After President Donald Trump fired Bureau of Labor Statistics commissioner Erika McEntarfer, claiming that jobs data were “faked” and “rigged,” an urgent question arose: Could private sector data fill the gap if political meddling erodes the integrity of official statistics? Could private companies even take over producing them?
Private sector data increasingly supplement government statistics, especially as survey response rates decline. Experts advocate shifting toward administrative records and proprietary data. But while private sector data are valuable, they cannot replace official statistics. They are an excellent complement—but not a substitute. Recent administration actions, however, have made it harder for the private sector to be even a complementary partner. It’s time to reverse course.
Private sector data offer speed and specificity that official statistics cannot match, despite not being representative or comprehensive. At the online real estate platform Trulia during the foreclosure crisis in the late 2000s and early 2010s, my team built a list-price index showing home-value trends months ahead of leading sales-price indexes. At the employment website Indeed during the COVID-19 pandemic, we tracked job postings weekly, capturing the labor market’s crash and recovery in near real time. Credit card transactions, satellite imagery, and cell phone locations all also help reveal economic behaviors that fill gaps in government data.
Companies can quickly create indicators for sectors, geographies, or demographic groups too small to be broken out in federal surveys. They can define and track trends using their own taxonomies—like identifying “data scientist” roles before official classifications exist. The tech industry brings a strong culture of user feedback, with lessons for how agencies could make data easier to find, visualize, and interact with.
Still, private sector indicators cannot match the authority, credibility, and durability of official US government data, which remain the global gold standard. Most importantly, official statistics are representative by design. Through sampling or near comprehensive administrative records, they aim to reflect the full population or economy. In contrast, companies only see their users or customers. Credit card data miss the unbanked. Job sites do not capture those who have stopped looking for work. Cell phone data miss people without phones. In fact, companies use official statistics to benchmark their own data—making private data only as good as the public data they rely on.
Official agencies exist to serve the public good. Their statistics inform policy, determine the allocation of public money, and guide private sector decisions. Most companies have other priorities. Legal teams worry about liability, executives about trade secrets, public relations departments about optics. Firms might hesitate to release unfavorable data—especially when politicians fire agency leaders for bad news.
Proprietary data can quickly adapt to new trends—like emerging job titles—but official statistics maintain consistent methodologies, with documented changes that enable long-term comparisons. A private data series launched last year will not show how today’s economy compares with the past.
Government data are subject to strict quality controls, confidentiality protections, and public scrutiny. Errors are documented and corrected. Private data lack this rigor, and errors can go unnoticed. The private sector has little incentive to fund the kind of quality control and public access that are part of the mandate of government agencies.
Making the most of private data therefore means using both private sector data and public sector administrative records to supplement traditional surveys. A promising proposal now open for public comment would create opt-in digital data feeds between companies and the Census Bureau, eliminating the need for manual survey responses.
Unfortunately, the administration has weakened the foundation for private sector collaboration on government statistics, in three ways.
First, to make efforts like this work, government agencies need funding and talent to ingest private data, test it rigorously, and assess how new methods affect long-standing series. This is not a job for amateurs. It requires running old and new methods in parallel to evaluate methodological changes and establish reliability.
However, statistical agencies have suffered severe attrition, including senior career staff with deep expertise. Ironically, former BLS Commissioner McEntarfer is best known for her pioneering work with administrative data at Census. Basic operations are under budgetary pressure, leaving modernization impossible. The president’s budget proposal for the next fiscal year calls for further cuts to the Bureau of Labor Statistics, the Bureau of Economic Analysis, and Census Bureau programs aside from the decennial Census.
Second, companies and their private sector customers must have conviction that their data will be kept confidential and not misused. Data sharing—survey-based or digital—depends on trust. But the administration is pushing boundaries here, such as trying to use IRS taxpayer data and insurance claims data for immigration enforcement. This violates long-standing norms and could make companies and other partners reluctant to open data spigots. (More urgently, this also discourages immigrants from paying taxes and seeking medical care.)
Third, statistical agencies should engage private sector experts to understand proprietary data and build trust. While keeping data confidential, discussions about limitations and methodology should be open to outside experts. Without transparency, public confidence is impossible. However, the administration disbanded the statistical agencies’ outside advisory committees. These committees included private sector experts from companies that would be essential data sharing partners.
The private sector could modernize and improve official statistics—as a partner, not an alternative. But the administration must change course on resourcing, staffing, data sharing, and engagement. Weakening statistical agencies will not just harm public data. It will impede private-public partnership and will reduce the quality of data that the private sector produces on its own.
Data Disclosure
This publication does not include a replication package.
