September Payrolls Preview: It Will Be A Beat, The Question Is How Big

Tyler Durden's Photo
by Tyler Durden
Friday, Oct 08, 2021 - 06:15 AM

After a strong initial claims report and a solid ADP private payrolls print, all eyes turn to the most important economic data point of the week, and the month, Friday's nonfarm payrolls report due at 830am ET on Friday, where consensus expects a 500K print- more than double last month's disappointing 235K print - as well as a drop in the unemployment rate to 5.1% and an increase in average hourly earnings to 4.6%. And unlike last month, when we correctly predicted the big miss in August payrolls, this time we agree that tomorrow's report will be a beat, the only question is how big.

Here is a snapshot of what to expect tomorrow:

  • Total Payrolls: 500K, Last 235K
  • Private Payrolls: 450K, Last 243K
  • Unemployment Rate: 5.1%, Last 5.2%
  • Labor force participation rate: 61.8%, Last 61.7%
  • Average Hourly Earnings Y/Y: 4.6%, Last 4.3%
    • Average Hourly Earnings M/M: 0.4%, Last 0.6%
  • Average Weekly Hours: 34.7, Last 34.7

As Newsquawk writes in its NFP preview, September’s jobs data, the last before the Fed’s November 3rd policy meeting, will be framed in the context of the central bank’s expected taper announcement, where a merely satisfactory report would likely to be enough for the FOMC to greenlight a November announcement to scale-back its USD 120BN/month asset purchases.

Goldman economists are more bullish than normal, and estimate nonfarm payrolls rose 600k in September, above consensus of +500k, and they note that "labor demand remains very strong, and we believe the nationwide expiration of enhanced unemployment benefits on September 5 boosted effective labor supply and job growth—as it did in July and August in states that ended federal benefits early." As a result, Goldman is assuming a 200k boost in tomorrow’s numbers and a larger boost in October. The bank also believes the reopening of schools contributed to September job growth, by around 150k. Despite these tailwinds, Big Data employment signals were mixed, and dining activity rebounded only marginally.

Labor market proxies have been constructive for the month: ADP’s gauge of payrolls surprised to the upside, although analysts continue to note that the direct relationship between the official data and the ADP’s gauge is tenuous, despite the gap being under 100k over the last three reports. The number of initial jobless claims and continuing claims has eased back between the survey periods of the August and September jobs data, although analysts note that more recent releases have shown an uptick in claims potentially clouding the outlook. The ISM business surveys have signaled employment growth in the month, with manufacturing employment rising into growth territory again, but services sector hiring cooled a little in the month, but remains expansionary; survey commentary continues to allude to a tight labour market. The Bureau of Labor Statistics will release the September employment situation report at 13:30BST/08:30EDT on October 8th.

POLICY: The September jobs report might have reduced relevance on trading conditions given that Fed officials have effectively confirmed that, barring a collapse in the jobs data, it is on course to announce a tapering of its asset purchases at the November 3rd meeting. Accordingly, trading risks may be skewed to the downside, rather than to the upside, where a significant payrolls miss may present obstacles to the Fed announcing its taper. Additionally, it is worth being cognizant of how efforts in Washington to raise the debt ceiling are progressing; as yet, officials have not struck a deal, and are in the process of enacting stop gap legislation to allow funding into December; some analysts suggest that the Fed may be reticent to tighten policy in the face of potential default risks.

PAYROLLS: The consensus looks for 500k nonfarm payrolls to be added to the US economy in September (prev. 235k), which would be a cooler rate of growth than the three- and six-month average rate, though in line with the 12-month average (3-month average is 750k/month, the six-month average is 653k/month, and the 12-month average is 503k/month – that technically at least suggests an improving rate of payrolls growth in recent months). Aggregating the nonfarm payrolls data since March 2020, around 5.33mln Americans remain out of work relative to pre-pandemic levels.

MEASURES OF SLACK: The Unemployment Rate is expected at 5.1% (prev. 5.2%); Labour Force Participation previously at 61.7% vs 63.2% pre-pandemic; U6 measure of underemployment was previously at 8.8% vs 7.0% prepandemic; Employment-population ratio was previously 58.5% vs 61.1% pre-pandemic. These measures of slack are likely to provide more insight into how Fed officials are judging labour market progress, with many in recent months noting that they are closely watching the Underemployment Rate, Participation Rate, and the Employment-Population Ratio for a better handle on the level of slack that remains in the economy. Analysts would be encouraged the closer these get to pre-pandemic levels.

EARNINGS: Average Hourly Earnings expected at +0.4% M/M (prev. +0.6%); Average Hourly Earnings expected at +4. 6% Y/Y (prev. +4.3%); Average Workweek Hours expected at 34.7hrs (prev. 34.7hrs). Aggregating the nonfarm payrolls data since March 2020, around 5.33mln Americans still remain out of work relative to pre-pandemic levels.

ADP: The ADP National Employment Report showed 568k jobs added to the US economy in September, topping expectations for 428k, and a better pace than the prior 340k (revised down from 374k initially reported). ADP itself said that the labor market recovery continued to make progress despite the marked slowdown in the rate of job additions from the 748k pace seen in Q2. It also noted that Leisure & Hospitality remained one of the biggest beneficiaries to the recovery, though said that hiring was still heavily impacted by the trajectory of the pandemic, especially for small firms. ADP thinks that the current bottlenecks in hiring will likely fade as the pandemic situation continues to improve, and that could set the stage for solid job gains in the months ahead. On the data methodology, analysts continue to note that ADP's model incorporates much of the prior official payrolls data, other macroeconomic variables, as well as data from its own payrolls platform; "Payrolls were soft in August, thanks to the hit to the services sector from the Delta variant, and that weakness likely constrained ADP data," Pantheon Macroeconomics said. "The overshoot to consensus, therefore, suggests that the other inputs to ADP’s model were stronger than we expected, but none of the details are published, so we don’t know if the overshoot was model-driven or due to stronger employment data at ADP’s clients."

INITIAL JOBLESS CLAIMS: Initial jobless claims data for the week that coincides with the BLS jobs report survey window saw claims at around 351k – little changed from the 349k for the August jobs data survey window – where analysts said seasonal factors played a role in boosting the weekly data, while there may have been some lingering Hurricane Ida effects; the corresponding continuing claims data has fallen to 2.802mln in the September survey period vs 2.908mln in the August survey period. In aggregate, the data continues to point to declining trend, although in recent weeks the level of jobless claims has been picking up again.

BUSINESS SURVEYS: The Services and Manufacturing ISM reports showed divergent trends in September, with the service sector employment sub-index easing a little to 53.0 from 53.7, signalling growth but at a slower rate, while the manufacturing employment sub-index rose back into expansionary territory, printing 50.2 from 49.0 prior. On the manufacturing sector, ISM said companies were still struggling to meet labour-management plans, but noted some modest signs of progress compared to previous months: "Less than 5% of comments noted improvements regarding employment, compared to none in August," it said, "an overwhelming majority of panelists indicate their companies are hiring or attempting to hire," where around 85% of responses were about seeking additional staffing, while nearly half of the respondents expressed difficulty in filling positions, an increase from August. "The increasing frequency of comments on turnover rates and retirements continued a trend that began in August," ISM said. Meanwhile, in the services sector, employment activity rose for a third straight month; respondents noted that employees were flocking to better-paying jobs and there was a lack of pipeline to replace these staff, while other respondents talked of labor shortages being experienced at all levels.


  • End of federal enhanced unemployment benefits. The expiration of federal benefits in some states boosted labor supply and job-finding rates over the summer, and all remaining such programs expired on September 5. The July and August indicated a cumulative 6pp boost to job-finding probabilities from June to August for workers losing $300 top-up payments and a 12pp boost for workers losing all benefits. Some of the 6mn workers who lost some or all benefits on September 5 got a job by September 18—in time to be counted in tomorrow’s data. Goldman assumes a +200k boost to job growth from this channel, with a larger increase in subsequent reports (+1.3mn cumulatively by year end).

  • School reopening. The largest 100 school districts are all open for in-person learning, catalyzing the return of many previously furloughed teachers and support staff. While full normalization of employment levels would contribute 600k jobs (mom sa, see left panel of the chart below), some janitors and support staff did not return due to hybrid teaching models, and job openings in the sector are only 200k above the pre-crisis level (see right panel). Relatedly, the BLS’s seasonal factors already embed the usual rehiring of education workers on summer layoff, so if fewer janitors returned to work than in a typical September, this would reduce seasonally adjusted job growth, other things equal. Taken together, assume a roughly 150k boost from the reopening of schools in tomorrow’s report.

  • Job availability. The Conference Board labor differential—the difference between the percent of respondents saying jobs are plentiful and those saying jobs are hardto get - edged down to 42.5 from 44.4, still an elevated level. Additionally, JOLTS job openings increased by 749k in July to a new record high of 10.9mn.
  • ADP. Private sector employment in the ADP report increased by 568k in September, above consensus expectations for a 430k gain, implying strong growth in the underlying ADP sample. Additionally, schools generally do not use ADP payroll software, arguing for a larger gain from school reopening in the official payroll measure.


  • Delta variant. Rebounding covid infection rates weighed on services consumption and the labor market in August. And while US case counts began to decline in early September, restaurant seatings on Open Table rebounded only marginally. leisure and hospitality employment rose in September, but probably not at the ~400k monthly pace of June and July.
  • Employer surveys. The employment components of our business surveys were flat to down, whereas we and consensus forecast a pickup in job growth. Goldman's services survey employment tracker remained unchanged at 54.5 and the manufacturing survey employment tracker declined 0.4pt to 57.8. And while the Goldman Sachs Analyst Index (GSAI) decreased 0.8% to 68.5, the employment component rose1.9% to 71.9.


  • Big Data. High-frequency data on the labor market were mixed between the August and September survey weeks, on net providing little guidance about the underlying pace of job growth. Three of the five measures tracked indicate an at-or-above-consensus gain (Census Small Business Pulse +0.5mn, ADP +0.6mn,Google mobility +2mn), but the Homebase data was an outlier to the downside. At face value, it would indicate a large outright decline in payrolls. The Census Household Pulse (-0.6mn) was also quite weak, though encouragingly, it also indicated a large drop in childcare-related labor supply headwinds as schools reopened.
  • Seasonality. The September seasonal hurdle is relatively low: the BLS adjustment factors generally assume a 600-700k decline in private payrolls (which exclude public schools), compared to around -100k on average in July and August. Continued labor shortages encouraged firms to lay off fewer workers at the end of summer. Partially offsetting this tailwind, the September seasonal factors may have evolved unfavorably due to the crisis—specifically by fitting to last September’s reopening-driven job surge (private payrolls +932k mom sa).
  • Jobless claims. Initial jobless claims fell during the September payroll month, averaging 339k per week vs. 378k in August despite a boost from individuals transitioning or attempting to transition to state programs. Across all employee programs including emergency benefits, continuing claims fell dramatically (-3.3mn)–but again for non-economic reasons (federal enhanced programs expired). Continuing claims in regular state programs decreased 106k from survey week to survey week.
  • Job cuts. Announced layoffs reported by Challenger, Gray & Christmas rebounded 11% month-over-month in September after decreasing by 14% over the prior two months (SA by GS). Nonetheless, layoffs remain near the three-decade low on this measure (in 1993).