March Revisions Yield Sub-Par U.S. Jobs Report

April was on the mark, but March was weaker than we thought

People check the jobs board at a Denver Workforce Center.

Photographer: Matthew Staver/Bloomberg

At first glance the April employment report appears broadly in line with expectations, as both the consensus forecast for the headline payroll change and unemployment rate were largely on the mark. However, significant downward revisions to March shade the tone of the report to the weak side. Given the drift of market expectations for the Fed liftoff to occur in the latter half of the year, the April jobs report is not likely to significantly alter this view. The source of first quarter weakness is unclear. Those who blamed weather will take comfort in the sizable rebound in construction hiring, but a virtual stall in manufacturing payrolls over the past two months highlights ongoing problems in the factory sector. Factory hours worked continue to fall, as manufacturers deal with a big inventory overhang from the first quarter, which in turn suggests industrial production will again be weak in April. The unemployment rate fell, but a weaker-than-anticipated increase in average hourly earnings kept the year-on-year trend within its post-recession range. Thus, wage inflation remains contained.

• The U.S. economy added 223k nonfarm payrolls in April, pretty much in line with Street expectations of 228k. However, March was revised down to 85k from 126k. So while the consensus forecast was close to the mark, net negative revisions to the prior two months totaled 39k. Sceptics will argue that the “real” payroll change was still sub-par, with net payroll change effectively 184k (223k minus 39k). Furthermore, there now appears to be a consistent pattern of downward revisions, as net revisions were negative in February, March and now again in April. The second quarter is off to a modest start, stronger than the 184k job average during the first three months of the year, but softer than the 324k average registered during the fourth quarter of 2014. The April increase was the 55th consecutive monthly increase in nonfarm payroll.

Screen Shot 2015-05-08 at 9.33.00 AM

• The rebound was broad-based outside of manufacturing and mineral extraction. Private payrolls rose 213k (versus 94k previously) and government jobs rose by 10k (a complete turnaround from the 9k contraction last month). Within private payrolls, goods-producing jobs rose by 31k, driven entirely by a 45k gain in the construction sector. The construction gains are probably weather related, as this sector fell 9k in March. Manufacturing posted a troubling 1k increase following an unchanged March, and mining accelerated its contraction to -15k from -12k. The cumulative decline in mining/mineral extraction is now -48k over the past four months. While the soft patch in manufacturing may yet pass, weakness in the energy sector continues to take a toll on employment. Private services rose by 182k, versus 115k in March. Trade and transportation were the weak spots here -- possibly reflecting the ongoing impact of the stronger dollar -- with transportation decelerating to 24k after holding strong at 43k prior, while wholesale trade contracted by 5k compared to a 10k expansion previously. Retail trade slowed to just 12k compared to 25k previously. Temporary jobs rose 16k following a 13k increase previously, possibly reflecting employers’ reluctance to hire permanent employees in the current growth soft patch.

• The manufacturing side of the report was not encouraging. Not only was manufacturing job growth weak, rising by just 1k over the past two months, but weekly hours in the manufacturing sector continued to slide (40.8 vs. 40.9 previously). Manufacturing overtime fell for the fourth consecutive month. In turn, aggregate hours worked in the manufacturing sector fell for the second month in a row following a flat reading in February. This is a negative omen for factory sector activity at the start of the current quarter, and should temper expectations for a strong rebound in the broader economy in Q2.

• The unemployment rate declined in line with expectations to 5.4 percent. This is consistent with the fact that the insured rate of unemployment, which is part of the weekly jobless claims data, has pushed to cyclical lows over the past several weeks. This is a particularly good decline in the unemployment rate, because unemployment fell by -26k at the same time that the labor force increased, meaning that marginal participants re-entered the labor force. While one month does not make a trend, this will be looked upon favorably by policymakers who are watching to see if there is a substantial cyclical component to the recent trend in labor force participation.

• The U-6 underemployment rate declined to 10.8 percent from 10.9 percent, for the ninth straight monthly decline. That leaves the spread between U-6 and the U-3 official national unemployment rate at 5.4 percent for the second straight month -- still about 1 percentage point above its prior two cyclical peaks. Given the recent bumps in headline job growth, it is encouraging to see progress continue in this measure of slack.

Screen Shot 2015-05-08 at 9.33.59 AM

• The labor force participation rate (LFPR) rose by 0.1 percent to 62.8 percent, from what had been its lowest in decades. This will cast a more positive light on the steadiness of the unemployment rate and the drop in U-6 underemployment. Over the last two months, comments by Fed Chair Yellen and Vice Chair Fischer have implied a persistent difference of opinion over how much recent declines in LFPR can be attributed to structural factors versus cyclical slack that could be absorbed if monetary policy remains accommodative.

• Other key measures of labor market slack in this report also posted modest improvement. Involuntary part-time workers declined, to re-attain their cyclical low of 6.6 million, from 6.7 million last month, which was the first rise in this measure since June 2014. Long-term unemployment’s share of total unemployment fell to 29.5 percent from 29.9. percent -- a new post-recession low. The decline across measures of slack ought to help reassure policymakers that despite weakness in manufacturing, broader non-factory employment retains substantial momentum, consistent with the non-manufacturing ISM survey, low jobless claims, and the NFIB labor indicators.

Screen Shot 2015-05-08 at 9.34.50 AM

• Average hourly earnings (AHE) slowed to a 0.1 percent gain in April after rising by a downwardly revised 0.2 percent in March. That was enough to push the year-on-year rate of change up to 2.2 percent from 2.1 percent, but that still remains within the range established over the past several years. By this metric, wage pressures remain muted.

• Aggregate weekly hours averaged 102.9 in the first quarter, an annualized increase of 2.1 percent from the final quarter of 2014 (102.37). During the first three months of the year, real GDP inched up 0.2 percent, but this is likely to be revised to a modest contraction in light of recent trade figures. April’s aggregate weekly hours worked index was 103.0 -- a 0.2 percent increase from the downwardly revised 0.3 percent decline posted in March. While this is only one data point, the gain is consistent with a positive, albeit soft, reading in second quarter real GDP.

Screen Shot 2015-05-08 at 9.35.37 AM

• Given the drift of market expectations for the Fed liftoff to occur in the latter half of the year, the April jobs report is not likely to significantly alter this view. Of course, with New York Fed President Dudley scheduled to speak at 9:45am this morning, it may not take long to get a sense of policy makers’ response to this data. The April jobs report shows the economy moving forward at a moderate pace despite drags from manufacturing and the energy sector, but wage pressures remain tame. The Fed is inclined to err on the side of caution, so Bloomberg Economics remains of the view that they will wait for more information -- namely the Q2 GDP report and a few more jobs reports before committing to a rate liftoff -- and hence September remains the likely meeting for the initial rate increase.

This post is courtesy of Bloomberg Intelligence Economics.