The US unemployment rate fell to 7.9per cent in September, from 8.4per cent in August, a big drop that in normal times would be welcome news for a presidential incumbent seeking reelection in just over a month.
These are not normal times.
As the best-known summary statistic of the labor market, the US unemployment rate is "a psychologically important number" for voters, said Michael Brown, principal US economist at Visa.
But President Donald Trump's announcement on Friday that he had tested positive for the novel coronavirus pushes that number into the background: Voters may be "weighing news related to the virus a bit more than the economic data right now," Brown said.
The drop in the September jobless rate, reported by the Labor Department on Friday, extends a steep downward trend from the 14.7per cent registered in April, which was the highest level since the Great Depression.
But other details in the report do not easily fit into Trump's narrative of an economy roaring back to life.
Monthly job gains slowed. Overall, of the 22 million jobs lost since February, the economy has recouped about half.
"Regaining the other half is going to be a whole lot harder," said Michael Arone, chief investment strategist at State Street Global Advisors.
Notably, some 865,000 women left the labor force last month, the data shows, about four times the number of men. Latinas accounted for more than a third of that decline, the report showed.
Those populations are key to Trump's reelection hopes as well as those of Democratic presidential nominee Joe Biden.
The mass exit of women from their jobs coincided with the start of the US school year, with many children learning online and at home.
"These numbers are really just what parents have been screaming for months, but in the form of economic data," said Michael Madowitz, an economist at the left-leaning Center for American Progress. "I can't imagine this is going to help win voters over."
Friday's report counted more than 12 million Americans among the unemployed, a demographic less likely to show up at the polls than the employed, studies here have consistently shown.
But in the current context, there's a twist: voter turnout among those out of work tends to rise when unemployment overall is high.
Amber Wichowsky, a political science professor at Marquette University in Wisconsin, studied thousands of state and local elections and found that higher unemployment is associated with higher turnout, and Republican incumbents "are more likely (than Democrats) to be punished by bad unemployment numbers."
In the Nov. 3 election, she said, it could well be different: the public health crisis could suppress voting if people are worried about casting ballots in person.
Moreover, it's hard to find a clear pattern linking unemployment rates to turnout in presidential elections, or to the outcome.
Where was unemployment before past elections?
The jobless rate was nearly as high as it is now when voters picked Jimmy Carter, a Democrat, over Republican President Gerald Ford in 1976, when Carter lost to Republican challenger Ronald Reagan in 1980, and when Democratic Arkansas Governor Bill Clinton ousted Republican President George H.W. Bush in 1992.