Employment seems to be growing very slowly, if not shrinking, while output (measured with lots of noise) seems to be rising, into Q3. Would NBER ever determine a recession occurred when GDP was rising?

Figure 1: Quarter on Quarter annualized growth rate of GDP (bold blue), GDPNow (light blue square), nonfarm payroll employment implied preliminary benchmark (red), and civilian employment experimental series with smoothed population controls. (tan). Employment data is quarterly average of monthly data Source: BEA, Atlanta Fed, BLS via FRED, BLS, and author’s calculations.

There are instances of NBER recession dates encompassing both negative NFP employment and positive GDP growth, and vice versa. Usually, former is during periods in the middle of, or latter end of recessions. In the 2001 recession, there is only one quarter of negative GDP growth during the recession dates (and one quarter during the peak quarter), and not two consecutive quarters of negative GDP growth (so much for the two quarter rule!). However, there are three consecutive quarters of negative employment growth, encompassing peak-to-trough.

At the monthly frequency (going back to 1992 using Macroeconomic Advisers/SPGMI series), one finds the 2008-09 recession has NFP employment growth negative from 2008M01 onward, while monthly GDP growth is negative in 2008M02, and then 2008M07 onward.

Figure 2: Month-on-Month annualized growth rate of monthly GDP (bold blue), nonfarm payroll employment implied preliminary benchmark (red), and civilian employment (tan). NBER peak-to-trough recession dates shaded gray. Source: BLS via FRED, S&P Global Market Insights (nee Macroeconomic Advisers, IHS Markit) (9/2/2025 release), and author’s calculations. 

What’s the picture look like today?

Figure 3: Month-on-Month annualized growth rate of monthly GDP (bold blue), nonfarm payroll employment, implied preliminary benchmark (red), nowcasted NFP (pink square), and civilian employment (tan). Source: BLS via FRED, S&P Global Market Insights (nee Macroeconomic Advisers, IHS Markit) (9/2/2025 release), and author’s calculations. 

I estimate the level of nonfarm payroll employment for October using ADP’s private nonfarm payroll (NFP) employment to estimate BLS private NFP, and then add 22K (September government employment growth) and subtract 150K (for Deferred Furlough Program).

Nonfarm payroll employment and civilian employment have both been bouncing around zero (annualized) growth, while monthly GDP growth has been volatile. Of these indicators, the NBER’s Business Cycle Dating Committee places highest weight on employment and personal income ex-transfers. So, it could well be that we’re in a recession now, focusing on employment indicators, but we won’t know for a long time (and we’ll have only a little inkling come early December, for October NFP).