Self-employment, entrepreneurship, and the NLSY79: researchers have used the rich data from the 1979 cohort of the National Longitudinal Survey of Youth to investigate the relationship between self-employment and various job and earnings outcomes; future inquiry may afford valuable insights into other interesting consequences of self-employment

Monthly Labor Review, Feb, 2005 by Robert W. Fairlie

The 1980 wave of the NLSY contains a special section on participation in illegal activities, including questions on selling marijuana and other "hard" drugs. The answers to these questions, together with data from subsequent years of the survey, are used to examine the relationship between drug dealing as a youth and legitimate self-employment in later years. Using various definitions of drug dealing and specifications of the econometric model, the survey finds drug dealers to be 11 percent to 21 percent more likely than those who are not drug dealers to choose self-employment, all else being equal. After ruling out a few alternative explanations, this article interprets these results as providing indirect evidence that aversion to risk, entrepreneurial ability, and preferences for autonomy are important determinants of self-employment.

In addition to offering detailed information on criminal activities in the 1980 wave, the NLSY79 includes information on whether respondents were interviewed in jail or prison in each year. This information is useful because convictions and incarcerations may have different effects on current and future wage and salary and self-employment earnings. In particular, ex-offenders who choose self-employment do not face discrimination, either pure or statistical, by employers in the labor market, but may face other forms of discrimination, such as that by consumers or lending institutions. Using the NLSY79, the aforementioned study by Robert Fairlie provides evidence on the relationship between incarceration and self-employment. (5) Estimates from probit regressions indicate that having a previous incarceration increases the probability of self-employment by 0.36 percentage point to 0.39 percentage point, or 5.2 percent to 5.9 percent. Thus, self-employment may provide an important alternative to wage and salary work for at least some ex-convicts.

Another finding reported in Fairlie is that AFQT scores have a small and insignificant effect in probit regressions for the probability of self-employment. (6)

Interestingly, previous research using the NLSY79 finds that AFQT scores have a large positive effect on earnings. The general argument is that the scores represent a measure of basic skills that help predict job performance. Although youths who have low levels of these basic skills may have limited opportunities in the wage and salary sector, that barrier does not translate into higher probabilities of self-employment.

The longitudinal nature of the NLSY79 also allows one to explore the effects of previous labor market experience on current self-employment. Ellen Rissman analyzes one aspect of the dynamic relationship between unemployment and self-employment among men. (7) She finds that the probability of being self-employed in the current year increases significantly if the person was unemployed in the previous year. Stratifying her sample by race, she also finds a positive and significant effect for whites, but not for nonwhites.