A window into America’s high schools slams shut

The choices you make as a teenager can shape the rest of your life. If you take high school classes for college credit, you’re more likely to enroll at a university. If you take at least 12 credits of classes during your first year there, you’re more likely to graduate. And those decisions may even influence whether you develop dementia during your later years.

These and insights from thousands of other studies can all be traced to a trove of data that the federal government started collecting more than 50 years ago. Now that effort is over.

On a single day in February, the Trump administration and its Department of Government Efficiency canceled a long-running series of surveys called the high school longitudinal studies. The surveys started in 1972, and they gathered data on more than 135,000 high school students through their first decade or so of adulthood — sometimes longer.

“For 50 years, we’ve been mapping a timeline of progress of our high school system, and we’re going to have a big blank,” said Adam Gamoran, who leads the William T. Grant Foundation and was nominated to head up the Education Department’s research and statistics arm under President Biden, but was never confirmed. “That’s very frustrating.”

The data collection effort has been going on since before the founding of the modern Department of Education. Thousands of journal articles, books, dissertations and reports have relied on this data to form conclusions about American education — everything from how high school counselors should be spending their days to when students should start taking higher-level math classes.

The Department of Government Efficiency first canceled contracts for the collection of new long-term high school data and then started laying off staff. The National Center for Education Statistics used to have nearly 100 employees. Today, only three remain.

“The reduction — annihilation — of NCES functionally is a very serious issue,” said Felice Levine, former executive director of the American Educational Research Association, one of the groups suing the administration over these actions. “Maybe it doesn’t appear to be as sexy as other topics, but it really is the backbone of knowledge building and policymaking.”

The Department of Education is reviewing how longitudinal studies “fit into the national data collection strategy based on studies’ return on investment for taxpayers,” according to an email from its spokesperson. The statement also said the department’s Institute of Education Sciences, which is in charge of overseeing research and gathering statistics, remains committed to “mission-critical functions.”

“It seems to me that even if you were the most hardcore libertarian who wants the government to regulate almost nothing, collecting national statistics is about the most innocuous and useful thing that a government could do,” said Stuart Buck, executive director of the Good Science Project, a group advocating less bureaucracy in science funding.

“The idea of a Department of Governmental Efficiency is an excellent idea, and I hope we try it out sometime,” he said. But the effort, “as it currently exists, I would argue, is often directly opposed to efficiency. Like, they’re doing the exact opposite.”

He likened the approach to “someone showing up to your house and claiming they saved you $200 a month, and it turns out they canceled your electricity.” 

‘We can’t just pick this back up later’

Since the effort began in the early 1970s, the federal government has collected data on six large groups of high school students, each numbering in the tens of thousands. Researchers surveyed each group at least once during high school, along with their parents and teachers. Researchers then contacted the students periodically after that, generally over the course of a decade or so — sometimes longer. They collected transcripts and other documents to track progress, too. In total, the data set contains thousands of variables.

The studies are called longitudinal, because they take place over a long time. The methodology is similar to studies that track twins over their lifetimes to determine which traits are genetic and which are caused by their experiences. Such data sets are valuable because they allow researchers to tease out effects that can’t be seen in a single snapshot, but they are rare because they require sustained funding over decades. And the high school data covers a large number of participants selected to represent the national population, giving insights that can be broadly applicable across states.

That vast repository of data affects students “indirectly, but profoundly,” said Andrew Byrne, who runs the math department at Greenwich High School in Connecticut. For example, research based on the data has shown that high school students who take classes for college credit have a better chance of finishing their bachelor’s degrees on time.

Byrne said that research informed the school’s decision to start offering a new Advanced Placement precalculus class when the College Board unveiled it two years ago. The new offering gave some high school students in lower-level math classes the opportunity to get college credit for the first time.

“Success in AP precalculus could empower them to believe they can succeed in college-level classes overall,” Byrne said. A student probably would not read the academic research, but “they live the results of the decisions that data informs,” he said.

Follow-up surveys for the group first contacted in 2009 — made up of people who started high school during the Great Recession — and for students who were high school freshmen in 2022 have been canceled. The latter group, who were middle schoolers during the pandemic, will be graduating next year.

Elise Christopher oversaw the high school longitudinal studies at the National Center for Education Statistics until she was laid off in March along with dozens of her colleagues. Christopher, a statistician who worked at the center for more than 14 years, is concerned about the data that was scheduled to be collected this year — and now won’t be.

“We can’t just pick this back up later,” she said. “They won’t be in high school. We won’t be able to understand what makes them want to come to school every day, because they’ll be gone.”

Researchers were hoping to learn more about why chronic absenteeism has persisted in schools even years after Covid-19 abated, Christopher explained. They were also hoping to understand whether students are now less interested in attending college than previous generations.

“Every single person in this country who’s been educated in the past 50 years has benefited from something that one of these longitudinal surveys has done,” she said.

Levine said the planned follow-up with students from the 2009 high school group would have helped reveal how a greater emphasis on math, science and technology in some states has influenced student decision-making. Were they more likely to study the hard sciences in college? Did they continue on to careers in those fields?

“These are the kinds of things that the public wants to know about, families want to know about, and school administrators and counselors want to know about,” she said.

Data could help solve the mystery of dementia

About 25,000 people who completed the high school survey in 1980 were contacted again by researchers decades later.

Rob Warren, the director of the University of Minnesota’s Institute for Social Research and Data Innovation, is hoping those people — now in their 60s — may help him and other researchers gain new insights into why some people develop dementia, while others with similar brain chemistry don’t.

“Education apparently plays a big role in who’s resilient,” Warren said. “That’s kind of a mystery.”

The people who participated in the high school study may offer a unique set of clues about why education matters, Warren explained.

“You need all that detail about education, and you need to be able to see them decades later, when they’re old enough to start having memory decline,” he said. Other studies can measure cognition, but to measure whether education plays a role in dementia outcomes, “you can’t really test (that) with other data,” he said.

So Warren’s team got permission from the federal government to contact the group in 2019. Researchers asked all the usual types of questions about their jobs and lives, but also gave them cognitive tests, asked medical questions, and even collected samples of their blood to monitor how their brains were changing as they aged.

Warren is continuing his research even though the federal government has canceled future high school surveys. But the staffing cuts at the Department of Education have hampered his ability to hand the data off to the center or share it with other researchers. To do that, he needs permission from the Department of Education, but getting it has been a challenge.

“Very often you don’t hear anything back, ever, and sometimes you do, but it takes a very long time,” Warren said. Even drafting legal agreements to make the data available to the National Institutes of Health — another federal agency, which funded his data collection effort and would be responsible for handling the medical data —  has been a bottleneck.

Such agreements would involve a bunch of lawyers, Warren said, and the Department of Education has laid off most of its legal team.

If the data isn’t made available to other researchers, Warren said, questions about dementia may go unanswered and “NIH’s large investment in this project will be wasted.”