254,000 jobs, despite the hurricanes

Non-farm payrolls rose by 254,000 in September, far above forecasts. Earlier months were revised up by 72,000 and the unemployment rate fell to 4.1 per cent. Economists had expected regional hurricane disruption to drag the figure down. It did not.

Why the storms barely registered

The way the figures are collected explains much of the surprise. The payroll survey asks employers how many people they paid during the pay period that includes the 12th of the month. A worker counts as employed if paid for any part of that period, so a storm has to keep someone off the payroll for the whole of it before it shows up as a lost job. Short closures, lost hours and delayed shifts rarely move the headline.

Timing mattered too. Hurricane Helene came ashore in Florida in the last week of September, well after the survey's reference week, so its effect on hiring will appear in October's figures, not these. Storms also hit response rates, which are already low in the first estimate each month. That is one reason weather-affected months tend to be revised more heavily later, in either direction.

No slowdown yet

The size of the gain wrecked the view that the US economy was sliding into a cyclical slowdown. Bond investors sold duration hard and cut back their bets on another 50-basis-point cut from the Federal Reserve. An economy adding more than a quarter of a million jobs a month hardly needs emergency stimulus.

The detail was as firm as the headline. The unemployment rate has now fallen for two months running after peaking at 4.3 per cent in July, the reading that tripped the Sahm rule and helped push the Fed into its half-point cut in September. Average hourly earnings rose faster than expected, which is not what a labour market on the edge of recession usually produces. Hiring was spread across services, with restaurants, health care and government all adding workers.

A noisy series, read carefully

Caution is still warranted. The payroll survey is a sample, and its monthly change has a margin of error of more than 100,000 jobs at conventional confidence levels, so a single strong print says less than it seems. The Bureau of Labor Statistics has also warned that job growth in the year to March was overstated by 818,000, a revision large enough to show how far the monthly figures can drift from the complete count based on unemployment insurance records.

The household survey, which produces the unemployment rate, is noisier still, and it has diverged from the payroll count for much of the past year. When the two tell the same story, as they did in September, the signal is stronger. A three-month average smooths some of the noise, and after this week's revisions it too has moved up, which weakens the case that July's weak report marked a turn in the cycle.

What the market repriced

Short-dated Treasuries took the brunt. The two-year yield, which tracks expectations for Fed policy over the next couple of years, rose sharply on the day, and the ten-year yield climbed back towards four per cent. Futures markets that had priced a meaningful chance of another half-point cut in November swung to pricing a quarter-point move, and some traders began to entertain the possibility of a pause.

The dollar strengthened as rate expectations rose, reversing part of its slide since the summer. For borrowers, the repricing means that mortgage rates, which follow longer-dated yields, are likely to drift up rather than down in the weeks ahead. Equity investors took the report more calmly, since stronger growth supports profits even if it delays cheaper money.

The winners and losers from a slower easing path are easy to identify. Savers in money-market funds keep higher yields for longer, while small firms and households with floating-rate debt wait longer for relief. Firmer wage growth helps workers but keeps services inflation, the stickiest part of the price index, under upward pressure, since labour is the main cost for most service businesses. That is the channel through which a hot labour market would eventually worry the Fed.

The case that the Fed still has work to do

The counterargument is that one month does not make a trend, and the Fed moves on forecasts. Officials cut in September because policy was restrictive, not because the economy was collapsing, and that logic still holds: with inflation near target, a policy rate between 4.75 and 5 per cent remains well above most estimates of neutral. A strong jobs report changes the pace of easing, not its direction.

History offers some comfort for that view. In 1995 and again in 1998 the Fed made three insurance cuts, after which the economy reaccelerated. In the first case officials simply stopped and held rates steady for more than a year; in the second they reversed course and began raising rates in the summer of 1999. Neither episode required the Fed to keep cutting once the labour market showed renewed strength, and neither ended in recession.

That argument supports continued quarter-point cuts, but not a larger one. The distinction between insurance and rescue, which the Fed itself has drawn, now points firmly towards insurance.

The next readings are compromised

October's data will be harder to read. The payroll report due early next month will carry the effects of Hurricane Helene in the south-east and of the strike by machinists at Boeing, which began on 13 September, just after this month's reference week, and took about 33,000 workers off the payroll. A weak October figure will therefore say little about underlying demand, and officials meeting just after the presidential election will have to look through it.

The case for another half-point cut has gone, for now. Bond traders have priced a slower path of easing, and quarter-point moves at the next few meetings look like the most on offer. Unless October's distorted figures hide a genuine collapse in hiring, the Fed will cut by 25 basis points in November and debate a pause by the new year.