EVENTS

Date:	Tuesday, August 5, 2003
Time:	10:00 AM -- 1:30 PM
Location:	Wohlstetter Conference Center, Twelfth Floor, AEI 1150 Seventeenth Street, N.W., Washington, D.C. 20036

August 2003
Passing the Buck on Medicaid: How the States Pledge and the Feds Pay

Rapidly growing Medicaid expenditures, which constitute over 20 percent of state
budgets, are now an urgent action item for state legislatures. Many states are considering cutbacks and new ways to control cost--all are demanding increased federal funding. What are the causes of this latest Medicaid crisis? Are they forces beyond the states’ control: rising healthcare costs, increasing poverty, and federal mandates? Or have the states brought this crisis on themselves?

On August 5 at AEI, Michael S. Greve provided his own answers to these questions based on a study (coauthored with Jinney Smith) of differences in state Medicaid expenditures in good and bad economic times (1994–2000). They were joined in discussion by Dennis Smith of the Department of Health and Human Services and Alan Weil of the Urban Institute. AEI’s Robert B. Helms moderated. Following the panel, Nelson J. Sabatini, secretary of health and mental hygiene for the State of Maryland, delivered a luncheon address.

Michael Greve and Jinney Smith
AEI and Northwestern University

The purpose of this paper, “What Goes Up May Not Come Down: State Medicaid Decisions in Times of Plenty,” is to examine the explosion of Medicaid expenditures in the years between 1994 and 2000 and to try to understand the state-level decisions that have turned Medicaid from a presumably countercyclical economic safety net into a perpetually escalating drain on state budgets.

The period from 1994–2000 was a time of sustained economic growth and declining poverty. Yet while the poverty population declined 18.4 percent, Medicaid spending (state and federal combined) increased by 30.8 percent. Our research suggests that the culprit was neither healthcare inflation nor any increase in federal mandates or poverty rolls. Rather it was a substitution effect brought about by surging state revenues and the conscious decision on the part of state legislatures to sustain and expand Medicaid expenditures at an “affordable” rate. Eligibility and coverage expansions (including the popular waiver system), combined with the uncapped federal matching program, allowed states to use their rising tides to lift all boats at a maximum cost of only 50 cents on the dollar.

Our findings are consistent with three conclusions. First, the growth of Medicaid expenditures has been fueled by policy decisions at the state level--a national pressure such as health care inflation would not strike Vermont and skip New Hampshire. State spending has generally increased but has varied too much to reflect a single influence. Second, the federal funding structure encourages states to expand Medicaid, regardless of the economic situation at hand. In a state with a matching percentage of 60 percent, for every dollar directed to Medicaid, the state receives 60 cents from the federal government; for every dollar cut from the program, the state recovers only 40 cents. Third, this expansionist approach to Medicaid provides a roadmap for the implementation of universal socialized healthcare. The states have demonstrated what they do when left to their own devices with federal money. Continued bailouts will only further the shift away from private markets.

Dennis Smith
Department of Health and Human Services

Medicaid serves three very distinct populations and really is two very distinct programs.
Low-income families, the low-income elderly, and people with disabilities all benefit from the healthcare that the Medicaid program has to offer. Seniors in particular have benefited because Medicaid has covered areas that Medicare does not--most notably prescription drugs. This view of Medicaid eligibility is radically different from the program’s original goal of providing healthcare to families on welfare. In fact, if Medicaid were the program that it was in 1965, it would be much smaller and much less expensive.

As it has evolved, however, Medicaid has taken on larger goals. In addition to the eligibility expansion, Medicaid has developed into two major versions of care: long-term and acute. Long-term care expenditures are rising rapidly--the current annual cost is $100 billion per year; by 2013, that figure will be $225 billion. The Bush administration has made it a priority to encourage the development of new models for the delivery of long-term care, particularly at the state level, with an emphasis both on recipient satisfaction and cost containment. In terms of acute care, we believe that the State Children’s Health Insurance Program (SCHIP) provides a model for reform. SCHIP has demonstrated the workability of both capped funding and increased state flexibility.

The mission of Medicaid has grown, and reform proposals should recognize and attempt to incorporate that fact. The Medicaid that we have before us today is a vitally important program, and reform is no longer a question of “if,” but “when.”

Alan Weil
Urban Institute

There are five major analytical problems with Greve and Smith’s paper. First, and most importantly, the premise of the paper--which is that the Medicaid program should be cyclical--is belied by the authors’ data. Seventy-five percent of all Medicaid expenditures go to benefit the elderly and disabled, not the poor, and therefore the base unit of analysis--total state Medicaid spending when contrasted with poverty rates or population--shows nothing that we do not already know.

Second, the regression model included is of questionable value because the same variable is used on both sides of the equation. For basic statistical reasons, the model cannot be used to explain anything.

Third, in 1994 state spending on several aspects of the Medicaid program was higher because of political circumstances, not the least of which was the suggestion of 1994 as a base year for future Medicaid block grants. More importantly, using only a period of economic growth skews the observations because during periods of economic difficulty states are likely to put off inflationary adjustments to service providers. When budgets increase again, the belated adjustments are made in one shot. Thus, what looks like a sudden surge in spending may only be a delayed response to inflation.

Fourth, with the implementation of programs like SCHIP, Medicaid included an eligible population with a relatively low cost of care: children. Therefore it is difficult to make conclusions based on the per capita costs of the program. Because of the varying rates of SCHIP implementation, state per capita costs for Medicaid do vary for this time period.

Fifth, after rejecting the tested hypotheses, the paper assumes the validity of its only unexamined theory.

In terms of the paper’s conclusions, yes, state decisions have driven Medicaid spending. However, ending coverage to “optional” populations would mean taking an eighty-five-year-old widow with a pension income of $10,000 per year out of her nursing home. As for the expansionist nature of Medicaid, it is a basic principle of the Medicaid program that states do not have to bear the full marginal costs of increased spending--and therefore the program is inherently expansionary, which is neither a bad thing nor very controversial. Finally, the current Medicaid structure is in no way an optimal framework for the implementation of a single-payer system.

Discussion

Jinney Smith
Northwestern University

In terms of the regression analysis equation, had the two variables (total Medicaid growth, 1994–2000 and Medicaid growth as a percentage of state revenue, 1994–1997) been too highly intercorrelated to give a reliable conclusion, that problem would have surfaced in the intercorrelation checks and in the standardized coefficients of the variables. As the regression chart shows, the other three variables taken together account for more variation than the growth of Medicaid as a percentage of state revenue. In theoretical terms, that variable was included to examine the effect of early budget choices on later budget decisions.

Michael S. Greve
AEI

A disjunction exists between the expert and public perceptions of the Medicaid program. Medicaid has been disconnected from its original purpose for many years, but most people still believe that it is a program for the poor. Pointing that out is a useful exercise. In reference to the increased heterogeneity of Medicaid, it is true that demographics and technology use have differed across states, but it would be very surprising if those variables alone could account for the spending variations. Finally, the expansion of Medicaid is a path to socialized medicine because the lack of final responsibility will allow shifts that no eligible politician would openly advocate.

AEI research assistant Kate Crawford prepared this summary.