The press reports the bad news this morning that the unemployment rate closed in on 10 percent last month, hitting 9.7 percent from 9.4 percent in July.
That’s a truly awful number—the highest in twenty-six years. But, as we and others have said many times before, it just begins to capture the true misery of joblessness out there.
That’s because the standard measure of unemployment, known as U-3, doesn’t include people who are so discouraged that they’ve quit looking for work. And that doesn’t mean they haven’t looked for work in three years, though that’s part of it. It also doesn’t include people who haven’t looked for work in the last four weeks, for whatever reason.
But the Bureau of Labor Statistics does report that all-inclusive number in its U-6 measure. For August it jumped half a point to 16.8 percent. A year ago it was 10.7 percent.
So how’s the press done so far (the BLS released the number this morning)?
The Journal’s lead story on its Web site right now doesn’t mention the U-6 number. Nor do any of the triple Timeses: New York, Financial, or Los Angeles.
But the Washington Post gets it done. As do Reuters, Bloomberg and NPR’s Planet Money.
At this point, there’s not much excuse for major papers like the Journal and Times to skip the U-6 datapoint, even in their early spot-news stories. It would take them all of thirty seconds to include this crucial number.
UPDATE: Just as I published, Twitter popped up a link to a good WSJ blog post saying the U-6 gives “a clearer picture of the broader employment situation” than the traditionally reported figure. Again, more reason why it should be in the paper’s main story.
I suppose U-3 is OK duriing normal times in the labor market, but these are not. Among other faults, U-3 has a sort of natural limit on it. As soon as the number of people running out of unemployment benefits equals the number of new applicants, U-3 stops rising.
On a related point, why doesn't anyone ever refer to ShadowStats? OK, it's not "official", but since when did a politically manipulated set of numbers take on such a holy glow?
#1 Posted by Benedict@Large, CJR on Fri 4 Sep 2009 at 11:12 AM
Actually, you're argument is extremely tenuous and you should search out the full facts before publishing. Following is a very cogent article from Barron's Economist and economics writer from the issues of Aug. 31 about why U3 is a better stat than U6. COpied below:
IS THE ECONOMY IN UNCHARTED WATERS, SAILING AGAINST the current without a compass? Not really -- but much that has been said and written about the labor markets leaves that impression. Very few of these assertions are strictly false. But all are potentially misleading, since they tend to imply that the unprecedented has happened, and that there is no adequate explanation for it. And if you believe in the unprecedented, it's a small leap to the belief that all talk of an imminent recovery from the worst recession since the Great Depression is also a leap in the dark.
Main case in point: widespread disdain for the official unemployment rate, which ran 9.4% in July, about unchanged from its 26-year high of 9.5% the month before. The official measure covers only those who actively looked for work over the previous four weeks. Another, more inclusive measure covers those who "have looked for work sometime in the recent past" (to quote the Bureau of Labor Statistics), plus the huge number of involuntary part-timers. This number ran 16.3% in July, it has been pointed out, and is therefore the more "valid" yardstick.
Or as one reader put it to me in a recent e-mail, the fact that I "persist in peddling the fiction" of the official unemployment rate shows only that "professional pride has no bearing on" my reportage.
The main reason I generally cite the official rate, dubbed "U-3" by the Bureau of Labor Statistics, is that it has by far the most history. The 16.3% jobless rate, which the BLS calls "U-6," goes back only to 1994, when the survey questionnaire used to gather the data was revised to such a degree as to virtually redefine the series.
But U-6 would certainly be newsworthy if there were any convincing evidence that its trend were behaving in some unusual way. All the evidence suggests that U-6 is about on trend. In July 2000, for example, the official unemployment rate, U-3, was close to its 30-year low of 4.0%, accompanied by much celebratory talk about the full-employment economy and even by concern that joblessness was running too low. Naysayers could have pointed out, however, that U-6 was still running 7.0% in July 2000, which hardly seemed cause for self-congratulation.
But at 7% and 4%, respectively, U-6 was 75% higher than U-3, roughly in line with the long-term trend. If we apply the same 75% to the July 2009 official unemployment rate of 9.4%, we would expect U-6 to be 16.5% on a rounded basis. The fact that it was actually 16.3% even suggests that it's running a hair below trend.
Since, as mentioned, U-6 doesn't go back earlier than 1994, we can't know how high it went the last time that U-3 was at 9.4%, back in 1983. But take the very highest figure on U-3 in prior cycles, starting in 1994. The official unemployment rate averaged 6.6% for two months in a row in January-February '94, while U-6 (at 11.6%) ran nearly 76% higher. Apply this 76% to the July 2009 U-3 of 9.4%, and again we get 16.5%, which still suggests that, at 16.3%, U-6 isn't quite as high as we'd expect it to be.
BLS research has shown that the more inclusive measures of unemployment are somewhat less reliable than the official measure, U-3 ("Persons Outside the Labor Force Who Want a Job," Monthly Labor Review, July 1998). But U-4 through U-6 are still worth tracking, to remind us that there is more to labor's pain than official unemployment, and to check on the overall trends.
ANOTHER TREND THAT HAS BEEN disconcerting is the greater percentage loss in jobs than at any time since the Great Depression: nearly 5% of the total, compared with about 3% in the early 1980s, the period of the next-worst performance. Here
#2 Posted by matthew Greco, CJR on Mon 7 Sep 2009 at 11:20 AM
If you want the employment stats that are hardest to manipulate, look to the BLS' employment-to-population ratio (fallen off a cliff in 2002 and never really recovered before current huge collapse)...
#3 Posted by cas127, CJR on Thu 10 Sep 2009 at 04:16 AM