If there is one thing that can stop the Fed dead in its tracks from hiking rates by 50 bps for the foreseeable future, it's a negative one million (or slightly better) print in monthly payrolls. While we won't get a print that ugly tomorrow, it is almost certain that will get a big drop on Friday morning followed by negative print in the coming months, especially when considering the plunge in new job postings...
... and the mass corporate layoff announcements which have yet to be captured by government data (as we discussed last week in "We Could See A Million Layoffs Or More" - Here Comes The Job Market Shock).
So what should one expect tomorrow? As Newsquawk writes in its May NFP preview, headline measures of payroll additions are expected to continue deteriorating in May. The jobless rate likely fell back to pre-pandemic levels (consensus 3.5%, down from 3.6%), while the participation rate may inch upwards, which would be an encouraging development. However, the focus will be on wages - assuming there is no shockingly bad payrolls print which will steal the show - to see how employment compensation is faring amid surging consumer prices; the expectation is that the annual rates of pay growth eased, but the monthly rate may pick-up relative to April’s data (consensus expects a 0.4% increase in average hourly earnings, up from 0.3%).
While the May jobs data will only have a limited impact on near-term Fed policy, given that the central bank is widely expected to lift rates by 50bps in both June and July, any signs of moderation in wages may support the argument that the Fed could slow or pause the increments of rate rises in September. Of course, a huge negative outlier and all bets are off.
Some more details on what to expect:
- Nonfarm payrolls: consensus expects 325k jobs (Goldman is below consensus at 225K) to be added to the US economy in May (down from April's 428k), as the pace cools relative to recent trend rates (three-month average 523k, six- and 12-month averages both at 552k). Yesterday's surprisingly ugly ADP print, driven by a decline in small-business jobs is not helping. Furthermore, as shown below, there is a non-trivial chance of a negative print tomorrow, in line with what big data indicators suggest.
- Average Hourly Earnings M/M: exp. 0.4%, up from 0.3% in April
- Average Hourly Earnings Y/Y: exp. 5.2%, down from 5.5% in April
- Average Hourly Earnings all employees: exp. 34.6, unch from April
- Unemployment Rate: exp. 3.5%, down from 3.6% in April
- Underemployment Rate: no consensus, was 7.0% in April
- Labor Force Participation Rate: exp. 62.3%, up from 62.2% in April
Initial jobless claims for the week that coincides with the BLS’ survey window for the jobs data showed a pickup to 218k from 185k going into the April report, although continuing claims for the week eased to 1.346mln from 1.403mln. Pantheon Macroeconomics said the trend in claims is probably rising slowly, but still remains at very low levels historically, and a long way from signalling any sort of distress in the labour market.
The unemployment rate is expected to fall by one-tenth of a percentage point to 3.5% in May, a level not seen since before the pandemic in February 2020. It would be even more encouraging if the participation rate also rose from the prior 62.2% in combination with the unemployment rate falling, as some economists expect, as workers continue to return. If the jobless rate fell to the level that the consensus expects, it will be where the FOMC had projected it would fall to by the end of this year, while the central bank also sees unemployment remaining at these levels through 2023 too. That is consistent with its thesis that the labor market is virtually at maximum employment.
Accordingly, it is not the headline payrolls number or the jobless rate that is likely to influence Fed policy in the near- term (absent major shocks in either direction), but wage measures given the central bank’s concerns about runaway inflation and the potential for second round effects. Currently, the market expects the Fed to raise rates by 50bps at both the June and July meetings; what happens after that point in September (does the Fed continue with aggressive larger hikes, move to lower increments of around 25bps, or even pause) is assumed to be a function of how the incoming data on price pressures evolves.
The street is modelling average earnings to increase by 0.4% M/M, which would represent a quicker pace than the +0.3% in April, but in line with the average pace seen over the last year. The annual rate of average hourly earnings, however, is seen easing from 5.5% Y/Y to 5.2%. With core PCE prices – the Fed’s preferred gauge of inflation – running at 4.9% Y/Y in April, wages would (quantitatively at the headline level, at least) offset the diminished purchasing power as a result of persistent inflation (NOTE: headline PCE is running at a rate of 6.3% Y/Y, above the level of annual average hourly earnings).
Given that the May jobs data will only have a limited impact on influencing near-term monetary policy, the debate may refocus onto whether employment compensation gains have peaked, much like the debate with regards to headline inflation, which has been coming off annual peaks recently.
Capital Economics says that the combination of cooling labor demand and a gradual recovery in labor supply suggests that wage growth is close to a peak, with the survey evidence pointing to a stabilization in labor shortages, and says that surveys of compensation plans – like the NFIB's Compensation Plans Index as well as the average hourly earnings – point to a slight decline in pay growth. Even so, a monthly rate of wage growth similar to April would still be well above levels consistent with the Fed’s 2% inflation target, it says.
Arguing for a weaker-than-expected report
- Labor supply constraints. As Goldman shows in the chart below, when the labor market is tight, job growth tends to slow during the spring hiring season, particularly in May. This reflects the combination of a fewer available workers and the high seasonal hurdle - the BLS adjustment factors historically assume over 0.5mn of net seasonal hiring in May. The arrival of the youth summer labor force tends to ease these constraints in June and July, consistent with a peak drag on job growth in the May report. For example, payroll growth slowed to +75k in May 2019 from +263k in April on a first-reported basis.
- Big Data. High-frequency data on the labor market generally indicate weakness in May employment, with the four indicators Goldman track consistent with a below-consensus report. The Census Household Pulse Survey is an outlier to the downside; however, that series has been much more volatile than nonfarm payrolls recently, rising or falling by at least 3mn in each of the previous four months.
- ADP. Private sector employment in the ADP report increased by only 128k in May, far below consensus expectations. Goldman believes that statistical inputs likely weighed on the ADP print, and expects a firmer gain in nonfarm payrolls this month. We think that there is a non-trivial chance of a negative print tomorrow.
- Employer surveys. The employment components of business surveys generally decreased in May. Goldman's services survey employment tracker decreased by 0.3pt to 54.5 and its manufacturing survey employment tracker decreased 0.6pt to 56.1.
Arguing for a stronger-than-expected report
- Reopening. While covid infections have risen further, hospitalizations remain near the lows of last summer, and domestic business activity has not been significantly impacted. For example, as shown below, dining activity has been relatively stable at pre-pandemic levels since March. Our forecast assumes a roughly 50k rise in leisure-sector payrolls in tomorrow’s report (mom sa).
- Evolution of the seasonal factors. The May seasonal factors have evolved favorably in recent years, with a hurdle of 499k in May 2021 compared to the 605k May average for 2017-19 (mom basis). This lower seasonal hurdle represents a tailwind of roughly 100k, other things equal. However, May seasonality remains a net negative factor due to binding constraints on labor supply during a rehiring month.
- Education seasonality. Education payrolls remains 0.4mn below pre-pandemic levels—in part due to labor supply constraints and related shortages of janitors and support staff. Many of these support staff typically stop working for the May survey period, implying a seasonally adjusted gain in education payrolls in tomorrow’s report (GS assumes +75k, public and private). For reference, education payrolls rose 112k in May 2021.
- Job cuts. Announced layoffs reported by Challenger, Gray & Christmas decreased by 32% month-over-month in May, after increasing by 6% in April (SA by GS)
- Job availability. The Conference Board labor differential—the difference between the percent of respondents saying jobs are plentiful and those saying jobs are hard to get—decreased by 5.4pt to +39.3. JOLTS job openings decreased by 455k in April to 11.4mn. Despite the sequential weakness, the level of labor demand remains very high, but these prints are dated. As shown above, real-time indicators shows a record plunge in new job postings.
- Jobless claims. Initial jobless claims increased from very low levels during the May payroll month, averaging 200k per week vs. 175k in April. Continuing claims in regular state programs decreased 60k from survey week to survey week.
Finally, and perhaps the clearest indication that tomorrow's report will be a big disappointment, comes from the White House itself, which said that it is not expecting to see big increases in jobs data every month as the economy transitions to a new period of more stable growth.
According to White House press secretary Karine Jean-Pierre, “as we transition to this new period of stable, steady growth, we aren’t looking to see blockbuster job reports month after month, like we have over the last year,” she said. “But that’s a good thing. That’s the sign of a healthy economy with steady job growth, rising wages for working America, everyday costs easing up and a shrinking deficit.”