Time to get out of the forecasting business

The BLS jobs report was much bleaker than I had thought it would be. One can't overreact to every squiggle in the data, but the May employment numbers provide no evidence for the robust recovery scenario. Highlights:

Payroll employment: +431,000, of which 411,000 are temporary Census workers; I had thought it would be closer to +700,000

Private nonfarm payroll employment: +41,000 - close to the much-maligned ADP figure

Unemployment rate: down to 9.7%, but largely due to unemployed workers (net) dropping out of labor force

Employment from household survey: -35,000

Unemployment from household survey: -287,000

Labor force: -322

The one bright spot in the report: aggregate weekly hours rose by 0.3 percent following 0.4 percent increases in March and April. The economy continues to use more labor, but is finding it by increasing average weekly hours worked rather than new employees.

Bottom line: I can continue to be a contrarian for at least one more month.

Huh?

The New York Times says that persistent high unemployment threatens Obama's chance at reelection. No argument there, but the article drops this nugget: "No American president since Franklin Delano Roosevelt has won a second term in office when the unemployment rate on Election Day topped 7.2 percent. Seventeen months before the next election, it is increasingly clear that President Obama must defy that trend to keep his job."

Some trend. The unemployment rate has been over 7.2 percent on election day when an incumbent president is on the ballot exactly once since the Truman Administration: 1980 (7.5% in November) when Jimmy Carter lost in his attempt at reelection. Ronald Reagan won reelection in 1984 when the November unemployment rate was exactly 7.2 percent. These are the only two times that the unemployment rate has exceeded 7 percent when an incumbent president was up for reelection, so it is equally true to say that "50 percent of American presidents since FDR have won a second term in office when the unemployment rate on Election Day topped 7 percent." Not as compelling a headline however.

Matthew Yglesias on Christina Romer on the unemployment problem

Christina Romer argued in a talk at Princeton that the reason the unemployment rate is so high is not structural problems in the economy but a simple deficiency of aggregate demand. Matthew Yglesias draws the wrong conclusion:

Every time there’s a downturn a certain swathe of the elite starts to label it unfixable and structural. And the worse the downturn, the louder come the calls. Look at the history of the Great Depression and you see an enormous chorus of voices from the right arguing that nothing could be done and people would just have to suffer through it. They were countered by a chorus of voices from the left arguing that nothing could be done and people would just have to stage a revolution. It wasn’t true then and it wasn’t true now. The fact of the matter is that key people responsible for running the global economy—people at the European Central Bank and the Federal Reserve Board, and the Bank of Japan, people in the United States Senate, people in the Germany cabinet—are screwing up. In the developed world, those countries who’ve been able to respond aggressively to the crisis with aggressive expansion-via-devaluation are all doing pretty well. The bigger developed economies can’t do that exact thing, but they can mount more aggressive expansionary responses—they just aren’t.

That's not how I interpreted her remarks, nor how I think she intended them to be interpreted. I believe Romer was responding to a view that is common among those who criticize the Administration's policies from the left. Their argument is that the disproportionate rise in the unemployment rate during the recession reflects structural problems - firms have figured out how to do more with fewer workers - and therefore we are unlikely to get a significant drop in unemployment unless we have extraordinarily high GDP growth. Romer's point, with which I agree, is that normal recovery-level GDP growth - I would say in the 4-5 percent range for several quarters - should be sufficient to bring the unemployment rate down at a reasonable pace (reasonable meaning as fast as can be expected, not fast enough). The Obama Administration's focus on stimulating demand through fiscal policy, bailing out the financial sector, and supporting the Fed's monetary policy, is the correct approach to the problem.

Unemployment insurance claims

Thursday's ETA report showed that unemployment insurance claims rose in mid-April, potentially bad news for those of us who have been predicting big gains in employment this month. Both initial claims and continuing claims have flattened a bit in recent months:

But all is not lost. The WSJ reports that April's UI figures are lower than they should be because of Easter and a "special holiday" in California (is that what they call it when they have to shut the whole state down because of the budget crisis?). The WSJ report repeats the canard that initial claims "have to drop to 400,000 or lower to indicate an accelerated hiring trend." Not true, at least based on the experience of recent recoveries, as I've noted in earlier posts. Compare the current recovery to the one following the 1981-82 recession:

The black line at August 1983 shows the point in the recovery comparable to where we are now. At that time initial claims were at about the same level as they are now (450,000 or so) and squiggling up a bit. Nevertheless, 1983 was a pretty good year for job creation (3.5 million jobs created!). We're well behind the pace set in 1983, but it still looks like a much stronger recovery than we had after the last two recessions.

Bob sez...

Bob Barbera explains what's going on with the population numbers. Every January the BLS rebenchmarks the household survey numbers, forcing an adjustment in the population, employment and unemployment numbers for January. But they don't go back and revise the previous months' data in light of this, so you can't really compare December to January. It's possible that this phenomenon is more pronounced in the year after the decennial census. So hidden in the employment report is Table C:



This table says that had the BLS gone back and revised December's data in light of their new population estimates, they would have found that household employment rose by 589,000 in January! That's on top of December's household employment increase of 297,000. These are huge numbers that suggest that the two month decline in the unemployment rate from 9.8 percent to 9.0 percent is not a fluke but a meaningful turn in the labor market (here I'm giving my own opinion, not necessarily Bob's). It may just be that the payroll numbers are, for reasons no one can explain, failing to reflect accurately the true state of the labor market. At any rate the household survey, ADP report, ISM reports, initial claims data, GDP report, not to mention retail sales, durable goods orders, etc. are all telling a different, more positive story than payroll employment.

Ick

The Employment Situation for September was worse than expected. Unemployment rate up to 9.8%, 263,000 jobs lost in September. The consensus forecast was more like 167,000 jobs lost.

No one thought the economy would start adding jobs until the end of 2009 anyway, so one shouldn't conclude from this data that the recovery is sputtering. Nevertheless, it would be prudent to start thinking seriously about a second stimulus, as Paul Krugman suggests today. This one should be focused on supporting state and local spending. Likelihood of Congress actually taking this on: zero.