Laid-off workers in Florida fared the worst in receiving timely unemployment benefits when businesses closed at the peak of the coronavirus pandemic, according to a newly published nationwide study.
Researchers surveyed 2,500 laid-off service industry workers in April and May and found that only a quarter of them had received unemployment benefits during the survey period, including those who had been laid off for two months or longer, according to the report released Wednesday by the Harvard Kennedy School and University of California at San Francisco.
The report offers a snapshot of the well-documented delays that unemployment claimants have faced during the pandemic. While the problem peaked in April and May, delays and claims backlogs have persisted in some states, as evidenced by California’s temporary suspension of new applications and formation of a strike team in September—a move that continues to cloud the overall data for weekly U.S. unemployment claims.
The authors—sociology professors Daniel Schneider from Harvard and Kristen Harknett from the University of California—reported wide differences in survey responses by state. On the extremes, Minnesota had paid benefits to 77% of respondents while Florida had paid only 8%. In the middle, roughly 50% of respondents in California and New York had received benefits at the time of the survey.
“Delays and dysfunction in the UI application process have consequences,” the authors wrote. “Essential workers, even when employed, live close to the economic margin—struggling to make ends-meet on low-wages and in the face of unpredictable schedules.” Their survey also found 26% of laid-off workers who hadn’t received benefits ran short of food sometime during the previous month.
State unemployment agencies faced historic challenges in responding to the widespread business closures imposed starting in March to slow the spread of the coronavirus contagion. Agencies saw record-high numbers of unemployment claims, coupled with the requirements of federal legislation that they reprogram their systems to handle new types of benefits—including the $600 weekly federal benefit and the Pandemic Unemployment Assistance for self-employed and gig workers.
“Florida’s UI computer system, like those in many other states, crumbled under the onslaught of UI filings after the lockdown was implemented in March,” said Rebel Cole, a business and finance professor at Florida Atlantic University who studies unemployment trends.
The 2,500 laid-off workers surveyed for the new research brief is “not a really representative sample” of the millions of people who applied for unemployment nationwide, Cole said. Still, he said the figure reported for Florida didn’t strike him as surprising.
“In Florida here it was a total disaster,” he said. “They did finally get the money out and they paid back benefits” for previous weeks. “You still got that money. You just didn’t get it in a timely fashion.”
A spokeswoman for Florida’s Department of Economic Opportunity, which is in charge of paying out state unemployment benefits, declined to immediately comment on the report.
Gov. Ron DeSantis (R) acknowledged in the spring that the state’s unemployment processing system was dysfunctional, stripping the department’s director, Ken Lawson, of overseeing the benefits payout and ordering an investigation into the $77-million contract that Deloitte’s technology consulting business had won in 2013 to build a new unemployment computer system for the state.
Deloitte has said it isn’t responsible for the current condition of the Florida system because it wasn’t hired to maintain it.
Lawson resigned on Aug. 31. Similar crises led to the resignation or firing of officials in at least seven other states.
Shortcomings of Survey
The Harvard/UC research might not accurately reflect the timeliness of benefit payments by state, according to Cher Haavind, a spokeswoman for Colorado’s labor department. The report showed Colorado had paid only 25% of survey respondents, but Haavind said her department has consistently ranked in the top 10 this year on U.S. Department of Labor scorecards for speed of benefit payments.
“We wholeheartedly disagree with the findings in the Harvard study,” Haavind said. “Colorado’s UI program has in fact been cited for a number of achievements in performance, including our ability to consistently and timely pay benefits.”
Doug Holmes, the former director of Ohio’s unemployment program and now a lobbyist on behalf of employers, also questioned the research. He noted the survey size was small and relied on outreach via Facebook to respondents who were offered a financial incentive to participate.
“A lot of people have recognized the need to improve systems that states rely on to process these applications. No question about that,” Holmes said. But he questioned some of the specific conclusions of the research and said its authors seemed to mistakenly think of unemployment as a public assistance program.
“If the purpose is to address broad macroeconomic issues with a public assistance program, that is hard to administer on top of an unemployment insurance program that’s paid for by employers,” he said. “It’s a very important distinction to make.”
Harknett, the report’s coauthor, said she and Schneider have been conducting similar surveys for five years as part of the Shift Project—a research effort they cofounded to interview hourly workers about their experiences and the effects of unpredictable work schedules. The sample size of this survey is larger than many presidential election polls, she said, and they have techniques for avoiding biases and overweighting of particular demographic groups.
Addressing the skepticism about the results, Harknett said any potential biases in the survey sample should have pushed all states up or down together.
“When we’re making comparisons across groups” within our own data set, “we worry less about that problem,” she said.
“I do think it’s important to acknowledge, if we went out four or five months in the survey, we will see the percent getting UI creep up over time,” Harknett added. “But what we can say is there are a whole lot of people not getting UI for a long time.”
More Resources Needed
The research reinforces the observations many policy analysts have made since March: that state unemployment systems often run on outdated computer systems and without adequate staffing or other resources, said Isabel Soto, director of labor market policy for American Action Forum.
“It kind of feels like a broken record at this point. We need to update the systems,” she said. But because the nation was enjoying unusually low unemployment rates and strong economic forecasts prior to the pandemic, the problem wasn’t on the radar for state policymakers until it was too late.
“We were in a moment of record unemployment,” Soto said. “There was no reason to be looking at these things.”
The U.S. Labor Department is in the process of reviewing how the largest states have handled unemployment benefit processing during the pandemic, with an eye toward potential improvements and policy changes.