Fed: the silent crisis

The Federal Reserve, the central bank of the United States of America, has recently lost access to a valuable macroeconomic indicator: anonymized data on employment and wages provided by ADP, the payroll-processing giant. More than just a technical anecdote, this statistic—no longer available to the Fed—poses a threat to stability and hinders its ability to steer the economy.

ADP is a private company that manages the payrolls of 20% of the U.S. private workforce—several million workers—and since 2017 had been generously sharing with the Fed a set of aggregated and anonymized data. Available with only a one-week delay, these data offered an almost real-time snapshot of the U.S. labor market in terms of hiring, wages, and sectoral trends.

For the Fed’s economists, this precise analytical tool allowed them to detect weak signals—cracks not reflected in the official statistics of the Bureau of Labor Statistics (BLS), which are often published late and frequently revised afterward.

This Fed–ADP partnership was praised by Jerome Powell himself, the central bank’s chairman, who revealed that thanks to the ADP data, Fed economists were able to refine a method for predicting revisions to the BLS’s official data. A model of institutional intelligence, this alliance between public power and private expertise served the goal of better economic forecasting.

Fed researchers demonstrated the impact of these weekly ADP data, which allowed them to detect very early—during the 2020 pandemic—the disruptions shaking the labor market. As jobs imploded, the Fed could measure the damage in real time and act proactively—something impossible under the bureaucratic delays of the BLS.

Powell went even further, stating: “We would have seen the Great Recession coming earlier if we had had these data in 2008!”

Structural vulnerability

Everything, however, came to a sudden halt late last summer, shortly after one of the seven members of the Fed’s Board of Governors, Christopher Waller, mentioned in a mere footnote to one of his speeches some ADP indicators that, in his view, pointed to a slowdown in employment. Citing these data to support his concerns about the labor market, he noted that hiring was falling more sharply than suggested by the latest official government figures.

Officially, ADP said it was “reassessing its processes” to ensure “the highest standards of confidentiality,” and thus stopped transmitting its weekly data. In reality, no one knows what truly motivated this break.

Nevertheless, this sudden reversal exposes the risks of public institutions depending on private actors. In a country where government shutdowns have become a recurring partisan weapon, delegating economic monitoring to a publicly traded company now seems, in hindsight, ill-advised.

This abrupt suspension reveals that the production of economic knowledge—supposed to guide collective decisions—now largely depends on the goodwill of private players. What happens when those companies decide to withdraw?

Having tried in vain to dissuade ADP, Powell—making a rare admission of powerlessness for the head of the world’s most powerful central bank—had to acknowledge that, deprived of these data, the Fed was left only with “imperfect substitutes” to assess labor market trends.

A cruel paradox: the Fed now proceeds blindly at a time when monetary policy demands the greatest precision. It speaks volumes about our era and goes far beyond the issue of statistics and economic data. The crisis of confidence seems real: is the state still capable of producing reliable knowledge?

A warning for democracies

Beyond the American case, this episode illustrates a growing tension within integrated democratic economies. The alarm signal is indeed global: public authorities depend on data they no longer control and cannot even verify.

As the digitization of the economy gives private companies a massive lead in information collection, governments find themselves downgraded to the role of clients—sometimes tolerated, sometimes excluded.

The Fed is obviously not the only one affected. Tax authorities, health agencies, and even statistical institutes increasingly rely on data from platforms or tech firms. If one provider decides, for commercial or political reasons, to turn off the tap, an entire segment of public decision-making is thrown into disarray.

Statistical sovereignty

The ADP–Fed case demonstrates the urgent need to establish genuine statistical sovereignty. Fundamental economic data are undoubtedly a common good—just as vital as electricity or money. Their production, dissemination, and preservation must be guaranteed by independent institutions, protected from political whims.

What a predicament, what difficult choices, what risks of error might face less powerful and less visible administrations if even the Fed—an institution endowed with immense resources and global prestige—finds itself powerless before a private company?

Without massive investment in public data infrastructure, democracies risk losing both their ability to act and their legitimacy.