Various aspects of the school enrollment-labor supply decision have been examined in earlier studies, including those of Bowen and Finegan (1969), Cohen, Rea, and Lerman (1970), Duncan (1965), Katz (1973), Korbel (1966) and Lerman (1972). Although these authors were aware of the joint nature of the school-work decision, typically, they used an ordinary least squares (OLS) framework with dichotomous dependent variables to analyze enrollment conditional on the labor supply decision or labor supply conditional on enrollment. Several more recent studies, such as Antos and Mellow (1978), Stephenson (1978), and Ehrenberg and Marcus (forthcoming) use multinomial logit to analyze the joint determination of labor supply and enrollment. Mallar (1976) uses a bivariate probit model to deal with the simultaneous relation between school and work. But none of these studies considers both the wage offers and job availability. There is still available no single set of estimates which considers the roles of both job availability and the wage on the joint enrollment-labor supply decision. Consider what the interrelation between wages and job availability implies for estimation of the labor force participation-school enrollment decision. To the extent that downwardly rigid wages prevent the youth labor market from clearing, one would expect both wages and job availability to influence labor supply. If one of these measures is not included in explaining labor supply, the coefficient estimated for the other measure may, for reasons well known, be biased. For example, if the wage is omitted but the unemployment variable included, and if wages and unemployment are negatively correlated, then the magnitude of the coefficient of the unemployment variable will be increased. It is also unclear on a priori grounds how job availability affects school enrollment. On the one hand, if jobs are readily available to young people, youth from poor families may be able to enroll in school, supporting themselves through part-time work. With no jobs available, some may be unable to afford school expenditures and may drop out (Bowen and Finegan, 1969, p. 404). On the other hand, readily available employment opportunities for youth may simply raise the probability of dropping out of school and working full-time. Our findings will help to determine how young people react to this influence and to other influences of the market. From the viewpoint of public policy, there are a number of reasons why it is important that the enrollment-labor supply decisions be understood, and that the separate responses to wage and job availability be isolated. For one thing, there is a tendency to discount the welfare importance of unemployment or poor labor participation for young people enrolled in school (Feldstein and Ellwood, 1979). However, the joint nature of the enrollment-labor supply decisions suggests that a high enrollment rate may not be simply a cause of low participation rates, but may be an additional symptom of adverse labor market conditions. Indeed, our results for nonwhite males support an interpretation of this kind. Second, the joint nature of the decisions also implies that public programs such as the Received for publication January 31, 1980. Revision accepted for publication December 11, 1980. * Both authors are with Dartmouth College and the National Bureau of Economic Research. The research reported here was supported by the Division of Technical Systems in the Office of Technical and Analytic Systems, Office of the Assistant Secretary for Planning and Budget, U.S. Department of Education and is part of the NBER's research program in Labor Studies. Any opinions expressed are those of the authors and not those of the National Bureau of Economic Research or of the Department of Education. We would like to thank Gary Fields, Meir Kohn, Robert Plotnick, Martin Segal and participants in the labor seminars at Harvard and at Cornell for their helpful comments.

IN I958, the O.E.E.C. in Paris published a survey of the gross national products (I956) of their member countries. This survey purported to demonstrate that the total production per person in the Netherlands was the second lowest in Western Europe (see Table I, col. 2). Because this finding is in contradiction with the ideas of productivity in the Netherlands, Dutch economists and statisticians have applied themselves to explain this rather bad relative position. Their studies raised several important questions concerning method and principle in the study of comparative productivity. The first difficulty is presented by the case of exchange rates in calculating gross national product per head. Official rates of exchange are an unsatisfactory basis for calculation, because their use neglects factors of internal purchasing power and internal alteration in price structure. Economists in the O.E.E.C. have themselves dealt with this problem2 and provided the material necessary to correct the official exchange rate. After having corrected the figures in this way,3 excessive differences disappear, at least so far as the European countries are concerned (see Table i, column 3) . A second difficulty to be overcome is connected with the divisor employed to obtain productivity per head. Of course the total population, as used by the O.E.E.C. economists, cannot be maintained. If, for instance, account is taken of social structure to obtain an estimate of the economically active population, quite different results emerge. So, the population of the Netherlands has increased more rapidly than that of the other O.E.E.C. countries. Furthermore, this country also has the highest average age. If these differences are reckoned with, the Dutch position has improved. Still, it remains second from the bottom in European countries (see Table i, column 4).

This paper analyzes the pricing behavior of electricity generating firms in the restructured California market from its inception in April 1998 until its collapse in late 2000. Using detailed firm-level data, I find that conduct is fairly consistent with a Cournot pricing game for much of the sample. In summer and fall 2000, the market was slightly less competitive, yet the dramatic rise in prices was more driven by changes in costs and demand than by changes in firm conduct. The five large nonutility generators raised prices slightly above unilateral market-power levels in 2000, but fell far short of colluding on the joint monopoly price.

This paper analyzes how community heterogeneity influences resident decisions to undertake actions generating public benefits. The decision in question is completing and returning the 2000 Census questionnaire, an action which secures a significant amount of federal grants for the community. The model developed to explain this action allows members of societal groups to differentially value public benefits that accrue to other group members. Racial, generational, and socioeconomic class heterogeneity all predict significantly lower response rates at the county level. The potential for endogenous sorting into heterogeneous counties implies that the magnitude of true behavioral effects exceeds these estimates.

Studies attempting to link locational attributes and individual outcomes often focus on children or young adults, under the presumption that their location was exogenously determined by their parents. This strategy is more difficult to justify if parents migrate selectively and tend to transmit their own characteristics to their children. This paper uses Census microdata to document a strong link between selective migration in one generation and economic outcomes in the next. I show that selective migration is a possible explanation for a puzzle in the existing literature: the changing relationship between segregation levels and individual outcomes within the black population.

Instrumental variable tests for serial correlation can be carried out by adding lagged residuals from initial estimation to the regressors of the model under scrutiny and then checking their joint significance. It is shown that asymptotically valid tests are obtained if the lagged residuals are also added to the initial instrument set. Monte Carlo evidence suggests that useful improvements in finite sample behavior under null and alternative hypotheses can be produced when the instrument set is extended to include the relevant lagged residuals. Links with other tests are discussed and a modification allowing for conditional heteroskedasticity is described.

In the absence of futures markets, cobweb cycles and other behavior inconsistent with Muth rational expectations persist for long periods of time. When futures markets are introduced in commodities, these markets behave in a manner much more consistent with Muth rational expectations. By contrast, despite the existence of active forward and futures markets, the Muth rational expectations hypothesis is rejected in the financial and foreign exchange markets. The aim of this paper is to suggest an explanation of how futures markets change the structure of the supply response.

Using a new and unique data set, this paper applies a systems approach to the study of potential competition, prices, and entry relations in airline city-pair markets. Consistent with limit pricing models, future entry is directly influenced by current prices. Current prices thus appear to provide an important signal to potential entrants about the probability of profitable entry. While results indicate the existence of barriers to entry, these barriers appear to have no independent effects on price beyond there effect on actual competition through increases in concentration.

The recent debate questioning whether unemployment in the 1970s represents sectoral adjustment or cyclical variation is expanded to examine comparable causes of real wage variability. Using Panel Study of Income Dynamics panel data, real wages respond more to persistent sectoral shocks than cyclical shocks in the 1970s, making recent estimates of procyclical wage variability appear weak in perspective. Employing a model of endogenous sector-specific individual skills, older workers earning economic rents are shown to have the greatest wage response to sectoral shocks. These results are consistent with the hypothesis that short run cyclical shocks may be met with hours adjustment, as specified in implicit or explicit contracts, but that persistent shocks require wage adjustment.