This review culls noneconomic literature on education--by sociologists, anthropologists, and practitioners to present a new economic theory of students and schools. This theory elaborates two themes that have eluded economic analysis. First is the student as decision-maker whose primary motivation is her identity. Second is a conception of the school as a social institution. This framework suggests a new perspective on questions such as resource allocation and school reform. It explains why some educational policies succeed and others fail. We show how sociological variables may affect outcomes, and suggest ways economists can incorporate them into theoretical and empirical research.

The purpose of this paper is to summarize the results of an earlier paper (1979), and two written in collaboration with Ross Milbourne (1980a, b) and, in so doing, to illuminate certain questions as well as answers which are implicit in the approach that has been taken. All inventory-theoretic models of the demand for money can be summarized by their assumptions about the autonomous payments into and out of bank accounts, and the way in which bank accounts are monitored. An autonomous payment is one that is independent, both in timing and amount, of the level of the bank account. From time to time holders of bank accounts monitor their accounts, at which point they either raise the level in their accounts, if deemed too low, or reduce the level in their accounts, if deemed too high. The payments made in response to this monitoring are called induced payments, in contrast to the autonomous payments. The most intuitive proposition in the theory of money, indeed, the proposition that historically spawned monetary theory as a special discipline, has been that an increase in transactions will cause a proportionate (or almost proportionate) increase in the demand for money as long as the ways in which bank accounts are monitored remain unchanged. And since changes in monitoring might well be considered a long-run, rather than a short-run, response to changes in endogenous economic variables, the demand for money will be almost proportional to the level of transactions. The preceding is a rough description of Irving Fisher's monetary theory (1911), and the proposition has been considered so obvious as to occasion virtually no comment anywhere-not even by John Maynard Keynes, whose transactions demand for money is proportional to income, which is his measure of transactions (1936, p. 201). Elementary textbooks often show the demand for money as proportional to transactions (or income) by a simple diagram, wherein an individual receives Y dollars of income at the beginning of the payments period and spends this income at a constant rate over the payments period, ending it with no cash balances. In this case the demand for money is Y/2. This diagrammatic exposition seems quite convincing, and for this reason, as well as the convincing verbal arguments of Fisher and earlier quantity theorists, it is expected that income should enter the demand for money, especially in the short run, as a strong argument. The dependence of the demand for money on income does not automatically make the LM curve vertical, however, because the demand for money may also depend on the rate of interest. As interest rates rise, it pays to monitor bank accounts more frequently and, as a result, to hold less money since the opportunity cost of money holdings is the interest lost from not holding alternative interest-bearing assets. Thus, according to William Baumol (1952) and James Tobin (1956), the demand for money depends both on income and interest. But calculations of the gains to the average money holder from adjusting his monitoring policies to new optima, following changes in income and interest, show these gains to be quite trivial, especially prior to the current recent period of double-digit interest rates. The smallness of these gains and the inability of any speculator to aggregate the gains of the many individual money holders, suggest that the adjustment of the monitoring of *University of California-Berkeley. I would like to thank Janet Yellen for invaluable help and criticism. I also thank the National Science Foundation for generous financial support under Research Grant no. 79-05562 administered by the Institute of Business and Econonmic Research of the University of California-Berkeley.

*This paper is based on a long-term research program with Rachel Kranton on the implications of identity for economic behavior and also a manuscript we are currently writing on the missing motivation in economics. Our previous joint papers (Akerlof and Kranton (2000), (2002) and (2005)) have explored implications outside of macroeconomics of utility functions dependent on peoples notions of what ought to be. Conversations with Kranton have also been the basis for the section on economic methodology. I have also benefitted from conversations with Robert Shiller, with whom I am co-authoring work on behavioral macroeconomics. In addition, I

Think about Richard Scarry’s Cars and Trucks and Things That Go. Think about what that book would have looked like in sequential decades of the last century had Richard Scarry been alive in each of them to delight and amuse children and parents. Each subsequent decade has seen the development of ever more specialized vehicles. We started with the Model T Ford. We now have more models of backhoe loaders than even the most precocious fouryear-old can identify. What relevance does this have for economics? In the late 1960’s there was a shift in the job description of economic theorists. Prior to that time microeconomic theory was mainly concerned with analyzing the purely competitive, general-equilibrium model based upon profit maximization by firms and utility maximization by consumers. The macroeconomics of the day, the so-called neoclassical synthesis, appended a fixed money wage to such a generalequilibrium system. “Sticky money wages” explained departures from full employment and business-cycle fluctuations. Since that time, both microand macroeconomics have developed a Scarry-ful book of models designed to incorporate into economic theory a whole variety of realistic behaviors. For example, “The Market for ‘Lemons’ ” explored how markets with asymmetric information operate. Buyers and sellers commonly possess different, not identical, information. My paper examined the pathologies that may develop under these more realistic conditions. For me, the study of asymmetric information was a very first step toward the realization of a dream. That dream was the development of a behavioral macroeconomics in the original spirit of John Maynard Keynes’ General Theory (1936). Macroeconomics would then no longer suffer from the “ad hockery” of the neoclassical synthesis, which had overridden the emphasis in The General Theory on the role of psychological and sociological factors, such as cognitive bias, reciprocity, fairness, herding, and social status. My dream was to strengthen macroeconomic theory by incorporating assumptions honed to the observation of such behavior. A team of people has participated in the realization of this dream. Kurt Vonnegut would call this team a kerass, “a group of people who are unknowingly working together toward some common goal fostered by a larger cosmic influence.” In this lecture I shall describe some of the behavioral models developed by this kerass to provide plausible explanations for macroeconomic phenomena which are central to Keynesian economics. For the sake of background, let me take you back a bit in time to review some history of macroeconomic thought. In the late 1960’s the New Classical economists saw the same weaknesses in the microfoundations of macroeconomics that have motivated me. They hated its lack of rigor. And they sacked it. They then held a celebratory bonfire, with an article entitled “After Keynesian Macroeconomics.” The new version of macroeconomics that they produced became standard in the 1970’s. Following its neoclassical synthesis predecessor, New Classical macroeconomics was based on the competitive, general-equilibrium model. But it differed in being much more zealous in insisting that all decisions—consumption and labor supply by † This article is a revised version of the lecture George A. Akerlof delivered in Stockholm, Sweden, on December 8, 2001, when he received the Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel. The article is copyright © The Nobel Foundation 2001 and is published here with the permission of the Nobel Foundation.