Knowledge that Transforms

The theory of the firm is a monumental achievement of neoclassical economics. Without this engine of analysis, it would be difficult, if not impossible, to comprehend the pricing, output, and input decisions of firms as they respond to routine events like the imposition of a tax, the opening of a new market, a technical innovation, and a sudden shortage of a key factor of production. It is remarkable that this theory has been successful in explaining behavior that to a large extent is motivated by both profit and risk when the theory itself has only explicitly considered the profit motive. This is not the place to dwell on the evolution of economic theory; it suffices to note that there are many important economic phenomena that the purely deterministic theory does not explain.' The purpose of this note is to elucidate the way in which risk influences the output decisions of riskaverse entrepreneurs and the number of riskneutral firms in a competitive industry. In both cases, the predicted behavior differs from that of the deterministic theory. In our study we shall restrict attention to the behavior of firms in a single period 2 setting. The firm has no control over price and, because storage makes no sense, simply sells all of its output at the going price. The source of uncertainty is the requirement that the firm produce before price is known, where the price is a random variable with a known probability distribution. The firm chooses output to maximize its expected utility. It is well-known that, in the presence of uncertainty, the optimal output of the risk-averse firm is less than that of the riskneutral firm; moreover, increases in risk aversion, in the sense of Arrow and John Pratt, lead to further diminutions in output. On the other hand, for a fixed degree of risk aversion, the change in output induced by a mean-preserving increase in risk depends on the shape of the cost curve as well as the sign of the third derivative of the utility function u and the sign of the second derivative of qu'(q). Next, competitive industry behavior under uncertainty is analyzed. In order to isolate the effect of uncertainty, firms are assumed to be risk neutral. We show that both the optimal number of firms in the industry and excess capacity increase as industry output becomes riskier. Results like this are important for probabilistic economics, for it would be unfortunate if the vitality of the stochastic theory of the firm relied solely on the controversial assumption of risk aversion.3

The purpose of this paper is to provide a simple framework in which to analyze the impact of a perfectly anticipated increase in the future selling price of houses on current house purchase decisions in the presence of capital market imperfections. Since, implicitly, it is the after-tax selling price of houses that is considered, this analysis is equivalent to one of a decrease in the capital gains tax on houses at retirement. In this paper assumptions about homeownership, mortgage markets, and liquidity constraints are added to a life cycle model of asset accumulation, taking the date of house purchase as exogenous. It is shown that the impact of an exogenous increase in the expected future price of houses depends on current mortgage market conditions, the future price of houses relative to the current price, and the ratio of liquid assets to future labor income, as well as preferences. Also, a simple model of a market for houses is presented to analyze the impact of an increase in the expected future price of houses on the current market-clearing price, the mortgage market, and the redistribution of the housing stock to consumers according to their wealth and liquidity characteristics. In the life cycle model as exposited by Franco Modigliani and Robert Brumberg, an increase in consumption in one period requires a decrease in consumption in another period. When owner-occupied housing is added to the model the tradeoff between consumption in different periods becomes more complex. Purchasing a larger house entails an increase in the consumption of housing services and an increase in assets held in the form of a house, but a decrease in either nonhousing consumption or the holdings of an alternative asset, or both. The response of the consumer to price or income changes is also limited by financial market constraints. Capital market imperfections have been introduced into the life cycle model by a number of authors (see, for example, Lester Thurow, Thayer Watkins, Thomas Russell, Clark Wiseman, and Christopher Pissarides). And, several authors have explicitly considered the impact of capital market imperfections on the house purchase decision (see William Dolde and James Tobin, William Poole, Donald Lessard and Modigliani, and Dolde). Most similar to this study, although developed independently, Roland Artle and P. Varaiya's paper includes the theoretical incorporation of the house-purchase decision into a life cycle model. They consider the problem of choosing the date of house purchase while treating the level of consumption of housing services as exogenous. In the life cycle model in this paper, the consumer chooses the size of house to purchase at an exogenous date. To use the life cycle model in the framework of a market of houses, it is assumed that the stock of houses and the prices of all other goods, including the rate of return on the alternative asset, are fixed. All consumers are identical, except perhaps for initial wealth. As noted by Irving Fisher (p. 325), in a world of perfect foresight and perfect markets the rates of return on all assets are equated in equilibrium. If markets are perfect, with the stock of houses and other prices fixed, the present price of houses will rise by an amount equal to the present value of an exogenous increase in a future price of houses. When mortgage market imperfections are introduced, this relationship no longer holds, and a future price increase may result in a redistribution of the current stock *Assistant professor of economics, University of Michigan. This paper is an extension of a chapter of my doctoral thesis at the University of Wisconsin-Madison. The helpful comments of Donald D. Hester, Charles A. Wilson, Hal Varian, and George Borts are gratefully acknowledged.

Inflation has become, the 1970's, an important economic phenomenon that must be taken into account analyzing the determinants of labor supply trends. Traditional labor supply theory has focused on the overriding importance of real wage growth as a primary determinant of both the long-run secular decline hours of work and the postwar increase the labor force participation rate of married women. Within this context, any labor supply effects of price level changes have been subsumed under the overall effect of real wages. However, despite unusually rapid price increases as well as stagnation the growth of productivity and real wages the last decade, the growth women's labor force participation has continued and, if anything, accelerated the last decade. Thus it is clear that, today, factors other than real wage growth must lie behind this continuing upward trend. Inflation is well worth exploring as a possible independent influence on labor supply, because of both its growing visibility and the frequency with which one hears comments along the lines of in these inflationary times, a family needs two salaries to make ends meet. Our objective this paper is to explore the possible effects of inflation on women's labor supply trends. As a first step this direction, we present some empirical results, relating changes the labor force participation rates of women various age groups to inflation as well as other independent variables, covering the period 1956-77. Inflation clearly appears to have an effect on labor force participation rates above and beyond the effect it generates through reducing the real wage. By next examining the primary sources of women's labor force growth the last decade and their implications for long-run labor supply, we suggest the likely importance of inflationary expectations sustaining the long-term growth of women's labor supply, particularly the prime age group.

It is suggested that the ethical notion of social contract can be formally modeled using the well-studied concept of the core of a game. This provides a mathematical technique for studying social contracts and theories of justice. The idea is applied to Rawlsian justice here. (This abstract was borrowed from another version of this item.)

The literature on the determinants of good performance has been largely devoted to empirical measurement. A complete review of the publications in this area would require at least an entire issue the size of this journal. Yet, it is virtually impossible to discover attempts to derive hypotheses about the determinants of good from explicit models of behavior. This has led researchers into the trap of exploring only the empirical determinants of good performance. Typically, the results of student evaluations are correlated with affective personality traits of the teachers, with peer evaluations of psychological compositions, or even with the results of various psychological tests. Examples are the works of Robert Isaacson et al., Frank Costin et al., and Wilbert McKeachie et al. In most of these studies, no substantive conclusions are reached regarding the determinants of effective teaching. In the few studies which do contain significant findings, the independent variable in question is difficult to explain as a direct determinant of effectiveness. For example, Isaacson et al. find that virtually the only statistically significant determinant of effective is the personality characteristic which they label culture. This may imply that simply wearing a tweed jacket and taking up pipesmoking is sufficient to improve teaching.' This paper takes a somewhat different approach to the analysis of teaching. Assume that the individual in question wishes to maximize effectiveness. Clearly, there are constraints on this maximization process. One must necessarily be the time budget constraint. A second constraint, however, adds a good deal of interest to the model. This constraint is best described as the bundle of characteristics available to the individual. These characteristics could be termed the components of the individual's personality. Such attributes include physical appearance, psychological makeup, speaking ability, and a myriad of other variables. Thus, this model bears a resemblance to the standard economic models of monopolistic competition (see A. Michael Spence). Before proceeding to a formal statement of the problem, a philosophical issue must be addressed. What is meant by teaching effectiveness? Generally, this is interpreted as maximizing student evaluations of teachers. However, this is clearly not always equivalent to maximizing student learning. Unfortunately, the latter variable is difficult (but not impossible) to measure. (Preand posttesting with standardized tests is becoming increasingly accepted, for example). The meaning of effectiveness in this paper must, unfortunately, remain ambiguous. Some teachers (concerned, perhaps, with obtaining tenure) will interpret good as obtaining good evaluations, the criterion by which they are judged. Others may take a more eclectic viewpoint. Since the problem below is an individual maximization problem, the precise definition of effective may simply vary from individual to individual (see W. R. Allen). Formally, this problem is exactly analogous to one first addressed by Kelvin Lancaster. He assumes that the consumer attempts to maximize satisfaction through purchase of bundles of attributes. Each commodity is possessed of certain of these characteristics. By consuming various combinations of these commodities, the consumer is *Assistant professor, department of economics, and Director, Center for Economic Education, California State University-Hayward. I would like to thank Alex Cassuto, Nancy Sanders, and an anonymous referee for helpful discussion. 'In a previously unreported experiment in the department of economics, California State UniversityBakersfield, faculty members were required to wear coats and ties in classes. It was subsequently observed that evaluations improved. Unfortunately, no detailed statistical results of this experiment are available for analysis.