This article reconsiders the durable goods monopoly problem when the monopolist's marginal cost is private information. We show that the Coase Conjecture implies the No Trade Theorem: In any equilibrium in which the lowest-cost seller's initial offer approaches her marginal cost, the aggregate probability of trade must vanish. However, we also construct non-Coasean equilibria which approximate the unique outcome of the rental version of the same model. These (stationary) equilibria are comparatively efficient. The results are equally applicable to the mathematically-equivalent problem of sequential bargaining with two-sided incomplete information where one party makes all the offers.

In this paper, we examine the impact of political constraints on economic reform plans, with special reference to the transition from centrally planned to market economies. We analyse the problem of an agenda-setting reform-minded Government facing a bureaucracy or industrial sector for which allocative efficiency requires redundancies and an increase in work intensity. The Government also tries to minimize the rents conceded to its heterogeneous workforce. We examine two types of political constraints: unanimity rule and majority rule, both in a one-period and a two-period horizon. The main results are the following. First of all, we show how adverse selection and time-consistency may generate the widely-observed feature of gradualism as an ingredient of an optimal reform. Second, under a majority rule, it is shown to be possible for the Government to obtain a majority vote for a reform scheme that intertemporally hurts majority interests. Indeed, the Government can improve rent extraction through the strategic use of the threat of future proposals: the group which expects to be in the minority tomorrow may accept concessions, while its votes can be used to extract rents from another group. These results suggest that, in a dynamic context, democratic constraints should not be overestimated as an obstacle against efficiency-enhancing economic reforms. The results of this paper may throw some light on the political economy of current reforms in Eastern Europe.

Estimated U.S. M1 demand functions appear unstable, regularly “breaking down,” over 1960–1988 (e.g. missing money, great velocity decline, M1-explosion). We propose a money demand function whose arguments include inflation, real income, long-term bond yield and risk, T-bill interest rates, and learning curve weighted yields on newly introduced instruments in M1 and non-transactions M2. The model is estimated in dynamic error-correction form; it is constant and, with an equation standard error of 0–4%, variance-dominates most previous models. Estimating alternative specifications explains earlier “breakdowns,” showing the model's distinctive features to be important in accounting for the data.