Coronavirus (COVID-19): Relief options and Additional Resources

A Longitudinal Analysis of Early Self-Employment in the NLSYs

August 2010
No. 367
A Longitudinal Analysis of Early Self-employment in the NLSYs
Yasuyo Abe and Hannah Betesh, Berkeley Policy Associates and A. Rupa Datta, NORC at University of Chicago 2010 [69 pages]. Under Contract SBAHQ-08-M-0470

While the existing literature on self-employment offers a wealth of information on the characteristics of self-employed workers at a single point in time, to date few studies have taken workers’ patterns of self-employment as their unit of analysis. Few studies describe how involvement in self-employment is changing for the new generation of workers. The purpose of this research is to provide policy-relevant analysis of the characteristics and career paths of those Americans who have chosen self-employ-ment. Specifically, this study will (a) provide new empirical findings regarding the dynamics of self-employment that underpin individual entrepreneur-ship during early adult work life; and (b) document generational changes in self-employment patterns in early adult work life between two cohorts born in the second half of the 20th century.To address these research issues, this study uti-lizes two National Longitudinal Surveys of Youth, the 1979 Cohort (NLSY79) and the 1997 Cohort (NLSY97), which offer extensive information on economic activity, as well as data on personal and family backgrounds, and allow detailed longitudinal investigation of self-employment activities.

Overall Findings

The key finding is that early exposure to self-employment increases individuals’ engagement in self-employment during the early- and mid-career years. There is a strong positive link between an indicator of self-employment during ages 20-22 and the self-employment outcome measures in ages 22-41. These outcome measures include total years of self-employment and the yearly self-employment probability.


  • The younger NLSY97 cohort has much higher self-employment rates than the older NLSY79 cohort had when the two are compared by age 23. This increase is driven by increases in Black and Hispanic self-employment, and to a lesser extent by female self-employment.

  • Demographic characteristics differ considerably between respondents who reported self-employment and those who did not. Respondents who have ever been self-employed are mostly male and White, have parents with higher educational attainment, and have grown up in a rural area or on a farm.

  • The availability of financial resources—family income and net worth—increases the probability of self-employment, although the size of the estimated effect is very small.

  • Total self-employment years are positively cor-related with economic outcomes, measured in terms of family income, the individual’s own income, and family net worth. An additional year of self-employ-ment increases the level of income and net worth significantly.

  • There are notable differences and similarities in the self-employment experiences in the early stage of work life between individuals born in 1960-1962 and those born 20 years later in 1980-1982.

  • Demographic characteristics such as gender, race, and ethnicity have different effects on self-employ-ment compared with employer-based employment.

Scope and Methodology

The research uses a series of univariate descriptive analyses to compare personal and family character-istics across individuals with varying levels of early self-employment. Then several different multivariate analytic methods are used to investigate the determinants of self-employment trajectories and to assess the economic outcomes attributable to these self-employment measures. Tobit regression analysis is used to estimate the effects of various background characteristics on self-employment (measured as total years). The generalized estimating equation (GEE) method is used to model the decision to engage in self-employment in any given year. This model also estimates the effects of various background characteristics on self-employment.

For comparison between the NLSY97 and the NLSY79, the authors use three birth cohorts of data from each survey: the 1960-1962 birth years from the NLSY79 and the 1980-1982 birth years from the NLSY97. The NLSY97 analysis sample includes 5,486 respondents. A comparison group from NLSY79 consists of 4,125 respondents.

Critical to the study is the preparation of a yearly employment status indicator, which defines respondents as “exclusively self-employed,” “dual-employed,” “exclusively employer-employed,” and “nonemployed” (not working and either in school or not in school). In addition, the analysis compares those who have been self-employed at any time with those who have never been self-employed.

Besides the employment status indicator, the study utilizes the NLSY’s extensive information about the respondents to create demographic profiles of the sample, which include country of birth, country of residence at age 14, South/non-South regional residence at age 14, household structure at age 14, highest grade completed by mother and father by 1979, race/ethnicity, gender, and a variable identifying persons born outside the United States with U.S. parentage.

This report was peer-reviewed consistent with Advocacy’s data quality guidelines. More information on this process can be obtained by contacting the director of economic research by email at or by phone at (202) 205-6533.

Ordering Information

Copies are available for purchase from:

National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
(800) 553-6847 or (703) 605-6000
TDD: (703) 487-4639

Order number: PB2010-112924
Paper A05 ($48.00)
Microfiche A05 ($33.00)
CD-ROM A00 ($30.00)
Download A00 ($15.00)

For email delivery of Advocacy’s newsletter, press, regulatory news, and research, visit

To read this full report, please view the attachment below.

Sunday, August 1, 2010