BONUS First Monday: What Led to the BLS's Inaccurate Jobs Report?

      The Bureau of Labor Statistics (BLS) releases significant job revisions on Friday mornings, eliminating almost 90% of the previously reported job gains for May and June. This raises essential questions about the calculation of our most trusted economic data.

      In this episode, we explain how the system operates, examining the reasons behind such large revisions and what they imply for our understanding of the actual economy.

      Friday comes, and the BLS reports what seems to be a normal update: 73,000 new jobs added in July. However, the revisions reveal a contrasting reality. The job gains for May, initially reported as 144,000, are adjusted down to 19,000, while June's initially robust figure of 147,000 drops to just 14,000. These corrections represent 87-90% overestimations and significantly change the economic outlook for those months.

      Each month, the BLS surveys 560,000 businesses and uses payroll data from the 12th day of the month. Yet, only 60-73% of those businesses respond by the deadline for the initial release. The remaining data is estimated through statistical modeling that relies on historical trends.

      Typically, this method results in revisions within the range of 20,000-50,000 jobs. However, throughout 2025, the average monthly revisions have escalated to 66,000, which is three times the usual amount. The statistical models are failing to effectively reflect the current economic conditions.

      This issue becomes apparent when economic environments shift quickly, rendering historical patterns unreliable. The annual revision for 2024 marked the largest change since 2009, a year characterized by the Great Recession—a time when traditional forecasting struggled to keep pace with rapid changes.

      ADP, a private payroll processor serving 460,000 companies, offers valuable comparative data. For May, their estimate of 37,000 private-sector jobs aligns reasonably well with the BLS's revised figure of 19,000. In June, ADP reports a loss of 33,000 jobs, while the BLS indicates a gain of 14,000.

      ADP's independent data serves to validate the revised figures and emphasize the extent of the initial inaccuracies.

      These statistics influence real-life decisions. Federal Reserve officials rely on employment data for interest rate policies, investors allocate capital based on these reports, and workers make career choices based on their perception of labor market strength.

      When the initial data is off by 90%, everyone is functioning with fundamentally flawed information.

      The revisions highlight the vulnerability of our economic measurement systems when conditions evolve more swiftly than the models can accommodate.

      Thanks to our sponsors!

      Indeed

      If you’re seeking exceptional talent to enhance your team, turn to Indeed. Visit indeed.com/paula for a $75 job credit to elevate your listing and begin hiring today.

      Shopify

      Broaden your business by offering both physical and digital products with Shopify’s comprehensive platform. Visit shopify.com/paula for a $1/month trial and gain full access to Shopify’s complete feature set.

      Policygenius

      Check out policygenius.com for free quotes and comparisons from over 30 insurers. With Policygenius, you can find life insurance policies starting at just $276 annually for $1,000,000 of coverage.

      Quince

      Quince provides a variety of high-quality items at accessible prices. Go to Quince.com/paula for free shipping on your purchase and a 365-day return policy.

Other articles