During the 1960s and early 1970s, strong economic growth and highly expansionary income security policies led to considerable progress for the entire American population with respect to major income security goals. However, in the last fifteen years much of this progress has been either arrested or reversed, particularly for the non-aged, as economic growth slowed and income security policies ceased to expand and, in some cases, contracted. This retrenchment was the inevitable consequence of numerous phenomena which preceded, and were reinforced by the Reagan era. American income security policies are not likely to contract generally in the future, nor to resume expanding in a direction characteristic of many Western European welfare states. Rather, the prospects are for slow economic growth, higher targetting of programs by income in some areas, and marginal expansions requiring minimal new commitments of public resources in others. Major income security problems, especially among the lower income population, will remain.

Advocates of pervasive social experimentation neglect to consider the effect that suchexperimentation might have on citizens' respect for the law and their inclination to obey it. Since habit plays a large part in law abidingness, frequent alterations in law, even when substantively justified on narrow policy grounds, weaken the power of law in general. Because of their great utility, we should use social experiments occasionally when they can shed light on crucial dimensions of truly important policy questions. But because pervasive experimentation threatens law abidingness and citizen support for the regime, we should decide against some otherwise justifiable experiments and neither expect nor wish for the sort of experimenting society advocated by many social scientists.

This paper aims to redress the asymmetry in the standard economic treatment of the shortcomings of markets and governments by developing and applying a theory of ‘nonmarket’ failure– that is, of government failure – so that the comparison between markets and governments can be made more systematically, and choices between them arrived at more intelligently. Several conclusions are drawn. First, the choice between markets and governments is not a pure one, actual systems inevitably involve combinations between markets and governments. Second, with respect to both static and dynamic efficiency criteria, markets generally do better than governments. Third, there are various ways in which government can contribute to improving the functioning of markets. Fourth, market forces can play a useful role in improving the functioning of government and reducing the incidence of nonmarket failures.

The quality of health care services has recently become a prominent political issue in both Britain and America. Everyone is in favour of quality. Yet closer inspection reveals that the term is being used with several different meanings, and that its fashionability seems intimately connected with the collision, in both countries, between rising health care costs and a desire on the part of the purchasers or paymasters of health services to restrict their expenditures. The different approaches to assessing quality are not neutral, technical matters. Each carries implications for the future relationships among health care providers, and between providers, payers and patients. Furthermore, changes in these institutional relationships may well lead the debate back to the larger, traditional and more obviously ‘political’ issues of resource allocation, equity and the role of the state.