The last two months have been nothing if not a lesson in the disater that is the economic-forecasters of the world. With a 3-sigma beat followed by a 5-sigma miss, hope abounds that April will be the 'goldilocks' print - just cold enough to leave the Fed on hold and just hot enough to 'prove' growth remains. Goldman expects nonfarm payroll job growth of 230k in April, in line with consensus expectations. While labor market indicators were mixed in April, the employment components of service sector surveys were strong and better weather conditions should provide a boost. In addition, they see some upside risk to the forecast from a calendar effect, and expect the unemployment rate to decline by one-tenth to 5.4% and average hourly earnings to rise 0.2%.
Not a great track record the last few months... extreme outlier prints...
[12]
Via Goldman Sachs' Jan Hatzius,
We forecast nonfarm payroll job growth of 230k in April, in line with consensus expectations. While not all major labor market indicators improved in April, jobless claims fell, the employment components of most service sector surveys improved, and weather conditions were more favorable. If realized, we would view a 230k gain in April as a strong print in light of last month's soft gain of 126k, the broader weakness seen across major economic indicators, and our expectation that the trend rate of payroll gains is likely to be closer to 200k in coming months than the 261k average seen over the last year. We also see some upside risk to our baseline forecast from a calendar effect that should be unusually strong this month.
Arguing for a stronger report:
Service sector surveys. Service sector surveys were encouraging in April. The employment component of the ISM nonmanufacturing index rose just a touch, but remained at a high level (+0.1pt to 56.7), and the employment components of the Markit services, Richmond Fed (+3pt to 9), and Dallas Fed (+3.7pt to 7.9) service sector surveys all rose, while the NY Fed survey's employment component declined a bit to a still-high level (-1.3pt to 15.7). Service-sector employment gains slowed to 142k in March and averaged 215k over the last year.
Jobless claims. The four-week moving average of initial jobless claims leading into the payrolls reference week fell 22k to 283k. Weekly claims remained high in Texas over the last month, likely reflecting the hit to energy-sector employment and possible spillover effects.
Weather. Based on our method for estimating the payrolls effect of deviations in weather conditions from seasonal norms, we expect that warmer temperatures should be a substantial positive factor for April payrolls. The weeks leading into the March reference week were much colder than usual, while the weeks leading into the April reference period were quite a bit warmer than usual. Furthermore, the change in employment in both the construction and leisure and hospitality industries was about 30k below trend in March. According to recent Fed research, these industries are among the most sensitive to weather fluctuations.
Arguing for a weaker report:
Manufacturing employment indicators. The employment components of the major manufacturing surveys were mixed in April, but overall manufacturing employment appears likely to remain soft. The employment components of the ISM (-1.7pt to 48.3), Empire (-9pt to 9.6), Kansas City Fed (-16pt to -18), and Markit manufacturing surveys declined, while the employment components of the Chicago PMI (+3.7pt to 54), Philly Fed (+8pt to 11.5), Richmond Fed (+1pt to 7), and Dallas Fed (+3.6pt to 1.8) surveys rose. Payroll employment growth in the manufacturing sector fell 1k in March, below the average gain of 16k seen over the last year. The manufacturing sector is more exposed to international trade than the services sector, and much of the recent weakness seen across manufacturing indicators likely reflects drag from the strong dollar. However, it is important to bear in mind that while there are a large number of manufacturing surveys available, manufacturing has accounted for only 5-10% of employment gains in recent years.
Job availability. The Conference Board's labor differential—the net percent of households reporting jobs are plentiful vs. hard to get—worsened by 2.8pt to -7.3 in April. The index had improved sharply through January, but has declined a bit since then.
Job cuts. Announced jobs cuts reported by the Challenger, Gray and Christmas report--which tend to be an early indicator of actual layoffs--jumped in April on a seasonally-adjusted basis. The increase was led by a roughly 18k jump in energy-sector job cuts, which returned to the elevated level seen in January and February. Oil and gas sector payroll employment has fallen by about 10k per month over the last two months, and the job cuts data point to further declines. Job cuts were also elevated in the chemical and consumer products industries.
Neutral factors:
ADP report. The April ADP employment gain of 169k was below consensus expectations and was the softest gain since January 2014. However, we suspect that this largely reflects the influence of the much weaker March non-farm payrolls report (the weak January 2014 ADP print also followed a much softer payrolls report). ADP uses outside data, including the lagged values of BLS estimates of employment growth, to produce its monthly estimates. In general, initial print ADP estimates have not been strong predictors of initial print payroll gains reported by the Labor Department.
Online job ads. According to the Conference Board's Help Wanted Online (HWOL) report, new online job ads held roughly steady in April, though total ads fell a bit from the high level seen in recent months.
We also see some upside risk to our baseline forecast from a calendar effect. There were five rather than four weeks between the reference weeks for the March and April payrolls surveys this year. While the relevant history is limited and the BLS's seasonal adjustment procedure does attempt to control for the four-versus-five week effect, we have found that five-week Aprils tend to see above-trend payroll growth. Moreover, we have found that this calendar effect is not fully captured by other employment-related data. We expect an unusually large and positive calendar effect this month.
We expect the unemployment rate to decline one-tenth to 5.4% in April, from an unrounded 5.465% in March. The headline unemployment rate declined by 1.1pp over the last year and the broader U6 underemployment rate declined by 1.7pp.
We expect a 0.2% increase in average hourly earnings in April, roughly in line with the recent trend. Average hourly earnings for all workers rose 2.1% over the year ending in March, in line with our Wage Tracker--an aggregate of the Employment Cost Index, average hourly earnings, and nonfarm compensation per hour--which also rose 2.1% over the year ending 2015Q1. We expect wage growth to pick up to around 2.75% by year-end, still below the 3-4% rate of wage growth that Fed Chair Janet Yellen has identified as normal.
