Mike joined BLS in 1983 after receiving a Ph.D. in economics from the University of Chicago. In 1986, the NLS program was transferred from the Employment and Training Administration (part of the U.S. Department of Labor) to BLS (also within DOL)and Mike became its first BLS director. He headed the NLS program until 1996, when he left BLS. He is currently a Senior Fellow at the Urban Institute.
Here are Mike’s reflections on the funding crises, non-scandals, and innovations that made his years at BLS memorable.
The 1980s began a tumultuous period for the NLS. With the election of Ronald Reagan, DOL’s budget was cut and two of its three research offices were decimated. To keep this from happening at BLS, Commissioner Janet Norwood removed “research” from every title in the agency so as not to draw attention. ETA tried to end the NLS program, but a lobbying effort saved all but the Young Men and Older Men cohorts.
ETA’s interest in the NLS waned and budget cuts continued, resulting in the administration of the NLSY79 by telephone in 1987 and a cut of $1 million (in 1986 dollars) that supported research at CHRR. When ETA offered the NLS to BLS, we vowed not to take it if it was underfunded. We negotiated to reinstate the funding to return the NLSY79 to an in-person survey, and got back part of the research budget which we used to fund a series of grants to study labor markets. Unfortunately, dollars and positions are not always interchangeable in the government, and BLS only received one position to run the entire NLS program. This was the position I took in 1986.
At the same time, we began to work toward a new youth cohort. Commissioner Norwood asked me to assemble an interagency working group to identify a broad base of support for a new cohort of youth. Through that group—representing about 10 agencies—we developed a plan for a new NLSY92. Funding a new cohort required getting it included in BLS’s budget submission, then DOL’s, through OMB into the President’s budget, and finally approved by Congress. We tried several times and couldn’t get it through all the stages. Finally, after Bill Clinton was elected and some key researchers came to work at DOL, including Katherine Abraham as the new BLS Commissioner, we were able to get funding for the new cohort. During those five years we faced another round of budget cuts (taking the NLSY79 to a biennial survey) before the NLSY97 became a reality.
We originally set 1996 as the year to launch the new survey. A questionnaire design team was assembled, with top researchers taking on questionnaire sections on various dimensions of the transition to adulthood. Ken Wolpin led the team, and he and I took the material supplied by the design team and wove it into a coherent questionnaire. We sought to improve known problems with the NLSY79, such as inconsistencies in longitudinal reports of educational attainment. We also tried some new things, including a full parent interview, measures of youths’ knowledge about the health impacts of risky behaviors, and quantifying expectations through use of self-assessments of the percent chance of events occurring in the youth’s lives in the future. We also surveyed the principals in all the high schools in the areas in which the youth lived, which gave information on the full choice set of schools.
We had hoped to try innovative data collection using graphical representations, but software was much cruder in those days and we didn’t have time to make those sorts of changes happen. The complexity of the design and the desire to be thoughtful caused time to tick away. We would have liked extra time, but because 1996 was an election year DOL wanted to be able to say they had started an important, new survey. We were struggling to be ready on time, and were “saved” when the government shut-down delayed the survey until 1997.
Another tribulation occurred in my first year, when DOL’s Inspector General selected the NLS contract as one of several to review. After searching high and low for waste, fraud, and abuse, they came up virtually empty—but their big conclusion was that the data weren’t being used by DOL. If DOL wasn’t using the data, why were we spending all this money on it? This led us to start a new publication, Work and Family (which no longer exists), as a means of using the data at BLS. Although this satisfied the Inspector General, it took us over a year to publish the first issue. Any Inspector General report open after a year is reported to Congress as still open. A reporter for the Dayton Daily News decided to write an article on government waste and chose to focus on open Inspector General reports (from across the government). The NLS was picked up, and the reporter set out to expose our waste. He struggled to find a story, concentrating on the cost of travel associated with the program—but most travel was interviewers’ and he was disappointed to find little expense one might consider out of line. I was the only government employee who travelled for the NLS and he requested (under the Freedom of Information Act) my travel expense reports. Discovering I had gone to conferences in Toronto and Vancouver, he noted that I had travelled to “exotic” places. (I told him that most of my travel was to Columbus, and only someone from Dayton thinks that Columbus is exotic.) His story got picked up by a Columbus TV news program and the anchors questioned whether we needed to collect data every year. After all, they asserted, people don’t change jobs every year. We showed in a later issue of Work and Family that even at age 30, about one-quarter of individuals had switched jobs in the previous year.
As I settled into my new role directing the NLS program, we identified several goals. One was to revitalize the Young Women and Mature Women cohorts. When the surveys began, they were meant to last five years with an in-person interview in years one and five and two telephone “check-in” interviews in between. When BLS took over the NLS, the women’s cohorts had been going for twenty years, but the interview pattern and basic data collection had not changed. I changed the data collection schedule to biennial, in-person interviews with one of the two cohorts going each year. This allowed us to smooth our budget from year to year, and to modernize the cohorts by use of event histories; later, it allowed us to move to CAPI interviewing. We also updated the topics covered to reflect the aging of the cohorts. In particular, we shifted the focus of the Mature Women to retirement, and collected pension plans of the respondents and their husbands.
Shortly after I took over the program, we started the Technical Review Committee to provide input on all aspects of the NLS program. I made a point of going to the major conferences as well as specialized meetings to keep abreast of how researchers use NLS data and to solicit ideas for future directions. We offered a small grants program to fund new research using NLS data and we sponsored a users’ workshop for researchers to learn about the data.
In addition, we connected with survey methodologists to engage them in using the NLS for their research, and to gain advise on issues that affect NLS survey design and analysis. Our interest in data collection issues focused in particular on attrition, an important dimension of longitudinal surveys. We commissioned a study on how representative the NLSY79 remained and joined with the PSID to sponsor a conference on attrition in longitudinal surveys and how attrition might have affected researchers’ results. [Link to Macurdy et al.]
I left BLS as the NLSY97 was going into the field. It felt like I was putting up my child for adoption. It was difficult to leave the NLSY97 at such an early stage, but I knew it would be in good hands, as would the other cohorts. I’ve gone on to do things I could never do while directing the NLS, especially while at a statistical agency—but I gave up being involved in facilitating the latest research in labor economics, demography, child development, risky behaviors, income and wealth, and a myriad of other fascinating issues. The NLS has survived for 50 years despite several significant reductions and outright attempts to end it because—despite having no legislative mandate— the breadth and depth of the surveys make the data valuable to researchers across many disciplines. This research provides important knowledge, but also the fundamental underpinnings of social policy.
… 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.