Jobs Report: Upside Surprise on Payrolls!

Payrolls were up 204,000 last month, well above expectations, and the unemployment rate ticked up slightly to 7.3 percent, due in part to the government shutdown, which lasted through the first half of the October.

The payroll number is a big, upside surprise (MarketWatch called it "shockingly strong"). Even though, as explained below, furloughed government workers were recorded as still on the payroll, analysts expected the impact of the shutdown on industries that depend on government activities to lead to a lower job count, around 120,000. In fact, most industries other than government added to their payrolls last month. In addition, payroll counts for the prior two months -- Aug and Sept -- were revised up by a total 60,000 jobs.

These numbers paint a pretty different, and more positive, picture of recent job growth than even the last jobs report, which came out just a few weeks ago on Oct 22 due to the shutdown. Averaging over the past three months, employment was up about 200,000 per month, compared to 150,000 in the prior three months. That acceleration is more in keeping with yesterday's report of 2013Q3 GDP growth, which also came in higher than expected.

Results from the household survey, however, are less positive and more clearly reflective of the shutdown (though note also that this survey, with its smaller sample size, always has a lot more noise than the payroll survey). Both the labor force and employment fell by more than 700,000 in the household survey last month, and the share of the population in the labor force fell a large 0.4 percentage points.

These declines are clearly shutdown related, as furloughed government workers are counted as unemployed in the household survey but on-the-job in the payroll survey (I explain the differences in methodology below). However, the BLS noted that because some furloughed gov't workers were misclassified as with a job but absent from work, the jobless rate should have been slightly higher.

At any rate, I expect these sharp declines to bounce back next month, though it will remain the case that both labor force participation and the share of the adult population with jobs are well below historical trends.

What to make of this report? The signal to noise ratio is always low in the monthly job numbers -- that's why revisions can change the broad outlook from one month to the next. Last month it looked like payrolls might be decelerating; now it looks like they might be accelerating. And with the shutdown in play, the noise is higher than usual. But I think it is probably fair to conclude that the job market is doing a bit better than we thought it was, though still not great, meaning job growth is still not fast enough to drive the jobless rate down to levels closer to full employment anytime soon.

October's results beat expectations, but expectations were low, and 200,000 per month -- the most recent 3-month trend -- is moderate job growth that will only slowly provide jobseekers and wage earners with a better chance of claiming a bit more of the economy's growth.

Still, the report contains some signal, and that signal is once again a reminder of how deeply resilient the U.S. economy remains, even after all the lousy policy, fiscal drag, and self-inflicted wounds that our dysfunctional politics have thrown at it.

It's enough to lead one to wonder how we'd be doing with good economic policy.

By the way, I wouldn't be surprised if equity markets fell when they open soon, based on the view that a stronger-than-expected report and faster recent payroll trend will lead the Fed to taper sooner. That's possible, but they're sensitive to the weak signal problem, and will want more solid evidence than is herein.

The Shutdown and the Jobs Report

The impact of the partial government shutdown (Oct 1-Oct 16) on the data is a bit tricky. Remember, there are two surveys in play here, one that gets its labor market information from people in their households and the other directly from workplaces. The latter gives us the payroll counts; the former, the jobless rate.

-- furloughed workers are supposed to be counted as unemployed, on temporary layoff, in the household survey that measures unemployment (as noted above, some appear to have been mis-classified as with a job but absent);

-- however, if someone is paid for a job, even if they're furloughed -- and furloughed federal workers were paid retrospectively -- then the payroll survey of workplaces counts them as employed;

-- average weekly hours from the payroll survey, which were unchanged last month, are for the private sector so they don't reflect direct government furloughs;

-- however, private businesses that cut work due to the shutdown, like a café outside of a shutdown national park, should show up in today's report.

This post originally appeared at Jared Bernstein's On The Economy blog.