In 17 states, women filed the majority of new unemployment claims in the weeks after their governors closed schools and workplaces to curb the spread of the coronavirus, a new analysis by The Fuller Project has found.
Over the previous 25 years, women have typically comprised just over one-third of all unemployment recipients in March. Even during the Great Recession, women never topped more than 47 percent of recipients.
But the consistency of the statistics across every state, along with other research published in the wake of the pandemic, indicates women may very well be the majority of the staggering 20.1 million new claims filed during the first four weeks of this recession.
The Fuller Project requested weekly breakdowns of new unemployment claims from every state, from March 1 through April 11, by gender, race, ethnicity, age, industry and occupation. Seventeen states provided some or all of the information requested. In all of those states, after governors issued the first social distancing orders, women’s share of all new unemployment claims surged to highs ranging from between 53 percent in Wyoming to 67 percent in Alabama.
The trend is particularly alarming, experts say, because millions of Americans are still waiting for their unemployment checks, weeks after losing their jobs, and because single mothers, particularly single mothers of color, were already experiencing high rates of economic insecurity.
“An event like unemployment, particularly if it’s a female headed household, can really be a punch in the gut, in terms of one’s ability to really make ends meet,” said Melissa Boteach, Vice President of Income Security and Child Care at the National Women’s Law Center.
Combined, the state data obtained by The Fuller Project represents 4.5 million of the 20.1 million new claims filed between March 15 and April 11.
The Fuller Project also analyzed the U.S. Department of Labor’s state-by-state breakdowns of unemployment claimants by demographic, industry, and occupation, from 1995 through 2019, for a historic benchmark.
Women’s share of new unemployment claims significantly higher
In the 14 states that provided data broken down by week, women’s share of new claims after the first coronavirus shutdowns happened in mid-March was significantly higher than their median share of unemployment recipients that month over the previous 25 years — from 11 percentage points in Oregon to 33 percentage points in North Dakota. (In Oregon, nearly half the applications filed after the shutdowns were missing gender information.)
“We need more data to better understand why so many women have lost their jobs,” Congresswoman Suzanne Bonamici (D-Oregon) said on Wednesday in an emailed statement.
In Nebraska, the top five categories seeking benefits were waitresses and waiters, hairdressers and cosmetologists, retail salespeople, bartenders, and childcare workers —professions made up primarily by women in the United States.
Employees in those sectors were among the first to become jobless, as governors issued social distancing orders that banned dine-in service at restaurants, child care for all but frontline health workers, and any services that required person-to-person contact, such as hair care, massages, and manicures.
The Economic Policy Institute estimates that women represented 59 percent of the job losses in early March, when unemployment related to coronavirus had just started creeping up, despite women being only half the workforce.
A poll conducted by ABC News and The Washington Post found higher rates of job losses and pay cuts among the most economically at-risk Americans, including low-income workers, people of color, and women without college degrees.
The Fuller Project also compared new claimants grouped by race, ethnicity, age, occupation, and industry, when states provided those statistics, with the groups’ typical share of unemployment recipients over the previous 25 years.
Workers under the age of 25 experienced the most consistent upward shift in their share of unemployment claims. In 10 of the 12 states that provided weekly data on age, these workers’ share of new claims was higher than the historic median by between 5 and 43 percentage points.
In several states, the share of claimants who identified as black or Latino was lower, and the share who identified as white or Asian higher. But tracking “white” applicants is problematic because the Labor Department’s format for exporting unemployment claim statistics, which the states generally followed, only disaggregates by one characteristic at a time, and separates race from ethnicity. Most Latinos choose “white” for race.
Lack of detailed breakdowns
The lack of more detailed breakdowns of groups, also makes it impossible to identify trends for certain segments of the population — such as low-wage workers of color, or single mothers who were forced to leave their jobs over childcare issues.
Meghan Cohorst, a spokeswoman for UNITE HERE, which represents 300,000 workers, said 98 percent of its members, most of whom are women and immigrants, were laid off due to social distancing measures.
One member, Christina Aguirre, a 50-year-old former housekeeper at the Fontainebleau Miami Beach, a luxury hotel, said she was laid off on March 16, after business dwindled due to the coronavirus. But she was only able to submit her application for unemployment last Saturday — by mail, because the online application kept crashing her friend’s computer, she said through a translator.
Aguirre, who emigrated from Cuba in 2010 with her husband, said she’s been their sole source of income after her husband was injured and lost his job. Now she wakes up every morning at 5 a.m. to stand in line at their local food bank.
“There are too many bills to carry,” Aguirre said. “Right now the people that are suffering are me and my husband.”
Over the weekend, Aguirre and other UNITE HERE members drove their cars in a caravan to Florida House Speaker Jose Oliva’s Miami office. Only 16 percent of the 680,000 unique claims submitted in Florida since March 15th have been covered, the Florida Department of Economic Opportunity recently acknowledged.
In the wake of the current health crisis, Congress added $600 each week to the amount states already provide for unemployment insurance, extended the duration of payments to to 39 weeks, and created the Pandemic Unemployment Assistance program for some groups of workers who would have normally been denied benefits under state guidelines. But reports from across the country have found the rollout has been inconsistent and plagued by delays.
Pre-existing state unemployment guidelines tend to exclude part-time and low-wage workers, “exactly the types of workers this pandemic is affecting the most,” according to a recent Brookings Institution report.
Workers denied traditional insurance — including child care workers and hairdressers who are self-employed, and single moms working low-wage jobs — would theoretically qualify for the expanded benefits under the Pandemic Unemployment Assistance program.
But there are limits to coverage. “If an individual was working very low hours and had very low earnings, they won’t be covered and exactly where the threshold is drawn will vary across states,” Susan Houseman, vice president of the W. E. Upjohn Institute for Employment Research.
She added that there still isn’t much information available on the specific groups of workers who are routinely denied benefits, which would be critical for informing policy.
This information is in the states’ databases, but it’s currently not being gathered by the Labor Department.
Congress, the Labor Department, and state agencies, continue to make adjustments to this troubled new relief program, but “I still think there’s likely to be many holes,” Houseman said.
The data used in this investigation is available for reporters in this Google sheet document, as well as Github. If you have more questions about the data, you can contact Sarah Ryley at: firstname.lastname@example.org.
A previous version of this story stated, “In 12 of the 14 states that provided weekly data on age, these workers’ share of new claims was higher than the historic median by between 5 and 43 percentage points.” This was inaccurate; their share was higher in 10 out of 12 states that provided weekly data on age.