumonia and complications increase following conversion to for-profit status. Other research, discussed in the paper, has found increased mortality rates at one year following admission for patients admitted to hospitals that had converted to for-profit ownership. There was no effect of such conversions on the propensity to admit uninsured or Medicaid patients. Overall, the evidence suggests a role for public scrutiny of ownership conversions prior to the occurrence of such transactions.
These papers will be published by the MIT Press in an annual conference volume. They are also available at Books in Progress under the title Frontiers in Health Policy Research, Volume 5.