Abstract

Anthony Downs is unusual in having been active in economic and political research during the 1950s and 1960s, and then turning to real estate and urban planning for the rest of his career. He quitted political economy in spite of the great and enduring success of his first book, "An Economic Theory of Democracy" (1957). A new Ms on the "Theory of Democracy" is currently circulating. This paper concentrates on the "Economic Theory", starting from the fact that the book presents not one, but many theories of democracy even if they are all generally couched in economic terms. Some of these theories (or models) of party competition and voting have mutually contradicting predictions, such as parties sticking to the same policies to demonstrate reliability and responsibility, while in terms of the two-party spatial model they will converge in policy terms. The confusion is heightened when assumptions and conclusions from one model are grafted on to another in the text. This paper expounds the argument of the "Economic Theory" systematically and sequentially, covering: a) the general non-spatial model of party competition; b) the effects of information shortage; c) party reliability and responsibility; d) the two party spatial model; e) the multi-party spatial model; f) further consequences of information shortage; g) problems of voting turnout. Evidence from quantified election programmes is used to show that the models leading to limited party policy movement and non-convergence are the more realistic ones. The Appendix provides the full set of 22 assumptions necessary to derive Downs' proportions at the end of the "Economic Theory". In spite of its defects the book laid down the agenda for mathematical political theory and empirical research over the last 50 years and has a fair claim to being the most influential single book written on politics in that period.