The headline non-farm payrolls jobs figure surprised on the downside in April, with 266k jobs created versus the median forecast of 978k. And the March figure was also revised downward by 146k. Obviously, it was private payrolls that disappointed (+218k vs. 893k survey) while manufacturing jobs contracted slightly (-18k). Government couldn’t make up for that but did offer a +48k contribution. The unemployment rate actually ticked back to 6.1%, a result of the 0.2% pt improvement in participation and the work week ticked up again.
February Non-farm payrolls recorded a 379k gain, comfortably exceeding the +182k median forecast. This leaves the 3m average at +80k. Private sector payrolls were significantly better than the headline print at +465k, manufacturing added +21k jobs while government dropped -86k. The surprise might have been even higher were it not for the weather events in Texas, which was behind the 939k people not at work due to the weather (Feb average is 385k). The unemployment rate ticked down to 6.2% and participation was unchanged at 61.4% despite this. The workweek did drop though, to 34.6, although this is from a rather elevated 34.9.
December Non-farm payrolls recorded a -140k drop, quite a bit below the markets +71k guess. This takes the 3m average down to +283k. Private sector payrolls were a little better than the headline print at -95k, helped by manufacturing which added +38k jobs while the decline in government moderated to -45k. That was still the fourth consecutive decline but is largely census related. Despite the headline miss revisions were very positive, +91k in November alone
November Non-farm payrolls number came in at +245k (mkt +469k), bringing the 3m average down to +5227k. Private sector payrolls were a little better again at +344k but there is a sequential slowdown happening there too. Manufacturing payrolls grew by 27k while we saw another month of government job losses, -99k, the third consecutive drop. Revisions overall were down a bit, Oct NFP revised down to 610k vs. the 638k initially reported.
October Non-farm payrolls number came in at +638k (mkt +600k), leaving the 3m average at +934k. But the private payrolls number comfortably bettered expectations once again at +906k vs. 690k mkt. Manufacturing payrolls grew but not as much as hoped (+38k vs. +50k est) while we saw another month of government job losses, -268k following on from -220k in Sep. Revisions overall were slightly positive though
September non-farm payrolls number came in under expectations at +661k (mkt +850k). But the private payrolls number bettered at +877k vs. 850k mkt, while manufacturing payrolls also beat (+66k vs. +35k est) with a loss of 216k on the government side explaining the headline miss. There was a sizable upward revision to the jobs gained in August too.
ADP reporting a further strong month of job gains (or jobs recovery) with the749k increase and net 53k upward revision to August quite a bit ahead of expectations, although consensus estimates should be taken with more of a pinch of salt than usual given the variables in play. It was mid-sized firms that drove the increase, alongside small businesses. There was actually a deceleration in job creation at larger firms, although they still added workers.
Another solid non-farm payrolls report, with 1.763mn jobs added in July, taking the 3-month gain to 9.27mn. That means the US has now recovered roughly 40% of the jobs shed during the turmoil of March and April. The report cements confidence that the labour market continues to repair, mirroring the conclusions of yesterday’s weekly jobless claims, if not the lower than expected ADP print we got on Wednesday.
The disruption from the Coronavirus continues to impact the data with the employment numbers for May showing a significant deviation from expectations and the near 2mn loss reported in the May ADP report. Indeed, based on the preliminary numbers the economy added 2.51mn jobs over the month, which compares to the 20.69mn jobs lost in April. Methodological problems are again in evidence ...
Bleak payrolls report. While the median survey was always going to be a short in the dark, the -701k decline in jobs was significantly more troubling than the mkts -100k guess. That’s the worst number since March 2009, a period of losses that saw five consecutive payrolls figures below the -700k mark starting from November 2008. Looking at the past fortnights jobless claims figures the April number and revisions to this release will paint an even bleaker picture of the labour market. In the household survey nearly 2.987mn jobs were recorded lost.
ADP reported fewer than expected job losses for March, with a modest -24k drop (mkt -150k), which compares to the 3.3mn jump in initial jobless claims last week. Partly this is due to the survey dates (it tallies responses up to the 12th of the month). So we will see catch up next month. Prior month was revised down to 179k from 183k.
The rapid and devastating effect of Covid-19 on the labour market is unprecedented based on this week’s jobless claims. Initial claims in week ending came in at 3,283k which looks like a data error rather than an actual real data point. But real it is, and further increases look likely in the coming weeks as the economic stop continues to push people out of employment.
Another positive surprise on the wages front with a strong pick up in the headline data. The rebound in weekly wage growth was partly due to a more favourable hours worked comparison though. Hourly wage growth has been steadier. Alongside steady inflation means real income growth has improved, continuing the recovery we’ve seen in recent months.