In April, Chairman Bernie Sanders and Ranking Member Bill Cassidy of the Senate HELP Committee asked for public input as they consider reauthorizing the Education Sciences Reform Act—the 2002 law authorizing the Department of Education’s main research and statistical collection activities. A reauthorization bill will likely be on the Committee’s agenda in 2024.
My colleague Robert Bellafiore and I have written a forthcoming paper with our recommendations for reforming ESRA that we will be publishing next month. As we argue, one of the simple recommendations for Congress would be to improve the timeliness of the National Center for Education Statistics data collection and reporting activities, to make this information more useful for policymakers and the public.
Institute of Education Sciences Director Mark Schneider made this point in a recent interview with Eduwonk:
Among my priorities is the insertion of a legislative mandate for timeliness in all of IES’ activities. Right now, far too much of our data and research findings are stale by the time they are released. This long lag time is often done in the name of accuracy—but being accurate at the fifth decimal point is false precision while being years late is disqualifying for much of what we are studying or gathering data about.
Yesterday, the National Center for Education Statistics (NCES) released tables with new information about state education practices. But several of the data categories are long out of date. For example, a table on “states with voucher programs” presents data from 2017. The University of Arkansas’s Patrick Wolfe summarized the problem on X: “I’m encouraged that NCES finally is posting information about #SchoolChoice programs, but ‘states with voucher programs as of 2017’? Really, that’s the best you can do? Seems like there have been important changes in the past 6 years.”
According to EdChoice, seven states enacted new private school choice programs, including education savings accounts, in 2023 alone. A researcher or policymaker who relies on the “new” table from NCES to analyze how or whether states enable private education choice would have little understanding of current laws and programs.
The “new” NCES data also has a breakdown of state assessment policies from 2018 and high school completion policies from 2017 and 2018.
The newly released data aren’t an anomaly. Other important information published by NCES about American schools is regularly out of date. In October 2023, NCES released per-pupil revenue and expenditure data for FY2021; NCES’s most current information about public school resources and expenditures is therefore at least two years out of date. Significant policy changes affecting school governance and revenues have occurred since FY2021, without the benefit of these data. Moreover, states have received nearly $190 billion in Elementary and Secondary School Emergency Relief funds during the pandemic, which suggests that current per-pupil expenditures are likely much higher than what was reported for 2021.
NCES should have access to timely information about public school expenditures. One effort to address this problem, the 2015 Every Student Succeeds Act, requires states to report each school’s per-student expenditure. Taking a different approach to the problem, the Foundation for American Innovation’s Labs team created Project Nickel, in partnership with EdChoice, to create an easy-to-use public website to search per-student spending by school. The website uses data collected by the Edunomics Lab at Georgetown University.
If academic researchers and technologists at FAI can collaborate to collect and publish per-student data by school, why can’t the National Center for Education Statistics publish timely data on average state per-pupil expenditures without a two-year lag?
As Congress considers reauthorizing ESRA, many reforms are needed to improve the value of the federal government’s research and development and statistical collection activities. But requiring the National Center for Education Statistics to collect timely and relevant data should be at the top of the list.
