Most Latino, Black and Native American households have experienced serious financial problems during the coronavirus outbreak — and Latinos have fared worst — finds a new survey from NPR, The Robert Wood Johnson Foundation and the Harvard T.H. Chan School of Public Health.
Nearly three-quarters of Latino households surveyed — 72% — reported serious problems paying for housing, bills and basic expenses since the coronavirus outbreak began in the U.S. in mid-March. That’s double the share of white, non-Latino households that reported difficulty affording the same things.
“The findings are not what we expected,” said Robert Blendon, health policy professor at the Chan School who helped lead the study, during a Sept. 16 media briefing. “They’re actually much worse.”
The survey taken earlier this summer found 46% of Latino households said they had exhausted all or most of their savings and 15% reported having no savings before the outbreak. Some 34% of Latino households surveyed said they have had serious problems paying debt, like outstanding credit card bills, while 32% had serious problems paying their rent or mortgage. Meanwhile, 33% had serious problems paying utilities.
Latino households fared worse in those categories than other households. The researchers define “serious problem” as “something that creates great difficulty for you and people living in your household.”
Among Black households, 60% said they experienced serious financial problems, 41% had used all or most of their savings and 10% didn’t have prior savings. Among Native American households surveyed, 55% said they had experienced serious financial problems, 41% had no or almost no savings left and 9% reported they were without prior savings.
For Asian households surveyed, 37% said they’d experienced serious financial problems. And for white households surveyed, 36% said they’d experienced serious financial problems.
Survey notes
The survey was conducted by phone from July 1 to Aug. 3 among a sample of 3,454 adults weighted to be nationally representative. Respondents were asked to report serious problems for themselves and others in their households. The researchers applied the race or ethnicity that respondents self-reported to the respondent’s entire household.
It’s important to note the margins of sampling error. For Native American households the error margin is 15%; for Asian households it’s 14.1%; for Black households it’s 9.1%; for Latino households it’s 8.8%; and for white households it’s 4%.
The researchers report a 95% confidence interval. That means if the survey were repeated over and over, 95% of the time the results would fall within the margin of error. For practical purposes, that means it’s likely the real-life number falls within the margin of error of the survey.
Take the 72% of Latino households that reported serious financial problems. The 8.8% margin of error means if researchers surveyed Latino households over and over, it’s highly likely that each survey would find between 63% and 81% of those households reporting serious financial problems. If the researchers took a census of every Latino household in the U.S., the result would very likely fall within that range.
As with all surveys, takeaways are less clear for populations that make up less of the sample — there‘s simply less information to work with. For example, there were 1,750 white respondents in this survey compared with 101 Native Americans. If the researchers ran the survey over and over, they would expect that 95% of the time between 32% and 40% of white households would report serious financial problems, considering the 4% margin of error for those households. Considering the 15% margin of error for Native American households, surveys conducted over and over would be highly likely to show between 40% and 70% of those households reporting serious financial problems.
Confidence intervals are sort of like shooting hoops in the dark. Think of a basketball player on a moonless night who comes up with a repeatable shot so good that she swishes it 95% of the time. Then she aims for a court that’s further away — but with a bigger hoop. The bigger hoop compensates for the distance, so she can maintain that 95% swish rate. The bigger the hoop, the bigger the margin of error.
Another fact worth noting: The survey period overlapped with the end of an extra $600 in weekly federal unemployment benefits, which expired late July and were part of the CARES Act that became law in late March. Congress and the White House have not gotten together on a new aid package.
“Regardless of the billions being spent, we have not put a [financial] cushion [in place] to get through this natural disaster in health among minority communities,” Blendon said.
Job loss
According to the new survey, 63% of Latino households reported having at least one household member who had lost a job or business, been furloughed or had wages or hours reduced since the start of the coronavirus outbreak. For Native American households the rate is 46%, for Black households it’s 44% and for white households and Asian households the rate is 42%.
In other words, American households writ large have experienced significant job and wage losses and reduced hours. Survey data from the Bureau of Labor Statistics show that, to the extent employment has rebounded, that rebound has only been beneficial for some racial and demographic groups.
Recovery for whom?
The official unemployment rate is 8.4%. Because of an ongoing misclassification error, unemployment could be as high as 9.1% — an “upper bound” that, according to the BLS, “probably overstates the size of the misclassification error.” The BLS error had a more pronounced effect on the unemployment rate in April, May and June. BLS conducted its survey for the current unemployment rate during the week of Aug. 9, about a week after the NPR/Robert Wood Johnson/Chan survey closed.
Today’s 8.4% unemployment rate is a marked improvement from 14.7% unemployment in April, and is markedly higher than the 3.5% rate in February.
A quick look beneath the overall unemployment rate shows jobs have returned unevenly for Black, Latino, Asian and white workers. White workers remain the only race or ethnicity with unemployment below 10%. The rate is 7.3% for white workers compared with 10.5% for Latino workers, 10.7% for Asian workers and 13% for Black workers.
There are no federal monthly unemployment numbers for Native Americans. Here’s why, according to BLS:
“Because the sample size for [American Indians and Alaska Natives] is relatively small, the U.S. Bureau of Labor Statistics (BLS) does not routinely tabulate detailed data on their demographic and labor market characteristics. The AIAN population is extremely diverse; its members have origins in hundreds of distinct and culturally diverse peoples from throughout North and South America. Well over 500 tribes are federally recognized within the contiguous 48 states and Alaska, and many more are either not recognized or are recognized by states only.”
Internet connectivity
Many workers with office jobs have shifted to working from home since March. Yet having an office job doesn’t automatically mean telework is doable — workers still need high-speed internet access. The new survey finds 27% of Native American households reported having serious problems with their internet connection to do work or schoolwork, alongside 24% of Latino households, 22% of Black households, 17% of white households and 11% of Asian households.
Internet connectivity aside, millions of American workers have jobs that can’t be done at home and are vulnerable to wage loss if they can’t physically be at their workplaces.
“One young lady was a frontline health care worker who ended up contracting COVID at work and then ended up having to get placed in a hotel because there were so many family members in her home that there was really a risk of transmitting to everyone else,” recounted Lisa Cooper, a professor of medicine at Johns Hopkins University who wasn’t involved in the survey, speaking during the media briefing about one of her patients. “Then at the same time her partner had to not go to work, to take care of the children. And he was an hourly worker, so he lost income during that.”
Read the full results of the survey and check out the rest of our COVID-19 related coverage.
(The Robert Wood Johnson Foundation, along with other grant-making non-profit organizations, funds Journalist’s Resource.)
Expert Commentary