In chapters 5 and 6 we have established the need in REQM for a theory of collective choice through government institutions. This area of theorizing is in one sense very new and, as we note below, in another sense, very old. In this appendix we delve a little more deeply into the theory to show more clearly the importance of vote trading and the importance of representative government as a vehicle for achieving it.
One line of theoretical development views government as being a quasimarket in the sense that it simply registers and transmits individual preferences, on the basis of the distribution of votes. The prototype model of this sort would be based on the assumption that each public issue is resolved by a referendum universally participated in by fully informed voters. The "market type" formulation of the political process has frequently been criticized by those who note that most political decisions are and must be made through representatives. The market type of model has, however, given rise to a result which has been discussed almost endlessly since it was published in 1951. This is the famous Arrow Paradox. The essential ideas of this paradox of collective choice can be simply illustrated.1
As can be readily shown, sets of individual preferences may lead to unambiguous collective preference. In the situation shown in table 6-A-1, two thirds prefer A to C and two thirds prefer C to B. Thus, if the choice is between B and C and then between A and C, A will have a majority. But two thirds prefer B to A, and a different order of choice can produce any one of the choices. To the aficionados, this phenomenon is known as cycling" or "intransitivity." It may be noted that the result is a consequence of the ordering of choices by one of the participants which leads to the preference for either extreme rather than a middle position. Thus, assume the order ABC proceeds along a spectrum politically. For concreteness, assume A is complete equality of income distribution, B is some redistribution, and C is no redistribution at all. In this case, individual 1 prefers complete equality but would rather have no redistribution at all than a moderate amount. The paradox has sometimes been cited as a
AUTHORS' NOTE: This appendix is based primarily on material prepared by Edwin T. Haefele.____________________