This paper is about criteria for assessing alternative political institutions in a society committed to popular sovereignty. The issues are classic, and we do not attempt a comprehensive review of the ways in which they run through political and economic thought since Aristotle. An aggregative perspective on political institutions, which describes politics in terms of aggregating exogenous, prior preferences of citizens, is contrasted with an integrative perspective, which describes politics more in terms of the development of preferences within a framework of rights and norms. The evaluation of aggregative processes and institutions highlights questions of efficiency, preferences, and endowments. The evaluation of integrative institutions focuses on questions of competence and integrity. Many of the most influential contemporary discussions of political institutions are based on an aggregative framework that is relatively inattentive to issues of political integration. This contemporary emphasis on metaphors of aggregation, however, represents a phase in a history of oscillation between aggregation and integration in politics and political theory, and there are signs that the emphasis may shift in the next few decades to a greater consciousness of integration. Although the cyclic process is not escaped through awareness of it, such awareness may provide a basis for a more thoughtful approach to the search for appropriate institutions within a political system dedicated to popular sovereignty.

Politicians are motivated primarily by the desire to avoid blame for unpopular actions rather than by seeking to claim credit for popular ones. This results from voters' ‘negativity bias’: their tendency to be more sensitive to real or potential losses than they are to gains. Incentives to avoid blame lead politicians to adopt a distinctive set of political strategies, including agenda limitation, scapegoating, ‘passing the buck’ and defection (‘jumping on the bandwagon’) that are different than those they would follow if they were primarily interested in pursuing good policy or maximizing credit-claiming opportunities. These strategies in turn lead to important policy effects, including a surrender of discretion even when it offers important credit-claiming opportunities.

The Conservative Government elected in Britain in 1979 wished to change the extent and pattern of government expenditure. We use econometric techniques to investigate whether it did so in the period up to 1984, concentrating attention on the welfare state. We also test the hypothesis that the observed changes favoured the middle classes. After discussing the channels by which the middle classes influence government policy, a model of government behaviour is outlined. The theoretical model indicates the forms of specification error that we might expect in our econometric results, which, in turn, suggest that the Conservatives tended to favour the middle classes, while the previous Labour administrations did not. However, the estimates for the Labour period appear to be misspecified, but those for the Conservative period survived tests of misspecification.

The Gift Relationship (Titmuss, 1970) argued that the UK system of voluntary blood donation was economically, medically, and morally superior to the American system, in which many blood suppliers are paid. The argument was badly put and fiercely attacked, but substantially correct. Developments since Titmuss wrote, especially AIDS, have tragically reinforced his main point: more than half of the severe haemophiliacs in the UK have been infected with the AIDS virus, almost certainly through imported blood products. Titmuss's normative argument is reconstructed, and defended against neoclassical economists who have been his fiercest critics. Finally, policy implications are briefly discussed.