AL: Ken, you were involved in conducting the NLS before my time, and returned later for an encore. Can you give a quick rundown of your various roles in the survey?
KW: I became the PI of the NLS in 1983. As PI, I was responsible for overseeing all tasks, from questionnaire design to data dissemination, associated with the original cohorts and the NLSY79. Prior to my arrival, discussions had already begun with NICHD about starting a survey of the children of the women of the NLSY79. Those discussions continued after I arrived and culminated in the Child cohort, which began in 1986 and which, as is well recognized, became a major new and innovative research tool for the study of child development. The Child cohort was born through the efforts of many, but I would be remiss if I did not mention those who led the effort: Wendy Baldwin (NICHD), Bob Michael (NORC) and Frank Mott (CHRR).
The day-to-day operation of the NLS project was handled by an incredibly dedicated and talented group of people. The fact that I was able to maintain an active research agenda, one that quickly became (and continued for decades to be) dominated by NLS data, is a testament to that. Without in any way diminishing the contributions of others, and recognizing that my interactions with many was limited, I would like to express my personal gratitude to Laura Branden, Patricia Rhoton, Carol Sheets and Mary Wildermuth.
After leaving Ohio State (and my position as PI) in 1987, I served for a number of years on the NLS Technical Review Committee. In the mid-1990s, BLS decided to start a new youth survey. I was asked to coordinate the questionnaire development. I accepted the invitation without a second thought. In many ways, the NLSY97 questionnaire built upon the innovations in the NLSY79, expanding on a number of substantive areas. Mike Pergamit, the director of the NLS at BLS, deserves special mention for his hands-on role in the design of the questionnaire.
Looking back, I can say unequivocally that being the PI of the NLS was one of the highlights of my academic career. I am thankful for having been given the opportunity.
AL: I entered the picture in 1993, when I joined the NLS team at Ohio State. During my first decade, I worked on design for the Mature Women, Young Women, NLSY79 and NLSY97, and contributed to various usability and user outreach efforts. I became PI of the NLSY79 in 2005, which means I head our collaborative design efforts and perform a grab-bag of other duties. In contrast to your era, when you were the sole PI for all active NLS cohorts, we now have a different PI for each cohort.
I am delighted that you mention a few of the many people who contributed significantly to the NLS over the years. Rather than provide my own shout-outs, I'll note that I orchestrated this NLS@50 website as a means of acknowledging everyone who helped make the NLS program such a success: sponsors, staff, researchers and, last but not least, the amazing respondents.
AL: At the risk of bringing the celebration to a crashing halt, let's change the subject: On a funding scale of one to ten, where ten means “all the money in the world” and one means “close to shutting down the whole project,” what was the financial climate during your stint with the NLS in the 1980s?
KW: Two funding issues arose early in my tenure. First, the NLS project, which had been a sole-source contract for more than 15 years, was for the first time competitively bid. Second, the Reagan era of government budget cuts struck the NLS. At that time, the NLS was housed within the Employment and Training Administration of DOL. Their interest in the NLS had diminished over time and the NLS became a target of the budget cuts they needed to make.
With respect to the first of these, the staff put together a winning proposal (although we surmised that there were likely no other serious bidders). The contract, however, called for cutting out the research component, about one million dollars if I recall. There were additional smaller cuts as well.
With respect to the second, Celia Homans of NORC and I, with help from a number of organizations, were able to have language added to the appropriations bill to maintain NLS funding. The key ingredient was having an Ohio congressman on the appropriations committee; it is otherwise doubtful that we could have achieved that outcome. Senator Ted Kennedy also gave a colloquy on the Senate floor in support of the NLS. However, these were only short-term fixes. A longer term fix came when Janet Norwood, the Commissioner of the Bureau of Labor Statistics, decided to give the NLS a new home at BLS. That arrangement has served the NLS and the government well.
AL: The NLS program has benefited greatly from being housed at BLS, but unfortunately funding problems re-arose under BLS stewardship. The NLS faced funding crises in the late 1980s, the 1990s, and again in recent years when the program was in danger of being terminated as a result of the sequester. These near-death experiences are awful, but they sure do reveal how many friends the NLS has. The program has consistently been aided by members of Congress who recognize its importance, by high-placed government officials, and by members of the research community. A few years ago, a "Save the NLS" campaign generated thousands of emails from NLS users to DOL officials.
AL: Let’s turn to questionnaire content. You have contributed significantly to the design of the 79 and especially the 97 cohort. What do you view as one of the most important or perhaps innovative components of the questionnaires for these cohorts?
KW: The NLS has continually pushed the frontier in the design and content of survey data, starting with the original four cohorts and culminating (so far) in the NLSY97. An overarching principle in the development of the NLSY97 was to continue and extend the event history format pioneered in the NLSY79 to additional domains. For example, the schooling module included an accounting of all post-secondary schools attended, their spells (beginning and ending dates), credits earned at each institution and a host of additional information relevant to post-secondary schooling decisions. Other new features of the NLSY97 were the incorporation of dating behavior in the marriage module, a separate module on crime that could be tied into employment and schooling histories, and a module on subjective expectations.
The challenge in coordinating the design of the NLSY97, and my goal, was to provide a strong foundation on which others could build as the cohort ages and the instrument inevitably evolves. To the extent that that goal was achieved, and I believe it was, it was because of the efforts of a design team that consisted of a who’s who of social scientists. It was both hard work and great fun. I hope someone has the chance to do it again with another new cohort.
AL: Because you referred to aging, I'll note that in recent years I've worked with numerous top-notch social scientists to make adjustments to the NLSY79 questionnaire, given that the "youth" who belong to this cohort are now well into their 50s. We have added many questions on health, retirement expectations, estate planning and financial preparedness, and will presumably make many more changes in the next few years. I am as hopeful as you that a new NLS cohort might be added before long, but I am equally hopeful that we will be able to follow the NLSY79 respondents into retirement. To collect data on these individuals from their school days into retirement will be a remarkable accomplishment.
KW: What do you know about the usage of NLS data? I’m curious to know whether usage is increasing for every cohort, and what the trends look like by academic discipline.
AL: BLS and CHRR compile all sorts of interesting usage statistics, many of which are drawn from the NLS Annotated Bibliography. Using publications as the metric and confining attention to the last decade, we find that all three ongoing cohorts (NLSY79, NLSY97 and Child/Young Adult) show an increase in usage, while use of the four original cohorts has declined. Unsurprisingly, the growth rate is greatest for the NLSY97, which averaged only a few publications per year for the first eight years of its existence.
One contrast that really jumps out at me is that publications in economics journals have grown by only about 1% per year over the last decade, while publications in psychology journal have increased by over 10% per year. I would love to see every empirical economist use NLS data, but there are just too many good ways to get data these days, including administrative records and field and lab experiments. While there appears to be a loss of market share among economists, it’s exciting to see researchers in other disciplines flock to the NLS. In addition to the psychologists, we see increased usage in health, criminology, consumer finance, and demography. I suspect you will agree that one of the best things to happen to the NLS over its 50 year history is that it morphed from “The NLS of Work Experience” into a survey of “work experience and everything related to work experience.” The broader focus has increased the constituency and produced an amazingly diverse mountain of NLS-based research.
KW: I’m sure that Herb Parnes never imagined that what he conceived 50 or so years ago not only would have survived this long but would have such a dramatic effect on empirical research in the social sciences. Although we no longer refer to the NLS as the “Parnes data,” the mountain of research you refer to is his legacy.
… to the individuals at the U.S. Department of Labor, the Center for Human Resource Research at Ohio State University, NORC at the University of Chicago, and the U.S. Census Bureau who teamed together to conduct the NLS for 50 years. Without the hard work and dedication of these individuals, there would be no NLS!
If you are current or former member of the NLS team and would like to add your “staff” profile to this website, please contact NLSat50@chrr.osu.edu and we will send you a link to the profile submission page.