Journal of Public Economics 39 (1989) 33344. North-Holland THE ROLE OF INTEREST RATE TAXES IN CREDIT MARKETS WITH DIVISIBLE PROJECTS AND ASYMMETRIC INFORMATION David DE MEZA University of zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONML Exeter, Exeter EX4 4RJ, UK David C. WEBB* London School of Economics, London W CZA ZAE, UK Received September 1988, revised version received February 1989 This paper assumes that entrepreneurs have divisible projects with random payoffs. However, managerial skill is unobservable to outsiders. Projects are assumed to be financed through debt. A pooling equilibrium is shown to exist in which good and bad entrepreneurs sell debt at the same price. If both types of entrepreneur have a positive probability of bankruptcy, project scale will be too large. This provides a prima facie case for raising interest rates and hence the cost of capital through taxation. If in equilibrium only the poor entrepreneurs go bankrupt, there is an outlying possibility that interest rates are too low and a subsidy is called for. 1. Introduction Banks can rarely identify the precise credit worthiness of potential borrowers. As a result, good risks typically end up paying higher interest rates than their characteristics merit, whilst poor risks obtain credit on terms that are better than actuarially fair. This raises questions of whether such an equilibrium is efficient and, if not, of what form of intervention is appropri- ate. In earlier papers [de Meza and Webb (1987, 1988)] we established that if projects are of fixed size and their quality is better known to entrepreneurs than to banks, then, under reasonable assumptions about the distribution of returns, an interest rate tax is efficient. This result contrasts with widely held intuitions concerning the performance of credit markets under asymmetric information [e.g. Akerlof (1970) Leland and Pyle (1977)] and may provoke suspicions that important elements of reality are missing from the model. The reason the tax is beneficial in our 1987 paper is because it improves the average quality of projects financed. In its absence, low yielding projects *We appreciate the constructive comments of two referees which have proved invaluable in revising this paper. 0047%2727/89/$3.50 0 1989, Elsevier Science Publishers B.V. (North-Holland)