Antecedents and predecessors of NLSY79: paving the course: a historical view of the NLSY79 development stages highlights lessons learned during an era filled with new concepts and innovations in sociology, economics, and computer science

Monthly Labor Review, Feb, 2005 by James R. Walker

In 1965, at the prompting of the Assistant Secretary of Labor, Daniel Patrick Moynihan, individuals from the Department of Labor (DOE) and Ohio State University designed the National Longitudinal Surveys of Labor Market Experience. At the time, the participants did not realize that they were creating one of the premier, large scale national longitudinal surveys in the United States. Initially funded for 5 years by the Department of Labor, the "Parnes" data, as the Original Cohorts were called, continued for 37 years, with the last scheduled fielding of the women samples in 2003. (1) The success of the Original Cohorts led to the creation of the National Longitudinal Survey of Youth (NLSY79).

This article explores antecedents and predecessors of the National Longitudinal Survey of Youth, 1979. (2) Longitudinal data are now so plentiful that it is difficult to imagine the world in which they did not exist. Yet, in the mid-1960s, the large scale longitudinal household surveys that came to dominate areas of sociology, demography, and labor economics did not exist. Analyses that are now commonplace were either not possible or inference was restricted to small or specialized samples.

Yet to suggest that there were no longitudinal data sources prior to 1965 is wrong; several longitudinal surveys predate the NLS. Two well-known studies reflect the nature of longitudinal data available before the start of the NLS. The Glueck study of juvenile delinquents from the Boston area followed 1,000 adolescents (500 juvenile delinquents and 500 non-delinquents) into adulthood to examine criminal behavior and contact with the justice system. (3) Sheldon and Eleanor Glueck started interviewing at the end of 1938, completing the first wave of interviews in 1948. Two more waves of interviews followed as the youth were interviewed at ages 25 and 32. Interviews continued until 1965.

The other study available before the NLS, and perhaps more visible to economists, is the National Bureau of Economic Research (NBER)-Thorndike sample, collected from Air Force volunteers during WWII In 1955, R. Thorndike and E. Hagen randomly selected 17,000 of the 75,000 Air Force volunteers who took the Aviation Cadet Qualifying Test in the second half of 1943 (a test similar in function to the Armed Services Vocational Aptitude Battery (ASVAB) tests that NLSY79 respondents took to set a norm in recruiting standards for the Department of Defense). In 1969, with funding from the NBER, Paul Taubman and his colleagues reinterviewed about 5,000 of the original 17,000 members of the Thorndike sample, obtaining information on current and retrospective earnings, education, and occupation. These data have been widely used to study the determinants of earnings, ability bias, and the return to schooling (that is, benefits associated with higher levels of schooling). (4)

A number of other specialized longitudinal studies were launched in the decade prior to the NLS. These efforts surveyed teen mothers, drug users, gifted children, and children from privileged and underprivileged backgrounds. (5) These studies shared features like the Gluecks' study and the NBER-Thorndike study in that they were local in character with limited or irregular longitudinal followups. However, several studies are impressive and cover a long arc of their respondents' lives. (6)

Antecedents

Scientific frontiers. Two critical elements came together in the 1960s supporting the development of large, household surveys. First, the social science field had developed the conceptual foundation supporting the use of longitudinal data. Within the fields of psychology and sociology, researchers and scientists fostered the life course perspective, viewing human development as following a sequence of stages. (7) And second, in the economics field, human capital became the organizing conceptual framework. In his 1960 Presidential Address to the American Economics Association, T.W. Schultz presented his influential thoughts on human capital. (8) The human capital theory quickly became a central concept for understanding the determinants of wages, the structure of earnings, and more generally, the distribution of economic opportunities. Labor economists sought to measure the return to schooling, labor market experience, and tenure with an employer. Social scientists sought to understand schooling decisions, both in terms of quantity (the amount of schooling obtained) and in terms of quality (types of post secondary schooling).

Intervention and experiments. The intellectual primacy of measuring education and training fueled and was fueled by the era of big social science and policy interventionism of the mid-1960s. In 1964, the Johnson Administration announced the War on Poverty. Education and training programs were among the most important anti-poverty programs proposed. Thus, measuring and understanding the determinants and consequences of poverty required the collection of new, longitudinal household level data. The Department of Labor, Office of Economic Opportunity initiated a survey of the same name in 1967, followed by the Panel Study of Income Dynamics (PSID), conducted by the Census Bureau in 1968.