What Happened and How the BLS Produces Jobs Data

President Donald Trump recently dismissed Bureau of Labor Statistics (BLS) Commissioner Erika McEntarfer, saying the July jobs report had been “rigged” to damage Republican prospects. That allegation drew widespread attention because monthly employment figures are closely watched by policymakers, investors and the public. The controversy also prompted former BLS officials and labor data experts to explain how the agency compiles and revises its estimates, and why routine adjustments do not imply intentional manipulation.

Contrary to the implication that the commissioner personally prepares or alters the monthly numbers, former BLS officials say the commissioner does not directly participate in the hands-on tasks of data collection or initial tabulation. The BLS relies on established statistical processes and multiple staff layers to produce its employment measures, and the agency’s output is built from two primary surveys that capture different aspects of the labor market.

The first major source is the Current Population Survey (CPS), a household survey that interviews roughly 60,000 households each month. The CPS gathers demographic and employment information directly from individuals and determines measures such as the unemployment rate, labor force participation and the number of employed persons. Because it interviews households, the CPS captures people’s self-reported work status and hours, and it is an important complement to payroll-based measures.

The second principal source is the Current Employment Statistics (CES) survey, often referred to as the payroll survey. CES collects payroll data from a sample of businesses and government agencies to estimate changes in nonfarm payroll employment and average hourly earnings. This survey provides detailed industry-level information and tends to be the primary source cited for the widely reported monthly jobs gain or loss in the economy.

Monthly employment figures published by the BLS are initial estimates based on the responses received within reporting deadlines. Revisions to those initial estimates are common in subsequent months. Revisions occur for several routine reasons: late survey returns that arrive after the preliminary tabulation, updated payroll records that businesses provide in follow-up filings, and the application of seasonal adjustment factors that smooth predictable variations across months. These and other technical adjustments are part of normal statistical work and are disclosed by the agency.

Experts and former agency staff emphasize that such revisions are not unusual and do not constitute proof of data manipulation. Seasonal adjustment, for example, is a standard statistical technique used to remove recurring calendar-related effects—such as holiday hiring or school schedules—that would otherwise make month-to-month comparisons misleading. Delayed responses and additional administrative data that come into the system after the first release can also change the picture when the series are recalculated.

The BLS publishes methodological notes, revision histories and technical documentation alongside its monthly releases so analysts can see how and why numbers change. These documents typically describe the survey designs, sample sizes, weighting procedures and the mechanics of seasonal adjustment. Transparency around methods and documented revision practices helps outside researchers and the public evaluate the reliability of the estimates and the magnitude of any subsequent changes.

Public debate about the integrity of economic statistics often arises when headline numbers affect political narratives or market expectations. In this instance, the dismissal of the commissioner and the claim that the jobs report was “rigged” triggered scrutiny of how the numbers are produced and how routinely they are revised. Observers with experience in labor statistics reiterate that routine revisions reflect the realities of survey work and the layering of data sources, not an individual official’s intervention in the data.

Understanding the distinction between the agency leadership role and the technical operations that generate employment statistics is key to assessing such claims. While leadership sets policy and oversees operations, the compilation of monthly employment estimates is a technical process carried out by economists and statisticians under established procedures. For readers and stakeholders following employment trends, awareness of the two main BLS surveys—the household-based Current Population Survey and the payroll-based Current Employment Statistics survey—along with the normal causes of revisions, provides context for interpreting monthly job reports and the occasional adjustments that follow.