No Good Spin On the June Jobs Number

The June employment picture was worse than the headline number implied.

Nonfarm payroll jobs registered a pathetic 18,000 increase last month, and even that small amount was not statistically significant, falling as it did within the survey's confidence interval. Adding to the bad news, the changes in payrolls were revised downward in both April and May. Excluding the June drop in government jobs, private sector payrolls last month showed a 57,000 rise, but that modest amount also failed the labor department's test of statistical significance.

Total employment in June fared worse. According to the government's monthly household survey, total civilian employment fell by a whopping 445,000 last month, a drop that passed the test of statistical significance for that series. Consistent with that decline, the related civilian employment to working-age population rate, a measure of labor force utilization, dipped by 0.2 to 58.2 percent. The last time that rate was lower was in July 1983.

The labor force participation rate (labor force as a percent of the working-age population) also dipped last month to 64.1 percent - the lowest level since January 1984. The rate is down from 66 percent prior to the recent recession. A decline during a weak job market signals a rise in hidden unemployment as more frustrated jobseekers retreat to the sidelines and are not counted as officially unemployed. Even though the government-measured discouraged worker rate misses many such jobless dropouts, it nevertheless rose by 0.2 last month.

The official unemployment rate ticked up by 0.1 to 9.2 percent in June, not a statistically significant change. But it was up by 0.2 over two months ago, which does pass the significance test for that series, and it was up 0.4 over its recovery low point in March.

Hours and earnings in the private nonfarm sector also weakened in June, declining by 0.1 hour and 1 cent, respectively.

The major monthly labor market data series are seasonally adjusted, and such adjustments are subject to error. Also, seasonally adjusted data disguise what actually happened. Last month is a good example. The change in seasonally adjusted payroll jobs was insignificant. Before seasonal adjustment, however, payrolls rose by 376,000, a solid amount. Private sector jobs rose even more - by a colossal 840,000.

Another example: involuntary part-time employment, seasonally adjusted, was unchanged last month. Before adjustment, however, it shot up by 468,000. That was reality, and unnerving.

Prior to seasonal adjustment, the official unemployment rate last month was 9.3 percent, up sharply from 8.7 percent in May, again illustrating how drastically statistical processing can alter reality. In contrast to the 0.4 rise in the seasonally adjusted unemployment rate between March and June, last month's unadjusted rate was only 0.1 higher than in March.

You won't find the not seasonally adjusted numbers discussed in the government's summary press release on the employment situation. Unfortunately, we have been trained to ignore the churning undercurrent behind the processed data.

Policymakers will say - and rightly so for their purposes - that they need seasonally adjusted data to more clearly observe cyclical and trend changes in economic time series. But the public should know what actually happened. It wouldn't hurt to include a few sentences in the labor department's monthly press release to describe recent changes in seasonally unadjusted data for the major labor market indicators.

Nevertheless, the public, including voters, will continue to watch the seasonally adjusted economic data highlighted in the media and the reality they see around them in their daily lives. The data since March describe a labor market that is stagnant to declining. We don't know yet if this is a temporary step-down or if it signals a weaker trend in the months and quarters ahead.

Even with some improvement in employment over the next year, the unemployment rate could linger in the 8 percent range, if only because the hidden unemployed will share in the jobs created, thereby dampening what would otherwise be, at best, a modest downtrend in joblessness.

The state of the job market will likely be the single biggest determinant of who becomes our next president. Postwar history tells us that U.S. presidents don't get re-elected when unemployment is 8 percent or higher. Right now, the odds of the economy growing fast enough to reduce the unemployment rate to the 7 percent range or lower by election day look pretty long. Add to that the widespread disillusion over Obama's stimulus and regulatory policies and the chances of a Republican victory in the White House and both houses of Congress next year look pretty good.