In 2000, James Heckman, along with daniel mcfadden, received the Nobel Prize in economics. Heckman won the prize for “his development of theory and methods for analyzing selective samples,” highly technical work that it is difficult to explain to the layman. Nevertheless, the work rewarded by the Nobel committee has been valuable for economists’ studies of many issues that laymen do care about. The main technical problem on which Heckman has spent much of his professional life involves self-selection. An economist who wants to know, for example, how male workers will respond to a higher wage rate can take microdata on wages and hours worked and find a relationship. This approach has a problem that economists have long recognized: some men will not work at all and, therefore, will not be in the data set. And, presumably, these men who do not work will be disproportionately from the group that, had they worked, would have earned low wages. They have “self-selected” out of the workforce. So the economist’s estimates on the effect of wages on hours worked will be biased. How is the economist to deal with this fact if he wants to generalize from his sample to the male population in general?

Enter Heckman. Heckman came up with a clever econometric approach to figuring out how to correct for this self-selection problem. In fact, empirical economists now know his correction as the “Heckman correction” or the “Heckit method.” In 1985, using his own method to study the connection between work and wages noted above, Heckman found a bigger elasticity of labor supply among American men than had previously been thought to exist. That was because a given increase in wages did not just cause an increase in hours among those already working, but also caused relatively low-wage workers to reenter the labor market and get jobs.

In his Nobel lecture, Heckman laid out some important implications of self-selection for U.S. and European labor markets. One is that the vaunted diminution of the gap between black men’s and white men’s wages in the United States between 1940 and 1980 was due almost entirely to low-wage black men dropping out of the labor force. Another implication concerns the much-higher equality of wages in Europe compared with the United States that many economists and others have noted. Heckman pointed out that much of this difference is due to the fact that many low-skilled potential workers in Europe are not working. This is presumably because of high minimum wages, strong labor unions, and laws that make it hard for employers to lay people off (and therefore make the employers hesitant to hire).

The Nobel Web site refers to Heckman as “the world’s foremost researcher on econometric policy evaluation.” In 1999, Heckman and coauthors Robert LaLonde and Jeffrey Smith found that government programs to train workers are generally ineffective at increasing those workers’ wages and long-term employment prospects. The reason they appear effective is that the average wage gain of people in these programs is high. But Heckman and his coauthors showed that this occurs because the trainees typically have lost or quit their jobs prior to entering the training program and that training is a form (albeit inefficient) of job search. Much of the improvement in their posttraining fortunes would have occurred without a government program in place. This would probably come as no surprise to people who have been in such programs. In 2002, with Pedro Carneiro, Heckman also showed that lack of credit is not a major constraint on the ability of young Americans to attend college. They found that credit constraints prevent, at most, 4 percent of the U.S. population from attending.

Heckman has also weighed in on the issue of racial discrimination in U.S. labor markets. In 1998 he wrote, “[M]ost of the disparity in earnings between blacks and whites in the labor markets of the 1990s is due to the differences in skills they bring to the market, and not to discrimination within the labor market.”


Selected Works

 

1974. “Shadow Wages, Market Wages and Labor Supply.” Econometrica 42: 679–693.
1975. “The Common Structure of Statistical Models of Truncation, Sample Selection and Limited Dependent Variables and a Simple Estimator for Such Models.” Annals of Economic and Social Measurement 5: 475–492.
1976. “A Life Cycle Model of Earnings, Learning and Consumption.” Journal of Political Economy 84: S11–S44.
1979. “Sample Selection Bias as a Specification Error.” Econometrica 47: 153–161.
1981 (with Thomas MaCurdy). “New Methods for Estimating Labor Supply Functions.” In Ronald Ehrenberg, ed., Research in Labor Economics. Vol. 4. Greenwich, Conn.: JAI Press.
1998. “Detecting Discrimination.” Journal of Economic Perspectives 12, no. 2: 101–116.
1999 (with Robert LaLonde and Jeffrey Smith). “The Economics and Econometrics of Active Labor Market Programs.” In Orley Ashenfelter and David Card, eds., Handbook of Labor Economics. Amsterdam: North-Holland. Chap. 31.
2002 (with Pedro Carneiro). “The Evidence on Credit Constraints in Post-secondary Schooling.” Economic Journal 112: 705–734.