Imagine if highways and bridges suddenly vanished overnight.
Commerce would halt, supply chains would collapse, and millions would be disconnected or even stranded. While this scenario is far-fetched for physical infrastructure, it's what's happening to some of America's most important public data infrastructure.
In recent weeks, the federal government has removed key public datasets and more than 8,000 web pages, dismantling the internet’s roads and bridges connecting policymakers, businesses, and researchers to essential information for improvements and innovation.
The short-term impact is already reverberating through every sector of American society. The news media has had their hands full, publishing a flurry of stories in past couple weeks tracking the developments (see stories from The Hill, the Associated Press, The Week, NPR, POLITICO, The Verge).
The pace of developments is volatile and fluid. We know there have been data and web page removals, modifications, reposting, altogether causing much disruption at certain federal agencies. We do not know to what extent these actions will expand to other departments or agencies, or to what degree changes are temporary or permanent.
Frenetic and disruptive decision-making are not limited to the information from federal data and web pages. Late yesterday the Washington Post and Education Week reported the U.S. Department of Education cancelled at least 90 research contracts in one fell swoop. The long-term impacts are unclear because the communication and information continue to be vague and limited around these actions. But in the short-term, the uncertainty only raises more and more questions absorbing time and attention from public officials, federal workers and contractors, journalists, interest groups, and researchers.
The opportunity costs are only accelerating.
The Importance of Federal Data for EdChoice Research
In full disclosure (and probably an obvious one!), I am not a neutral observer. If disruptions and data removals continue, and even expand, they will negatively affect our work, as well as others in the nonprofit, private, and academic sectors.
At EdChoice, our research team regularly uses and incorporates federal data, particularly the data and information from the National Center for Education Statistics (NCES) and the U.S. Census Bureau. These data sources provide a rigorously collected and reliable source of information for our own data collection, surveys, and analyses, ensuring that our findings are credible, impactful, and useful for policymakers, researchers, and advocates.
Here are specific examples how we incorporate federal data and tools into flagship projects:
EdChoice’s annual almanac of K-12 choice programs across the United States relies on NCES and Census Bureau data that detail statewide student enrollments and most recent average per-student spending and revenue. We obtain this information from ELSi, the Common Core of Data, Private School Universe Survey, and online tables via Digest of Education Statistics, Condition of Education, and the State Nonfiscal Public Elementary/Secondary Education Survey. These foundational sources provide valid and reliable insights particularly regarding enrollments and public school spending.
2. Fiscal Effects of School Choice
Our Fiscal Research and Education Center (FREC), led by Marty Lueken, conducts regular fiscal impact analyses of K-12 choice programs. Every 2-3 years we publish a compendium of those studies, The Fiscal Effects of School Choice. FREC’s research and analyses depend on NCES financial data and Census Bureau household income and population demographics data for important comparisons and identifying fiscal gaps between public school spending and K-12 choice program spending.
3. Schooling in America Survey
The annual Schooling in America Survey (SIA) measures public opinion on K-12 education and school choice policies, utilizing demographic data from the U.S. Census Bureau to ensure our survey results reliably represent the national adult and K-12 parent populations. We use a mixed-mode method (online and phone), incorporating both online and telephone surveys, with statistical weighting aligned with the American Community Survey (ACS) 5-year estimates (provided by the Census Bureau) and also phone type usage estimates from the National Health Interview Survey conducted by the Centers for Disease Control and Prevention (CDC). We simply cannot ensure population representativeness without these benchmarking resources.
4. EdChoice Public Opinion Tracker
Like our SIA Survey series, EdChoice’s Public Opinion Tracker utilizes Census population and demographic data for benchmarking and post-field weighting adjustments for representativeness. Also like SIA, our Tracker polling uses NCES collected data and reporting for split-sample experiments to learn how providing information to respondents affects their views and attitudes about K-12 public school funding and spending. When we survey teachers, we use benchmarking data from the NCES National Teacher and Principal Survey for weight adjustments to achieve national representation of America’s teachers.
5. Enrollment Data for Public Schools and Private Schools, and EdChoice Share
We update statewide public school student enrollments annually and statewide private school student enrollments every two years thanks to data collected and reported by NCES. Our EdChoice Share metrics provide state-level insights into schooling sector enrollment rates and school choice program participation rates. By analyzing federal education statistics, we can provide information to policymakers, researchers, and advocates to better understand how K-12 choice programs are varied and evolving across different states.
Federal data and reporting helps to ensure EdChoice’s research and analyses remain grounded in reliable, transparent, rigorously collected, and nationally representative data in the cases of the wide range of Census and NCES survey programs. This approach not only strengthens the credibility of our work but also enhances our ability to advocate for education policies that empower parents/guardians and their children to find the best education that meets their needs and priorities.
Three Principles for Government Data: Transparency, Clarity, and Accessibility
There are some simple principles around data collection and reporting that are worth articulating.
If government collects information and data, then the public and taxpayers should be able to see it.
If data have been made public previously, but are no longer accessible, then it is wise to make public and accessible again.
Ultimately, it comes down to transparency, clarity, and accessibility.
Let’s start with transparency. The public should know what data exists and how it is collected. When data disappears, the government operates behind a curtain, making it harder to hold accountable public officials and government agencies and contractors.
Clarity is an important principle too. Even when data are publicly available, the information must be structured, standardized, and as easy as possible to understand and interpret. The lack of clear communication regarding recent removals has caused confusion among researchers and healthcare professionals, underscoring the importance of transparency and clarity in managing and reporting federal data.
And then there’s accessibility—a growing concern as some government datasets have fallen behind mounting online firewalls or more and more administrative hurdles. When government data are removed or locked away, the American public loses one of its most powerful tools for decision-making and accountability.
The need to uphold these principles are concrete and not hypothetical— they have real-world consequences. Businesses depend on labor market data to guide hiring and investment decisions. Schools and nonprofits use education data to develop programs that serve students effectively. Health researchers rely on federal statistics to track and respond to emerging public health crises. If these datasets are altered, obscured, or made harder to access, everyone from policymakers to parents loses out.
Policymakers from both sides of the aisle should recognize that having transparent, clear, and accessible government data isn’t a partisan issue—it’s a fundamental public good. The Census Bureau, NCES, NIH, and CDC provide crucial information that underpins economic growth, scientific research, and informed public policy. The government has a duty to ensure that this data remains open, reliable, and accessible.
Government Data Infrastructure for an Informed Future
Data collection and reporting doesn’t reflect a recent expansion of the federal government’s role—it's woven into our Constitution. Article I, Section 2, requires an apportionment of representatives every 10 years, providing the constitutional and legal purpose of the decennial census.
James Madison, writing to W.T. Barry in 1822, crystallized the importance of public information and knowledge-building: "Knowledge will forever govern ignorance: And a people who mean to be their own Governors, must arm themselves with the power which knowledge gives." This wasn't mere rhetoric—Madison understood that democratic self-governance requires more information, more knowledge, and less ignorance. Aspirations perhaps more critical today than 200 years ago.
Of course, no research project or survey is perfect. And the same applies to bureaucracies. There is probably wasteful spending and less-timely or lower-quality research that can be found in federal agencies and programs. There can always be improvements toward effectiveness and efficiency. But given the incredible importance of the work that federal statistical agencies do, and all of the related public policymaking and business decisions that rely on such information, any adjustments should be made with as much precision as possible—with a scalpel and not a chainsaw. That is not what we are seeing from the federal government right now.
What is the best way forward? We should expect policymakers and government officials to adhere to firm commitments to information and data transparency, clarity, and accessibility. The federal government’s data are not just a collection of information and numbers—it’s a public trust. And if we lose sight of that, we risk losing much more.
The economic costs are too high for neglect or a dismantling of such data infrastructure, the social implications too severe, and the damage to evidence-based policymaking and advocacy too devastating. We should protect it with the same commitment we bring to maintaining and improving our physical infrastructure. Our global and economic competitiveness, scientific progress, and our democratic republic depend on it.