Health Insurance Subsidies Research Articles

Health care financing among the elderly: who really pays the bills?

This paper investigates the issue of who pays the health care bills of the elderly by considering the types of subsidized health insurance protection enjoyed by the noninstitutionalized elderly and the way that increased Medicare cost-sharing efforts in the 1980s are affecting those without additional health insurance subsidies. In making this examination we estimate the out-of-pocket health care expenditures of the elderly either directly or as nonsubsidized medigap premiums by income level, taking into account four types of health insurance subsidies received by elderly persons: Medicare, Medicaid, Veterans Administration health care, and subsidized health insurance from either current or former employers. We find that increased cost sharing is likely to fall most heavily on those elderly least likely to afford it: the poor and near-poor elderly who have only Medicare as a health insurance subsidy, particularly those who are older and sicker and who use Medicare services more heavily. These persons are caught between well-intentioned federal cost-cutting efforts and the often confusing panoply of health insurance programs for the aged, and they will bear an inequitably large portion of any future Medicare cost-sharing initiatives.

Deinstitutionalization in Switzerland: The Incrementalist Model

Processes of deinstitutionalization of mental health care are particularly difficult to assess in Switzerland because of the complexity of the country's health care delivery system. Each of the 26 states, or cantons, that make up the Swiss Confederation is fully autonomous in structuring its health care facilities. Administration and financing of hospitals are a cantonal responsibility; most psychiatric hospitals are publicly owned. It is only a terms of federal financing, health insurance subsidies, and issues of consensus of the medical profession that some nationwide parameters impinge upon the diverse policies of the cantons. Deinstitutionalization of psychiatric services is precisely one of the processes in which cantonal differences have manifested themselves in the extreme. Six cantons have no psychiatric hospitals and rely on service agreements with neighboring ones; the five cantons with medical schools Geneva, Vaud, Berne, Basel City, and Zurich have university psychiatric hospitals in addition to state psychiatric hospitals. These five cantons together supplied 47.5% of Swiss psychiatric beds in 1975 [1]. The number of beds for psychiatric care per 1,000 inhabitants ranges from a high of over 6, in Appenzell (because of high out-of-state utilization), to a low of 1.2, in the Valais; the average is 2.4. There are similar variations in the distribution of psychiatrists, the national average in 1979 be-

Abstract: Recursive Experimental Design for Econometric Research: The Multiple Response Case

Until recently, most interest in econometrics was directed toward the data analysis problem. However, as opportunities for controlled experimentation in economics and business become more common, econometricians' interest in the statistical theory of experimental design is a natural consequence. Design techniques have and are being applied in marketing research, and in the contexts of such policy areas as the negative income tax and manpower experiments, the housing subsidy and health insurance subsidy experiments, and others. Experimentation in economics and business is an extremely complex research undertaking compared to the experimentation in other fields (such as agriculture, physics or chemistry) for which most of the design theory was developed. So, control experimentation in business and economics tends to be expensive in dollars, time and research talent. All this suggests the importance of carefully designing experiments to maximize their information output (efficient design) and the need for econometricians to extend the design theory in a number of directions of particular interest to economists and business researchers. This paper provides a solution to the problem of experimental designs for various multiple equation regression models frequently encountered by business and economic researchers.