Goldman's Take On NFP

Goldman's Hatzius, whose +100,000 NFP revision to 600,000 yesterday, officially means he has lost the magic touch, offers his take on the NFP disaster. Ever the optimist, he still manages to find a silver lining. Alas, the market doesn't care, and instead is holding its breath to the joint Obama-Biden teleprompter conference praising the -215k adjusted NFP number.

Weaker than Expected, a Mirror Image of April Results

BOTTOM LINE: In several respects, the May employment report is a mirror image of the April report: skinny and disappointing private-sector payroll gains, a drop in the jobless rate due to labor force exodus, and larger-than-expected wage gains. Manufacturing data strong in May; data on hours worked still support firm GDP growth in Q2, thanks to longer workweeks and large prior gains in payrolls.

Nonfarm payrolls +431k in May vs. GS +600k, median forecast +536k.
Private payrolls +41k in May vs. GS +150k, median forecast +180k.
Unemployment rate 9.7% in May vs. GS 9.7%, median forecast 9.8%.
Average hourly earnings +0.3% in May (mom, +1.9% yoy) vs. GS and median forecast +0.1%.


1. As widely expected, the hiring of temporary Census workers dominated the payroll gains reported for May, as a net increase of 411k in these workers drove overall government payrolls up 390k. In the private sector, job gains were a skinny 41k, and 31k of that was offset by downward revisions to March and April. Since September most revisions (to total payrolls) have been higher, so this is a cautionary sign.

2. The household survey told a similar story: its measure of employment was down 35k in May (though up 407k on a payroll concept adjusted basis). Although the unemployment rate dropped back to 9.7% in May, it did so because of a 322k drop in the labor force. Broader measures of underemployment also fell in May; the "U6" measure that incorporate those working part-time for economic reasons and marginally attached workers fell ½ point to 16.6%.

3. In many respects, these data mirror opposite moves - at least relative to expectations - in the April report. Thus, some averaging is warranted. For payrolls, this implies private-sector hiring of about 130k; for unemployment we see no meaningful change; for wages, it implies a 0.2% average monthly increase, as wages were up 0.3% in May vs. a 0.1% (upward-revised) April increase.

4. The one sector that did well in May was manufacturing. Payrolls were up 29k (0.2%), and the workweek lengthened by 18 minutes to 40.5 hours (+0.7%). So the report suggests another large increase in industrial production. Elsewhere, job gains were mostly thin but down in construction (35k), retail (7k), and finance (12k). In an encouraging sign of future growth, temporary help continued to show increases (31k).

5. With overall workweeks also longer, though only by 6 minutes, the index of hours worked in the private sector rose 0.3%. Given the larger gains reported for prior months, this is tracking at a 4% annual rate, suggesting potential upside to our 3% estimate for annualized real GDP growth for Q2.