seasonal adjustments, but the focus is generally on new net hires from one month to the next. Month-on-month job growth is seen as a sign of continued confidence in increasing revenues that pay for more workers and leave some room for profits after other expenses are paid. These data initially come from household surveys on job gains or losses and are then revised based on a core survey known as the Current Employment Statistics (CES) survey, which collects data from thousands of businesses and government agencies. CES survey respondents submit employment data, including hours worked and wages/salaries, for all paid workers to the BLS from their payroll records. Bureau of Labor Statistics economists then use models to finalize estimates, and the jobs report is created. When the June 2025 numbers were released in July, growth continued. However, as more complete payroll data come in and monthly and quarterly numbers are prepared (for what is ultimately the final statement on labor data, called the Quarterly Census of Employment and Wages or QCEW), revisions are made. If you visit the graphic source website, you will see that there are monthly revisions of various magnitudes, occurring in both directions, with more or fewer workers hired. In the accompanying figure, when the red column is taller than the blue column, the revised labor market outcome suggests that fewer jobs were created than initially estimated. What created controversy in July and August 2025, and may continue to do so as the BLS leadership position is debated and changed, is how politicians may view those numbers versus the qualitative story to tell. The revisions for May and June 2025
were somewhat larger than recent changes and consecutive, but not necessarily out of bounds in terms of percentage changes from historical norms, nor without both strong theoretical and empirical foundations. Suppose the sources, credibility and modifications to these methods were adjusted to consistently provide accurate job numbers, regardless of economic conditions. Unemployment, for example, is a condition affecting the working-age population, not an indicator of employer hiring conditions, as with non- farm payroll data. If more people lose their jobs and go on unemployment, the number of insurance claimants rises while jobs are in demand. The working-age population available for work is in question; economists will wonder where the disconnect may lie. Global financial markets will also get nervous quickly. The United States remains the global leader in data sophistication, accuracy and confidence. We need to closely monitor these data and ensure that the methods and reporting remain consistent and scientific at their core. The last thing the global economy needs is another concern about American economic outcomes. . g
Dr. Robert Eyler is professor of economics at Sonoma State University and president of Economic Forensics and Analytics in Sonoma County.
October 2025
NorthBaybiz 19
Made with FlippingBook Annual report maker