Nonfarm payrolls growth surprises to the upside (see Discussion here), but combined with the downward revisions in October and November, we are only at consensus levels. Below is a picture of some of the key indicators followed by the NBER Business Cycle Dating Committee, along with monthly GDP and GDPNow.
figure 1: Nonfarm payrolls included in preliminary benchmark (bold dark blue), using Bloomberg consensus implied levels as of 1/3 (blue+), civilian employment (orange), industrial production (red), 2017 excluding Transferred personal income $ (bold green), 2017 manufacturing and trade sales $ (black), 2017 consumption $ (light blue), 2017 monthly GDP $ (pink), GDP second release (blue Color bar) and GDPNow for 2023Q4 as of 1/3 (lavender box), all logarithms normalized to 2021M11=0. Source: BLS via FRED, U.S. Bureau of Labor Statistics preliminary benchmarksFederal Reserve, BEA 2023Q3 2nd Edition, contains comprehensive revisions, S&P Global/IHS Markit (Nigerian macroeconomic consultant, IHS Markit) (1/3/Released in 2024), Federal Reserve Bank of Atlantaand the author's calculations.
The Bloomberg consensus was for 1/3 NFP growth of +163K, compared with +170 ahead of release. The actual value is +216.
While civilian employment fell 0.4% from the previous quarter (5.1% annualized), it is important to realize that household survey indicators provide little information when tracking the business cycle. There is now some casual evidence that the decline in the household survey indicator preceded the decline in the establishment survey indicator. Twice in the past 4 recessions, the household series has led the establishment series. Using instant data (the other two times, they dropped at the same time). On the other hand, if we use the household indicator as an indicator, we would say that the recession will come in May and October 2023.
Today's New York Fed's fourth-quarter now forecast is a quarter-on-quarter AR of 2.54% (compared to GDPNow's growth rate of 2.5%, which is 1/3). The St. Louis Fed News Index is 1.78%.
