At the Associated Press, economics writer Christopher Rugaber used not seasonally adjusted data published by the government's Bureau of Labor Statistics on metro area employment and unemployment to crow about "widespread improvement in the job market." The predominance of part-time jobs among the new ones created and fact that houshold incomes have yet to recover from the recession apparently had no impact on his assessment.
The opening sentence of the government's report reads: "Unemployment rates were lower in July than a year earlier in 320 of the 372 metropolitan areas, higher in 38 areas, and unchanged in 14 area, the U.S. Bureau of Labor Statistics reported today." But the second paragraph of Rugaber's AP report, headlined "Unemployment Rates Fall in Two-thirds of U.S. Cities," tells readers that "[U]nemployment rates fell in 239 of the nation's 372 largest cities in July from June."
Rugaber looked at single-month changes instead of full-year changes. BLS's entire report uses not seasonally adjusted data, so a month-over month comparison really doesn't mean much.
To his credit, Rugaber eventually told his readers that the government data wasn't seasonally adjusted ("Unlike the national data, the metro figures aren't adjusted for such seasonal patterns"). But before he did, he mixed apples and oranges, throwing in the following seasonally adjusted data:
The U.S. unemployment rate fell last month to a 4 1/2-year low of 7.4 percent. That's down from 7.6 percent in June. Employers added 162,000 jobs. That's enough to lower the unemployment rate but below the average monthly gain of 192,000 this year.
Readers who didn't move past that point wouldn't know that Rugaber was mixing apples and oranges. The not seasonally adjusted national rate went from 7.8 percent in June to 7.7 percent in July.
Far more fundamentally at this point, if one is going to sing the praises of "widespread improvement" in the sense that more people are employed, it's more than a little negligent to avoid a couple of quite pertinent realities — even if, as BLS indicated, over 85% of metro areas showed lower not seasonally adjusted unemployment rates than a year ago.
The first, as seen below, in a trend driven by the impending train wreck of Obamacare, is that the new jobs being created this year are predominantly part-time:
Unless there's evidence that part-time work is all people are looking for, which is highly doubtful, this fact directly negates Rugaber's "widespread improvement" claim — even if all 372 metro areas were somehow seeing lower unemployment rates.
The second, as noted by the former Census Bureau employees at Sentier Research, is that household incomes are still well below where they were four years ago when the recession officially ended. A job market which is seeing high-wage jobs disappear, only to see them replaced with lower-paying and often part-time jobs is not what the average American would see as "widespread improvement."
Cross-posted at BizzyBlog.com.