We had not expected much from the August employment report, and we got even less—zero net jobs created. The outcome is not quite as miserable after extracting the temporary effect of the Verizon strike, but perhaps the worst news is that the August report is not that much different from the previous three. Without the effect of the Verizon strike, payrolls would have risen 45,000 last month. The average payroll gain over the previous three months was 53,000.
Excluding the Verizon effect, the last three months (including August) show average job creation of 50,000, with private-sector jobs up an average 98,000 and government jobs down an average 48,000.
The problem does not seem to be an upsurge in firings, since initial unemployment claims have not climbed, but a lack of hiring due to a lack of growth in demand and huge policy uncertainty. The extreme uncertainty over the outcome of the debt-ceiling debate probably did extra damage to the August figures.
In the payroll details, manufacturing lost 3,000 jobs, the first drop since October 2010. Primary and fabricated metals had been adding jobs steadily in previous months, but turned negative in August. Overall manufacturing production-worker hours fell 0.1%, also the first drop since October 2010, pointing to either a drop or at best a very small increase in manufacturing output during August. Construction lost 5,000 jobs, with residential payrolls up slightly but nonresidential jobs down.
Private services employment growth slowed to 20,000, from 104,000 in July, although without the Verizon effect that led to 48,000 drop in information jobs, the increase would have been 65,000. The big swing was in retail, where 8,000 jobs were lost (concentrated in electronics and appliance stores), after 26,000 had been added in July. Leisure and hospitality added only 2,000 jobs, after adding 12,000 in July. Professional and business services (up 28,000) and healthcare (up 30,000) both showed exactly the same increase as in July. Temporary help edged up 5,000. Financial services added 3,000 jobs, but we expect declines here in coming months.
The government sector shed 17,000 jobs. Federal employment fell by 2,000. State and local employment fell "only" 15,000, but the decline here would have been steeper but for the return of 22,000 workers previously furloughed in Minnesota. The main news in government employment was the big downward revisions to June and July. If we exclude the Minnesota effect, state and local employment now shows steady, steep declines of 39,000 in June, 44,000 in July, and 37,000 in August. The biggest declines are in local education.
The private workweek shortened from 34.3 to 34.2 hours. A shorter workweek combined with flat private employment to generate a 0.2% decline in hours worked. Hours rose quite sharply in the second quarter, by 3.5%, but the figures for July and August suggest that hours will probably fall slightly in the third quarter.
Average hourly earnings fell 0.1%, and were up 1.9% year-on-year. We do not attach too much significance to the drop in earnings, since it followed a surprising jump of 0.5% in July. It is better to focus on the year-on-year earnings increase, which was 1.9%, well below the 3.6% CPI inflation rate for July, meaning declining real purchasing power for wage earners. Overall payrolls (wages multiplied by hours) fell 0.4%, reversing most of July's 0.6% increase.
The unemployment rate was steady at 9.1%. That is odd given the lack of payroll jobs, but the household survey showed a huge monthly increase of 331,000 jobs, roughly matched by a 366,000 increase in the labor force. The participation rate—the proportion of the working-age population either in employment or looking for work—edged up from 63.9% to 64.0%. It makes no sense to try to analyze why household employment rose 331,000—just as it makes no sense to analyze why it dropped 445,000 in June. The household survey is much smaller than the payroll survey and consequently its jobs estimates are much more volatile.
The most comprehensive measure of underemployment (U-6)—which includes workers who would like a job but are not currently looking, plus those working part time who would rather work full time—edged up from 16.1% to 16.2%. There was a sharp increase this month (430,000) in those working part time for economic reasons (consistent with the decline in the workweek in the payroll figures). The proportion of long-term unemployed (27 weeks or longer) remained elevated, at 42.9%. The longer that potential workers remain either unemployed or on the sidelines outside the labor force entirely, the less likely that they will ever get back into employment. Recent trends are perilously close to stall speed for the economy. They do not yet say that we have tipped into recession, but that risk remains high (40% odds, in our view). They increase the pressure on policymakers to do something more to help.
The Federal Reserve will consider its options at its extended September 20–21 meeting. Even though we think that the Fed will take more action, including (at some point) more quantitative easing, we do not think that it has the tools necessary for the job. For the federal government, after the wholly unnecessary and damaging debacle over the debt-ceiling, the Obama administration is now trying to "pivot" to a jobs agenda, and the president will be delivering proposals on September 8. Unfortunately, his proposals are likely to be small compared to the magnitude of the problem. A major program of public works to re-employ unemployed construction workers would be an appropriate option (although it could not kick in quickly), especially since the government can borrow so cheaply to finance it (10-year Treasury yields are close to 2.0%). But the political argument over stimulus has been decided—stimulus is perceived to have failed—so a major new stimulus program is very unlikely. Probably the best that we can hope for is that the government will extend temporary payroll tax cuts and emergency unemployment compensation into 2012, to avoid piling on an even bigger fiscal contraction.
