Trust Issues

The United States Bureau of Labor Statistics (BLS) uses two surveys to assess labour market conditions: the establishment survey, which produces the monthly jobs numbers, and the household survey, which is used to calculate unemployment and participation rates.

The establishment survey External link. draws on a sample of 121,000 businesses and government agencies. The respondents are expected to submit their data at the end of the month, and roughly two-thirds typically do so on time. The BLS keeps the window open for another two months and reports the late responses as revisions. As such, the revisions reflect newly available information, rather than inaccuracies in the original submissions.

In the July BLS jobs report, the two-month downward revision for May and June, at -258,000, was the largest since the pandemic.

Economists have highlighted several reasons for the large revisions: a) Many firms reported later than usual. Late responders are typically small businesses and local government agencies, thus conditions at small enterprises may be weaker than at large ones; b) Rapidly changing tariffs, deportations and DOGE policies may have created genuine uncertainty about staffing decisions; c) falling response rates; and d) updated seasonal factors.

Given the above, accusing the BLS Commissioner of manipulating the data for political reasons is an extremely serious allegation. As the New York Times put it, “When President Trump didn’t like the weak jobs numbers, he fired the person responsible for producing them.”

Has one of the most critical official statistics for monitoring the US economy become politicized, and thus unreliable? William Beach, former BLS Commissioner appointed during Trump’s first term, sought to dispel such concerns in a Bloomberg interview. He emphasized that the agency is staffed by “patriots” who are deeply committed to objectivity, assuring that the integrity of the data will remain intact.

Still, many market participants will be more skeptical. Alternative, non-official data sources could start to carry more weight in shaping market expectations than before. This poses a challenge for the Federal Reserve, given that the unemployment rate is a core element of its mandate. Communicating future policy decisions may become increasingly difficult, especially when the president simultaneously calls for lower interest rates while questioning data that signals labour market weakness.

Perhaps the biggest surprise is how little this has rattled financial markets. In most other countries, such developments would trigger a sharp steepening of the yield curve and a currency sell-off. Yet, US treasuries have behaved as if not much has happened. Our measure of term premia has increased marginally, but long-end yields have fallen overall, with markets pricing in a steady pace of rate cuts starting in September. The US dollar has also remained broadly stable. Clearly, investors continue to place strong faith in the credibility of US institutions, at least for now.