The Wall Street Journal reports this morning, based on new figures from the Bureau of Labor Statistics, that unemployment rates continued to climb in July in metropolitan America.

It’s probably true, based on BLS estimates that the nation overall shed another 247,000 jobs in July. 

But BLS’ metropolitan data don’t exactly show us what’s going on from month to month.

That’s because they don’t take account of the natural fluctuations in employment levels that occur everywhere due to seasonal labor market changes. Or, in data-wonk terms, they’re not “seasonally adjusted.” 

Why does this matter for July unemployment figures? Well, school’s out. And more young people on the labor market generally translates into higher unemployment rates in June, July, and August. You can see the seasonal pattern in national unemployment figures over the past few years in the chart linked below. (The article acknowledges this distinction implicitly, by comparing metropolitan rates to the national unadjusted figure for July.)

The other seasonal unemployment trend comes at the end of the year, when people are hired for retail positions over the holidays, then resume looking for work in January.

OK, back to the metropolitan story for a minute. The Journal article does nail the wide regional disparities in the unemployment picture. Things continue to look fairly grim in a combination of auto industry-dependent metro areas (Detroit, Toledo, Youngstown) and Sunbelt metro areas feeling the hangover from the housing crash (California’s Central Valley, coastal Florida, Las Vegas).

At the other end of the spectrum, several mid-sized “middle America” metro areas (Omaha, Des Moines, Provo, Oklahoma City) are faring relatively well, as are national and state capitals with a significant presence of government jobs (Washington, Madison, Honolulu, Salt Lake City).

And unemployment is up everywhere from where it was a year ago--which is how BLS (and Brookings, in its MetroMonitor) examines the metropolitan trend, in order to cancel out the seasonal “noise” in the data.

We’ll present a broader picture of metropolitan economic health through the second quarter of 2009 in our next MetroMonitor, due out in 2 weeks. Until then, celebrate back to school--for the kids, and for labor market statistics.