Knowledge that Transforms

Morton Paglin's recent article in this Review is an important contribution to the analysis of the distribution of income. He argues convincingly that inequality of incomes on a life cycle basis would exist even in a perfectly equalitarian society. He then provides a clear and simple technique by which inequality in excess of that related to the age-income profile can be measured. Paglin finds that on this basis the degree of inequality in the distribution of income fell significantly over the quarter century from 1947 to 1972; that is, the difference between the actual distribution and that which would have obtained if each family received the mean income for its age group decreased. Paglin's conceptual approach is a meaningful improvement over consideration of Gini coefficients without reference to underlying population change. At the same time, the Paglin technique raises further questions which shadow his conclusions on the trend of inequality. The objective of this comment is to discuss some of these questions and suggest alternative applications of the technique; the results will indicate that Paglin's technique must be used with caution for conclusions regarding the trend of inequality to be fairly drawn. The remainder of the comment will be divided into three sections. The first will consider Paglin's selection of the age-income profile as a measure of permissible income inequality. The second will examine the distribution of total family income as an indicator of equality. The third section will be a brief conclusion.

Availability is an important attribute of a good in many markets. Such markets include retail stores, restaurants, hotels, manufacturing, taxi cabs, airlines, parks and public utilities. Some of these markets are competitive, some noncompetitive, and some regulated. Fluctuating delivery time can be thought of as the consumers' equivalent to varying availability of a good. Here too, customers face some risk of being unable to obtain goods when they want them. There are good reasons why some markets do not always clear in the classical supply and demand sense at each instant. The three features that characterize these markets are temporary price inflexibility, demand uncertainty, and production lags. Prices do not instantaneously adjust in response to shifts in demand. If demand is especially heavy during the moming, prices do not rise in the aftemoon. Several justifications for such temporary price inflexibility are possible. Consumers may dislike price fluctuations, and firms may be providing a service of stabilizing prices in the very short run. Changing price may be costly. To provide an effective signal, prices may have to remain fixed for some time period. Unless it is evident that demand and supply have permanently shifted, firms may be reluctant to change price. Whatever the reason, it is a fact that for many markets, price once set does not vary for some time. Of course, there still remains the issue of how the price is initially determined. The second feature of these markets is that demand is uncertain. If demand were perfectly predictable, there would be no need to have unsatisfied customers. The final feature is that production takes time. If instantaneous production were possible, once again there need be no unsatisfied customers. We assume that recontracting or insurance markets do not develop. Such markets rarely develop in reality presumably because of high transaction and monitoring costs. This paper discusses the implications of markets characterized by price inflexibility, demand uncertainty over the time period for which prices are inflexible, and noninstantaneous production. A competitive equilibrium is defined and its properties examined. The social welfare implications of these markets and the socially optimal policy and its relation to regulation are analyzed. This social welfare problem is exactly the same as a peak load pricing problem under uncertainty. We next deal with the behavior of a monopolist, who tends to oversupply availability, but whose behavior is consistent with a smoothly functioning economy. Finally, the issue of firm interaction and incentives for vertical integration is addressed.

The imposition of price controls in the U.S. economy during the period 1971-73, together with the persistent inflationary bias in the decade of the 1970's, has brought forth a renewed research interest in the process of price determination.' The longstanding debate regarding whether firms tend to use full-cost or target return pricing rules on the one hand, or marginalist principles on the other has been intermingled with questions regarding the efficacy of price controls. In addition, the administered-price inflation hypothesis has been subjected to renewed empirical testing. Each of these areas of inquiry is of current policy relevance to the general problem of inflation. Although this paper will touch on each of the above areas, its principal objectives are narrow: 1. to specify a cross-sectional price equation which is consistent with both target return and marginalist principles; 2. to estimate the parameters of the price equation using data on year to year rates of change during the period 1958-72, for a large sample of four-digit SIC manufacturing industries; 3. to test hypotheses: a) regarding the relative effects of cost changes, demand changes, degree of concentration and price expectations on industry prices; and b) regarding the effects of the Phase I and II price controls of 1971-72. Many studies have examined the issues discussed in this paper. A discussion of selected studies which have treated some or all of the issues listed above will illustrate their current status. Most researchers agree that cost changes show up more strongly in price equations than do demand changes, although that result could be because the latter seem to be more difficult to measure. Researchers also agree that there are differences in the price adjustment process among concentrated and nonconcentrated industries, but the exact nature of these differences is controversial. The issue of rule of thumb versus marginalist pricing practices has also not been settled in the empirical literature since most price equations are consistent with either practice. In their study of the price equation, Eckstein and Fromm test the relative significance of target return and competitive elements in the determination of price changes. The equations which they estimate cover all manufacturing, durable manufacturing, and nondurable manufacturing based on two-digit SIC sectors. They utilize a time-series approach with data mainly for the period 1954:1 to 1965:4. Cost and demand elements are included in the equations. The authors make the general conclusion that: While the different forms of the equations yield varying results on the relative importance of the competitive mechanism vis-a-vis oligopolistic pricing, there is pretty strong evidence that equations combining both mechanisms are superior to equations using either approach in isolation (p. 1171). In a cross-section study of 395 four-digit SIC manufacturing industries, Frank Ripley and Lydia Segal estimate a price change equation for the period 1959-69. Changes in unit labor cost, changes in the materials cost, changes in real output, and changes in productivity are independent variables, with industries classified by broad concentration category. They find that over the elevenyear period covered by their study concen*Department of economics, University of South Carolina. i Rather than cite each item in the extensive literature, the following articles and their bibliographies provide broad coverage: William Nordhaus, Otto Eckstein and Gary Fromm, and Eckstein and David Wyss on price equations; Robert Lanzillotti, M. Hamilton and Blaine Roberts on Phase 11 controls; J. Fred Weston on alternative models of pricing behavior; and Steven Lustgarten and Ralph Beals on tests of the administered pricing inflation hypothesis. A more general survey on inflation has recently been provided by David Laidler and Michael Parkin.