Kevin Lang

Inference with Difference-in-Differences and Other Panel Data

We examine inference in panel data when the number of groups is small, as is typically the case for difference-in-differences estimation and when some variables are fixed within groups. In this case, standard asymptotics based on the number of groups going to infinity provide a poor approximation to the finite sample distribution. We show that in some cases the t-statistic is distributed as t and propose simple two-step estimators for these cases. We apply our analysis to two well-known papers. We confirm our theoretical analysis with Monte Carlo simulations.

Does School Integration Generate Peer Effects? Evidence from Boston's Metco Program

Most integration programs transfer students between schools within districts. In this paper, we study Metco, a long-running desegregation program that sends mostly Black students out of the Boston public school district to attend schools in more affluent suburban districts. Metco increases the number of Black students in receiving districts dramatically. Because Metco students have substantially lower test scores than local students, this inflow generates a significant decline in scores, with an especially marked effect on the lower quantiles. This paper investigates the impact of Metco on receiving districts. Aggregate data on schools from districts throughout Massachusetts and micro data from a single large district strongly suggest the impact of Metco is largely a composition effect, since OLS estimates show no impact on average scores in samples of White or non-Metco students. On the other hand, OLS estimates using micro data show some evidence of an effect on the scores of minority 3rd graders in Reading and Language. Instrumental variables estimates for 3rd graders are imprecise but generally in line with OLS. Further analysis shows the negative effects on 3rd graders to be clearly present only for girls. Given the highly localized nature of these results, we conclude that any peer effects from Metco are modest and short-lived.

Unemployment and the Structure of Labor Markets

In this book, 16 leading economists establish the connection between the forces of the labour market and the persistent but ever-changing problem of unemployment. These scholars examine structural unemployment in terms of efficiency wage and dual labour market models, as well as neo-classical theory. The effects of the decline of the manufacturing sector, the rise in services, and the upheaval in the oil industry on our understanding of cyclical and frictional unemployment are incorporated into the analysis. An examination of the relationship between structural shifts and unemployment considers the frequency and duration of unemployment of displaced workers, the effects of migration caused by structural shifts, and the phenomenon of unemployment within particular industries. Finally, the book presents intriguing thoughts on the conduct of labour research and experimental policy analysis. The contributors include: David M. Lilien, Alan B. Krueger, Lawrence H. Summers, William T. Dickens, Lawrence F. Katz, Kevin M. Murphy, Robert H. Topel, Mark Dynarski, Steven M. Sheffrin, Danny Steinberg, Frank A. Monforte, James W. Albrecht, John Haltiwanger and Mark Plant.

Are All Economic Hypotheses False

We develop an estimator that allows us to calculate an upper bound to the fraction of unrejected null hypotheses tested in economics journal articles that are in fact true. Our point estimate is that none of the unrejected nulls in our sample is true. We reject the hypothesis that more than one-third are true. We consider three explanations for this finding: that all null hypotheses are mere approximations, that data-mining biases reported standard errors downward, and that journals tend to publish papers that fail to reject their null hypotheses only when the null hypotheses are likely to be false. While all these explanations are important, the last seems best able to explain our findings.

Human Capital Versus Sorting: The Effects of Compulsory Attendance Laws

Under the educational sorting hypothesis a state compulsory school attendance law will increase the educational attainment of high-ability workers who are not directly affected by the law. Under the human capital hypothesis such laws affect only those individuals whose behavior is directly constrained. We find that compulsory attendance laws do increase enrollment rates in age groups they do not affect directly. Thus, our results contradict the human capital hypothesis and are consistent with the sorting hypothesis.

Labor Market Segmentation Theory: Reconsidering the Evidence

We argue that Labor Market Segmentation theory is a good alternative to standard views of the labor market. Since it is sometimes argued that labor market segmentation theory is untestable, we first consider the uses of theory and the attributes of a good theory. We then argue that labor market segmentation has these attributes. It is internally consistent and is based on plausible assumptions about behavior and technology. More significantly, many of the predictions of the theory have been tested and confirmed. Further, from a dynamic view the theory has done quite well. When the theory has suggested new tests, far more often than not the predictions have been validated. Labor market segmentation theory has had to make little recourse to post-hoc explanations for unexpected empirical results. In contrast, human capital theory has required a series of post-hoc rationalizations to explain a large and growing body of empirical work motivated by the labor market segmentation perspective. Finally, we consider the implications of labor market segmentation theory for the practice of labor economics. We argue that further exploration of the implications of the theory for unemployment, trade. industrial policy and income distribution will provide useful insights and further tests of the theory.

Racial Discrimination in Labor Markets with Posted Wage Offers

We analyze race discrimination in labor markets in which wage offers are posted. If employers with job vacancies receive multiple applicants, they choose the most qualified but may choose arbitrarily among equally qualified applicants. In the model, firms post wages, workers choose where to apply, and firms decide which workers to hire. Labor-market frictions greatly amplify racial disparities, so mild discriminatory tastes or small productivity differences can produce large wage differentials between the races. Compared with the nondiscriminatory equilibrium, the discriminatory equilibrium features lower net output, lower wages for both white and black workers and greater profits for firms.

The effect of minimum-wage laws on the distribution of employment: theory and evidence

Abstract Recent research casts doubt on the view that minimum-wage laws reduce employment. We show that in a simple model of bilateral search with heterogeneous workers, a minimum-wage law increases employment. However, the increased competition from higher productivity workers makes lower productivity workers worse off without making higher productivity workers better off. We provide evidence that minimum-wage laws shift employment from adults to teenagers and students. This raises concerns about the distributional consequences of minimum wages even when they increase employment.

The Intake Dispositions of Juvenile Offenders

This article presents the analysis of intake dispositions for a sample of 533 juvenile suspects. By using multinomial logit analysis, we found strong statistical confirmation of the relevance of prior record and demeanor of youth during intake for the selection of intake options by officers in Los Angeles County. The analysis indicates that when other factors are held constant, there is no consistent pattern of differential treatment between black and Mexican-American youth. However, whites are less likely to receive counsel and release, the most lenient option, and a detain petition, the most severe option.

Relative Wages, Wage Growth, and Quit Behavior

Using Italian Social Security records for male workers from a sample of firms in Turin from 1981 to 1983, we show that conditional on the workers own wage the average wage in the establishment for similar workers is negatively related to quits. We also find that this variable predicts future wage growth. This is consistent with an economic model in which workers compare the longrun value of employment opportunities when making quit decisions.