The recent downward revision of U.S. payroll figures by 818,000 jobs, the largest since 2009, has sparked debate over whether it signals an impending recession. The revisions, which cover the period from April 2023 to March 2024, have raised concerns about the accuracy of the economic models used by the Bureau of Labor Statistics (BLS), with some economists questioning whether the data overstates the economy’s strength amid signs of softening in the labor market.
Historical parallels have been drawn to 2009 when a similar revision overstated jobs by 824,000. At that time, the National Bureau of Economic Research had already declared a recession six months earlier. Jobless claims had surged to over 650,000, and the insured unemployment rate peaked at 5%. GDP had already shown negative growth for four consecutive quarters, signaling a clear economic downturn.
In contrast, the current situation presents a different picture. No recession has been declared, and key economic indicators such as the 4-week moving average of jobless claims and the insured unemployment rate have remained stable over the past year. The reported GDP has shown positive growth for eight consecutive quarters, further distinguishing the present from the 2009 scenario.
While the downward revision in job growth by an average of 68,000 per month during the revision period is significant, it only reduces the average monthly employment growth to 174,000 from 242,000. The impact of these revisions on current economic conditions will depend on how the BLS distributes the weakness over the 12-month period. If the revisions are concentrated towards the end of the period, they could have more relevance to the current situation and possibly influence Federal Reserve policy.
There is speculation that the revisions could modestly increase the likelihood of a 50 basis-point rate cut by the Fed in September, especially if the labor market is weakening faster than previously thought. However, the Fed is likely to prioritize current jobless claims, business surveys, and GDP data over these backward-looking revisions when making policy decisions.
It is also worth noting that revisions in jobs data are not always consistent. In the past 21 years, revisions have only followed the same direction 43% of the time, with negative revisions often followed by positive ones the next year, and vice versa.
Further complicating the issue, economists at Goldman Sachs have suggested that the BLS may have overstated the revisions by as much as half a million jobs. They attribute this discrepancy to factors such as unauthorized immigrants who were initially listed as employed but are not in the unemployment system, as well as a general tendency for initial revisions to be overstated.
Despite the noise and potential volatility in the jobs data, other macroeconomic indicators do not currently point to an economic downturn akin to 2009. The debate continues, but for now, the broader economic data suggests that the economy is not in the same dire straits as it was during the Great Recession.
