Knowledge that Transforms

In a recent paper in this Review, Sanford Grossman and Joseph Stiglitz challenge the Efficient Markets notion that in an asset market any time prices fully reflect all available (p. 404). The purpose of this comment is to argue that certain theorems and comments of Grossman and Stiglitz do not necessarily follow when an error in the specification of the informed trader's demand function for the risky asset is corrected. In Theorem 5(b) (p. 401), Grossman and Stiglitz show that when informed traders have perfect information concerning a risky asset's yield, no overall equilibrium will exist; that is, there will be no equilibrium value for the proportion of informed traders, X. Inspection of equations (8), (8'), and (9) (p. 396) helps to give intuitive reasons why this is so. If no individual purchases an observation of the information point 0, that is, if X = 0, then for a sufficiently small value of C, the cost of obtaining an observation of the information point 0, an individual could increase his expected utility by purchasing the information. If some fraction of traders purchase the information point 0, that is, if X >0, from equation (8) it is seen that as 0,2 0 (as informed traders obtain better information) small movements in 0 have a nearly infinite effect upon the informed trader's demand for the risky asset and hence upon the asset's price. Thus, according to Grossman and Stiglitz, for a fixed X > 0, as q2 0, the price system becomes perfectly informative as to the informed trader's observation of 0. Thus for X > 0 and a,2 = 0, no individual will wish to purchase 0 at cost C. But for XA-c2 0 and small C it pays an individual to purchase 0. Clearly then for q,2= 0 no equilibrium exists. Equation (8) plays a crucial role in Theorem 5(b). However, equation (8) ignores the fact that X1i, the ith informed trader's demand for the risky asset, is constrained by his wealth, Woi. In fact, if the informed trader has perfect information (so that 0 is the return on the risky asset), then the informed trader simply invests in the asset with the higher yield. Thus for UE2 = 0, equation (8) ought to read

A new study of benzene emissions from maleic anhydride plants - the subject of a standard proposed by the Environmental Protection Agency in April, 1980 - illustrates how benefit-cost techniques can provide useful guidance to decision makers, even with limited information, and how incentive-based approaches can provide workable and efficient alternatives to standards. The case study also suggests that regulators and economists both need to shift their foci from emissions to damages; that efficient regulations must be sensitive to variations across sources in marginal damages, as well as to differences in the costs of controlling emissions. 3 references, 3 tables.

For those who measure the greatness of a year by the assumptions and isms that have perished within its span, 1981 will be judged one of the great ones in theoretical macroeconomics. It is the year, I shall argue, of the demise of monetarism; if not the demise, then at least its demotion to a temporary position pending a more promising replacement. But monetarism was not the only fatality of 1981. It is also the year in which the efficacy of fiscal policy was shaken to its foundations. It is conceivable that the idea of stimulating (or contracting) employment by means of fiscal policy measures may yet be reborn in some sturdier theoretical frame. But, for now, the old Keynesian notion of fiscal stimulus is so beset by doubts that fiscal policy, if not truly incapacitated, is in a deactivated status. Thus theoretical macroeconomics is today in a state of crisis. The hope, of course, is that the crisis signals a transition, in a direction not yet foreseen, to some Eriksonian stage of new competence and restored confidence.