The effects of endogenously-generated information on specific economic institutions.

Abstract

The effects of endogenously generated information on decision making is studied within three economic institutions. A non-market institution, (voluntary contributions for the provision of a public good), and a market institution, (sealed bid auctions), which both have been extensively studied as non dynamic institutions were used to examine the effects of dynamically generated information. A third institution, the double auction, which has been studied as a dynamic institution, was used to develop prototype methodology for the simulation and estimation of decision strategies with dynamic environments. Experimental results are presented showing the effects of allowing real time revisions of voluntary contributions for the provision of a public good. Four public good payoff functions are examined, each of which generates specific equilibria. Evidence of increased provision of the public good is demonstrated for: (i) the case in which revisions are limited to increases and a provision point exists, and also (ii) the case in which there is a high initial marginal return from the public good. An experimental investigation of sequential first and second price sealed bid auctions is conducted examining the effects of known capacity constraints on bidding behavior. Results are presented and compared to a Nash equilibrium model developed by Robert Weber. These naturally occurring inter-auction limitations are shown to significantly affect revenue generation capabilities of sealed bid auctions. Bidding and price behavior is presented and is found to be inconsistent with the model. A methodology is proposed as a means for attempting to identify and estimate strategies utilized by participants within dynamic institutions. The methodology was used on the double auction. Computer automata are used to examine the contracting behavior of a variety of decision rules. Simulations of the decision model are run using parameters common to past experiments. The resulting contracts are compared to corresponding experiment contracts and are found to follow similar patterns. Two double auction experiments, each with ten subjects were used to generate actual contracting data. The parameters of the proposed decision model are estimated from data on the individual decisions of the subjects.