Irrational Economist_ Making Decisions in a Dangerous World - Erwann Michel-Kerjan [96]
In this approach I assumed that the parameters of the economic system, the preferences and production possibilities of the economic agents, were themselves random variables, in general changing over time according to known probability laws. A specification of the values of all these random variables is called a state of nature. I suggested that a market was defined not just by the physical characteristics of the commodity and the time of delivery (as in Hicks) but also by the state of nature prevailing at that time. The unit of transaction was the contingent contract, that is, contingent on the state of nature prevailing.
With this reinterpretation, the standard apparatus of general equilibrium theory could be used to demonstrate the existence and efficiency of equilibrium. I also suggested a role for securities (payable in money, at least in the sense of a unit of account). If individuals had what has been referred to as contingent perfect foresight—that is, the ability to predict the prices that would prevail if a given state of nature occurred—then it would be sufficient to have securities paid in money contingent on states of nature. These securities would be very much like the insurance policies we are well acquainted with, which reimburse you if you suffer a loss that is covered in the insurance contract; the insurance payment is contingent on the loss occurring.
There are indeed a large number of securities with payment conditional on some events. Apart from standard insurance policies, there are common stocks, clearly very numerous. In addition, corporate bonds and bank loans have payments that are nominally fixed in time and amount, but it is fully understood that there is a risk of default from the borrower, an event that is contingent on some events. What is noteworthy, however, is that these securities are contingent not on exogenous states of nature but on occurrences determined at least in part by economic variables within the system, especially prices but also quantities. A firm will fail to repay a bond or will pay a dividend when the input prices are high relative to the output prices. That the contingent contracts in the real world depend in major part on variables endogenous to the economic system has been especially emphasized by Mordecai Kurz (1974 and 1996, among other papers).
LIMITED INFORMATION OF ECONOMIC AGENTS
Probably the most important innovation in economic theory in the last fifty years has been the emphasis on what has been called asymmetric information. The term information is properly and appropriately used in the sense given by mathematical statistics or communication theory: an observation on one random variable that changes the probability distribution of some random variables relevant to decision making. It is clear and reasonable that different individuals have access to different information; they have varying life experiences and varying opportunities to make observations. A number of economists came to stress this concept from varying points of view; I myself came to it by considering the economics of medical care (Arrow, 1963).4 Insurance companies had long understood the consequences of asymmetry of information under such headings as moral hazard and adverse selection.
It is not always recognized that the most neoclassical approaches in economics also assume asymmetry of information. It is a standard claim for the usefulness of a system of markets that it requires an individual to know only his or her own utility function and production possibilities. The only information about the rest of the world is contained in the prices.
This has