Women job losses: A deeper look at the data
“And if you look at the damage early on, you know, most of the early job losses were in construction and manufacturing, and disproportionately affected men. But as the crisis intensified, as it did over the course of 2008 — again, before the president came into office — the damage spread and you saw state and local governments, for example, cut back into teachers, fired a lot of teachers, let a lot of teachers go. And a lot of women teach, so you saw the composition of those job losses change over the course of the recovery.”
— Treasury Secretary Timothy Geithner, on NBC’s “Meet the Press,” April 15, 2012
We criticized Republicans last week for promoting the “true but false” assertion that under Obama, women have lost seven times as many jobs as men. This did not stop the Mitt Romney campaign from trying to promote this idea, which Treasury Secretary Geithner on Sunday labeled “ridiculous and deeply misleading.”
We did not think much of the GOP claim because, even though the numbers added up, it is silly to think any economy begins or ends with a presidency. Over the course of the recession (December 2007 to June 2009), the number of jobs declined by just over 5 million, with women accounting for nearly 1.8 million of that figure.
Still, since the recession ended, 2.2 million jobs have been added under Obama, with women accounting for just 284,000 of that figure. So something’s going on. Geithner’s theory — that women started to lose jobs later in the recession — is an interesting one, so we decided to dig deep in the Bureau of Labor Statistics database for some answers.
Teaching (listed in the BLS database as “local government-education”) represents about 9 percent of the jobs held by women in the United States. The data show that, since Obama became president, a larger proportion of the overall job loss by women — 22.7 percent — has been in teaching positions.
In fact, more than a third of female job losses have been in local government jobs; the number of female-held jobs in the federal government has also declined by about 51,000, while state government declines have been marginal. (Since the numbers in the local government section of the database are not seasonally adjusted, we compared figures for January 2009 with January 2012. Comparing figures of the same month is more accurate, especially in teaching.)
Interestingly, the data suggest that local teaching positions held by women climbed by 124,000 in the year before Obama became president, even as the rest of the economy was collapsing. This made the decline feel even steeper when teaching jobs began to decline in earnest toward the beginning of 2010, as federal stimulus money from 2009 began to fade and local governments ran short of funds.
Okay, that kind of supports Geithner’s narrative, though he seemed to suggest that much of this happened before Obama became president, not after. Moreover, it does not explain where the rest of the jobs for women have gone during Obama’s presidency.
That’s because women have not just lost public-sector jobs under Obama. Women have also lost private-sector jobs — some 243,000, compared to a gain of 82,000 in the same period for men.
Clicking through the various job categories, one finds one negative number after another for women — in construction, in manufacturing, in financial services, in retail, in information and so forth. These categories do not even show any uptick from the end of the recession.
One of the few bright spots for women is professional and business services (which includes lawyers, accountants, computer systems and the like), which shows a gain of more than 400,000 jobs since the recession ended. The leisure and hospitality industry also shows a marginal gain.
While Geithner said that “early losses” were in construction and manufacturing, slightly more construction jobs have been lost after Obama became president rather than before; manufacturing jobs also have continue to drop sharply. A large chunk of the gains for men has also come in professional and business services, as well as durable goods (such as building automobiles) and in transportation. Women, meanwhile, have continued to lose jobs in durable goods and transportation.
A Treasury Department official defended Geithner’s statement, but would not do so on the record. “One feature of the recovery is that while employment in the private sector has started to come back, government employment is still falling,” the official said. “Moreover, of those changes in government employment, a large share are accounted for by changes in local government educational services (i.e. teachers).”
The official added: “If you compare employment changes from December 2007 to January 2009 versus January 2009 to March 2012, government represents the only sector that has performed much worse. Every other industry has either had relatively similar job changes on net across the two time periods (e.g. construction) or is doing much better (e.g. manufacturing, professional and business services, retail trade, leisure and hospitality, transportation and warehousing, mining and logging, etc.).”
“Every sector is doing better from February 2010 – March 2012 (most substantially so) than they were from January 2009 – February 2010, while government job losses have accelerated over that period,” the official said. “Teachers represent a large proportion of these state and local government job losses, which is why the Secretary highlighted them.”
The Pinocchio Test
Geithner’s explanation appears incomplete. Certainly the decline in teaching — and government — jobs has been a drag on overall employment for women in recent months, but it’s not the whole story. We continue to frown on measuring job losses or job gains from the start of a presidency, but even our preferred metric of dating the trends from the start of the recovery suggest that women are not faring as well as men in this economy.
We were tempted to say this was worth a Pinocchio, simply because Geithner’s explanation seems so facile. The data suggest a more complicated picture. But given that the data change depending on the dates used, we remain reluctant at this point to attach too much significance to the results. Certainly, the trends are worth watching.
Check out our candidate Pinocchio Tracker
Track each presidential candidate's campaign ads