How would mental health parity affect the marginal price of care?

OBJECTIVE: To determine the impact of parity in mental health benefits on the marginal prices that consumers face for mental health treatment. DATA SOURCES/DATA COLLECTION: We used detailed information on health plan benefits for a nationally representative sample of the privately insured population under age 65 taken from the 1987 National Medical Expenditure Survey (Edwards and Berlin 1989). The survey was carefully aged and reweighted to represent 1995 population and coverage characteristics. STUDY DESIGN: We computed marginal out-of-pocket costs from the cost-sharing benefits described by policy booklets under current coverage and under parity for various mental health treatment expenditure levels using the MEDSIM health care microsimulation model developed by researchers at the Agency for Healthcare Research and Quality. Descriptive analyses and two-limit Tobit regression models are used to examine how insurance generosity varies across individuals by demographic and socioeconomic characteristics. Our analyses are limited to a description of how parity would change the marginal incentives faced by consumers under their existing plan's cost-sharing arrangements for mental and physical health care. We do not attempt to simulate how parity might affect the level of benefits, including whether benefits are offered at all, or the level of managed care that affects the actual benefits that plan members receive. Rather, we focus only on the nominal benefits described in their policy booklets. PRINCIPAL FINDINGS: Our results show that as of 1995 parity coverage would substantially reduce the share of mental health expenditures that consumers would pay at the margin under their existing plan's cost-sharing provisions, with larger changes for outpatient care than for inpatient care. Because current mental health coverage generally becomes less generous as expenditures rise, while coverage for other medical care becomes more generous (due to stop-loss provisions), the difference in incentives between current mental health coverage and the assumed parity coverage widens as total expenditure grows. We also find that the impact of parity on marginal incentives would vary greatly across the privately insured population. CONCLUSIONS: Based on the large variation in the impact of parity on marginal incentives across the population under current plan cost-sharing arrangements, changes in the demand for mental health treatment will likely also vary across the population.

Reconciling medical expenditure estimates from the MEPS and the NHA, 1996.

This article compares 1996 estimates of national medical care expenditures from the Medical Expenditure Panel Survey (MEPS) and the National Health Accounts (NHA). The MEPS estimate for total expenditures in 1996 was $548 billion; whereas, the NHA estimate for personal health care (PHC) in 1996 was $912 billion. Much of this apparent difference, however, arises from differences in scope between MEPS and NHA--rather than from differences in estimates for comparably-defined expenditures. We adjusted the NHA for differences in included populations and types of services covered, finding a much smaller difference between MEPS and a comparably-defined NHA.

Designing employer health benefits for a heterogeneous workforce: risk adjustment and its alternatives.

Many health economists recommend that employers provide employees with a risk-adjusted choice among competing health insurance plans. However, formal risk adjustment is rarely if ever used by employers. This paper examines a range of health benefit design options available to employers, focusing attention not only on risk adjustment but also on its alternatives. We argue that while formal risk adjustment is rare, employers commonly use strategies that accomplish some of the same objectives at less cost.

How much did the Medicaid expansions for children cost? An analysis of state Medicaid spending, 1984-1994.

The authors examine the relationship between the Medicaid eligibility expansions for children and state Medicaid spending during the period from 1984 to 1994. They find that the Medicaid expansions had relatively low incremental cost per enrollee--substantially below the average Medicaid expenditure for children. Expansion children tend to be older and have fewer disabilities. Moreover, many of the most expensive expansion children would have been covered by Medicaid-medically-needy provisions had the expansions not occurred. The authors examine the implications of our findings for intensified Medicaid outreach efforts and for the State Children's Health Insurance Program.

Cross-subsidization in the market for employment-related health insurance.

This paper uses data from the 1987 National Medical Expenditure Survey to examine the nature of equilibrium in the market for employment-related health insurance. We examine coverage generosity, premiums, and insurance benefits net of expenditures on premiums, showing that despite a degree of market segmentation, there was a substantial amount of pooling of heterogeneous risks in 1987 among households with employment-related coverage. Our results are largely invariant to (i) firm size and (ii) whether or not employers offer a choice among plans. Our results suggest the need for caution concerning incremental reforms that would weaken the link between employment and insurance without substituting alternative institutions for the pooling of risks.

Medical savings accounts: microsimulation results from a model with adverse selection.

This paper examines medical savings accounts combined with high-deductible catastrophic health plans (MSA/CHPs), exploring the possible consequences of making tax preferred MSA/CHPs available to the entire employment-related health insurance market. The paper uses microsimulation methods to examine the equilibrium effects of MSA/CHPs on health care and non-health care expenditures, tax revenues, insurance premiums, and exposure to risk. If MSA/CHPs are offered alongside comprehensive plans, biased MSA/CHP enrollment can lead to premium spirals that drive out comprehensive coverage. Our estimates also raise concerns about equity, insofar as those who stand to lose the most tend to be poorer and in families with infant children.

Premium subsidies for health insurance: excessive coverage vs. adverse selection.

The tax subsidy for employment-related health insurance can lead to excessive coverage and excessive spending on medical care. Yet, the potential also exists for adverse selection to result in the opposite problem-insufficient coverage and underconsumption of medical care. This paper uses the model of Rothschild and Stiglitz (R-S) to show that a simple linear premium subsidy can correct market failure due to adverse selection. The optimal linear subsidy balances welfare losses from excessive coverage against welfare gains from reduced adverse selection. Indeed, a capped premium subsidy may mitigate adverse selection without creating incentives for excessive coverage.

The impact of ownership type on nursing home outcomes.

This paper examines the sorting of residents between for-profit and nonprofit nursing homes and the health outcomes of those residents conditional on ownership type. Using data from the 1987 National Medical Expenditure Survey, we find evidence of systematic sorting of residents by ownership type, and significant effects of ownership type on outcomes. These results are broadly consistent with the hypothesis that for-profit and nonprofit homes exploit their informational advantages to differing extents in a market characterized by asymmetric information.

Health care expenditures and GDP: panel data unit root test results.

This short paper presents unit root test results for time series on per capita national health care expenditures and gross domestic product in the OECD. Unlike the country-by-country test used by [Hansen, P., King, A., 1996. The determinants of health care expenditure: A cointegration approach. J. Health Econ, 15, 127-137], the test we employ exploits the panel nature of the OECD data. Using this approach, we are able to reject the null hypothesis that these series contain unit roots. No single test is likely to be definitive in this rapidly-evolving area of econometric research; however, our results help to mitigate concern that panel data analyses of national health care expenditures are misspecified.

How are net health insurance benefits distributed in the employment-related insurance market?

The recent health care reform debate has questioned whether the health insurance market effectively pools risks and transfers income across states of health. We use data from the 1987 National Medical Expenditure Survey to examine how net health insurance benefits are distributed in the employment-related insurance market. We find this market to transfer income from those in good health to those with health problems and the tax subsidy from employer health insurance contributions to be a crucial determinant of the net benefit distribution. To the extent society views these transfers as meritorious, our findings suggest caution regarding initiatives to limit or eliminate the tax subsidy.

Uncertainty and health care spending by the poor: the health capital model revisited.

This short theoretical paper examines health care utilization by the poor in the context of Grossman's (1972) health capital model. Earlier work has shown that uncertainty can reduce the attractiveness to the poor of health capital investments. This paper demonstrates that uncertainty can also have precisely the opposite effect. That is, while there may be cases in which it is excessively risky for the poor to invest in their health, there may also be cases in which the poor can ill-afford the risks that arise from not making timely investments in their health.

A model of capitation.

This paper presents a theoretical model of capitation contracts. The consumer's ex ante choice of medical plan is derived under flexible assumptions about provider-patient decision-making. The optimal medical plan is shown to combine full insurance with a provider payment system that is a mixture of capitation and partial reimbursement of provider costs. This solution strongly parallels the 'mixed payment' system derived by Ellis and McGuire (1986, 1990) in the context of prospective payment, though the optimal medical plan derived below may in fact be preferred to that solution in a world with endogenous admissions.