US state payrolls numbers confirm the cooling jobs narrative
Headline non-farm payrolls were significantly weaker than expected in October. Today’s state data show that this wasn’t entirely due to temporary strike- and hurricane-related factors and that there is a broader cooldown in play. This suggests to us that a December Federal Reserve interest rate cut remains more likely than not
State jobs numbers indicate a broad cooling
Admittedly, this is a little bit niche at this stage, but I suspect it will become an important story over the next couple of weeks... The US state non-farm payrolls numbers have been released today and give us a guide as to how impactful Hurricane Milton was in depressing the national payrolls figure from October – remember that was 12k versus the 100k consensus with 112k of downward revisions to the previous couple of months, which prompted a 10bp fall in the US 10Y yield and confirmed the Fed's 25bp rate cut decision on 7 November.
The hurricane made landfall on Wednesday, 9 October, in the middle of the data collection week for the jobs report. This was two weeks after Hurricane Helene had hit Florida, Georgia, the Carolinas and others. Many people in the impacted region were unable work and receive pay the week of 12 October and would therefore not have been included in the payrolls number. Today’s numbers show that Florida employment fell 38k in October relative to its six-month trend growth of 13k. This allows us to approximate that the hurricane effect depressed Florida payrolls by around 51k.
There is the hint of some minor impacts in the Carolinas and Virginia so we can perhaps round that to a 60-65k impact from the hurricane. We also know that strike activity, predominantly at Boeing, subtracted 44k so if we add the hurricane and strike activity to the 12k number that was published as the official non-farm payrolls number for October, we get a figure of 121k (12k+44k+65k).
At first glance you may say that is above the 100k consensus forecast at the time, but that is misleading as economists factored in hurricane and strike activity into their forecasts. Economists also didn't expect the 112k of downward revisions to August and September payrolls – so the October jobs report was an unambiguously weak report.
Moreover, the data shows 29 states reported negative payrolls growth in October and we certainly didn't have 29 states impacted by the hurricane and strikes. Only 14 states reported a negative monthly change in payrolls in September so the October state data hints at broader weakness.
Bias towards a December rate cut
What today's data means is that we have to have 109k as a base for December before we consider any actually payrolls growth – because the 44k striking workers have now returned to work and the approximately 65k people who weren't counted in October because of the hurricane will be there in the November data.
The current consensus for the 6 December jobs report is for payrolls to increase 175k. It is incredibly early and there are barely a handful of submissions so far, but given 109k of this is the technical rebound, it implies "true" payrolls growth of just 66k (175k-109k). If that is correct the Federal Reserve is likely to cut interest rates again in December.
Personally I am looking at something of the order of 200-225k for the headline figure, which would give a "true" payrolls increase of just over 100k, which I think would still be enough for the Fed to cut rates at the December FOMC meeting before signalling a likely pause in January. Something above a "true" 175k, implying a headline of 284k would be when the Fed says pause at December. Anywhere in between would leave it as a coin toss with the November CPI report then the focus, released on 11 December.