May’s Jobs Report Brought Good News — But Not For Everyone

We already knew the economic crisis of the coronavirus pandemic was moving much faster than our economic data could keep up with. But the most recent jobs report, which was released today, underscores just how little we know about the current state of the job market at any given moment. In a surprise development, the government’s data shows that 2.5 million more people were employed in May than in April. The unemployment rate, which had climbed to a stunning 14.7 percent in April, dropped to 13.3 percent in May. That’s a huge shock, to put it mildly: Forecasters had predicted that the report would show nearly 1 in 5 Americans out of work and millions more jobs lost.

Although experts told us that we’ll have to wait for the next report to get a sense of where the economy is truly heading, it’s clear that employment in some of the hardest-hit sectors started to recover last month. Almost 50 percent of the job gains for the entire economy were in the leisure and hospitality sector, which had cratered in March and April. This suggests that government efforts may have protected jobs — particularly through the Paycheck Protection Program (PPP), which allowed small businesses to receive forgivable loans if they spent the money on direct expenses like payroll — and that the first few weeks of states’ reopening allowed some workers to return to their jobs.

But this report isn’t all good news: Not only is a 13 percent unemployment rate extremely high by historical standards — for context, it hit a peak of 10 percent during the Great Recession — but also, the recovery doesn’t seem to have arrived for all workers. The unemployment rate fell for both men and women, which is noteworthy because women experienced more job losses during the first few months of the pandemic. But the unemployment rate for black workers (and particularly black women) actually rose from April to May — further exacerbating the racial disparities in which workers are out of a job. The trends aren’t as catastrophic as many expected, but the state of the labor market is still very grim.

Even though the monthly jobs report is one of the most accurate and comprehensive sources of economic data we have, it’s always a single snapshot of a moment in time — one that, in this case more than most, is rapidly disappearing in the rearview mirror. The May report is based on surveys of businesses and households from the middle of the month, before many states started to reopen and protests against police violence began to roil the country. So things could change even more in the next report — and it’s more difficult to predict in which direction.

“This almost feels like a quantum report — in the sense that it’s really hard to interpret this report without knowing the one after it,” said Nick Bunker, the director of economic research for North America at the Indeed Hiring Lab, a research institute connected to the job-search site Indeed. “It looks like the partial rebound that people were expecting a few months from now got here early. But we have no idea how this will continue. Is this what the whole recovery is going to be like? Do we go from a jump to a crawl?”

This report saw huge gains in some of the sectors of the economy where job losses were most concentrated in March and April — areas that were unable to continue to operate because of the pandemic, such as hospitality, education and health services, and retail trade. The leisure and hospitality supersector gained a total of 1.2 million jobs in May, the vast majority of which fall under food services and drinking places. Health care and social assistance saw a gain of 391,000 jobs, with dental offices making up 245,000 of those. And in retail, 368,000 jobs were added, with the biggest gains coming in clothing and accessories stores and general merchandise stores.

A notable takeaway from last month’s jobs report was the impact of the coronavirus recession across many different industries. Although leisure/hospitality continued to hemorrhage jobs in April, the effects of the pandemic were already rippling out to other sectors of the economy, such as local/state government, construction and health care. But in May, the trend of broad losses across the entire economy appears to have reversed itself. The diffusion index is an indicator that shows the share of private industries that added jobs, with an average score of 50. Last month’s index was revised to below 4 percent, meaning that almost no industries added workers. But this month, the index is 64 percent — meaning that many more sectors of the economy added jobs than lost them.

For instance, construction saw the second-biggest jobs increase for May among the major industries in the chart above, adding 464,000 workers — despite not being among the very hardest hit in March and April. That’s a sign that May’s gains are not coming only from workers returning to sectors that had seen the largest losses when the coronavirus first shook the economy.

It’s important to bear in mind that the headline jobs numbers are revised for accuracy in the following two jobs reports, so the total number of jobs gained or lost could change substantially in the coming months. The March number, which was initially reported as 701,000 jobs lost, was revised in the May 8 report to 881,000 lost jobs and to 1.4 million lost jobs in today’s report. But the revision for April’s payroll jobs number was smaller — from 20.5 million jobs lost to 20.7 million jobs lost. So it’s harder to predict what the revisions for May will look like, although Matt McDonald, a partner at the consulting firm Hamilton Place Strategies who writes a monthly analysis of the jobs report, told us in an email that he would be surprised, based on this month’s revisions, if it turned out that jobs were lost in May rather than gained.

One likely explanation for this sudden turnaround is that a significant number of small businesses started to get money from the federal government to bring their employees back to work. The program got off to a rocky start, with banks quickly running out of money and funds flowing first to parts of the country that were less affected by the pandemic. But according to a census survey conducted May 24-30, 80.6 percent of accommodation and food service small businesses had received PPP assistance, compared with an average across all sectors of 71 percent.

It’s also very possible that the handful of states that reopened in late April and early May made a difference. We won’t be able to dig into state-level information until a more granular set of data is released later this month, but McDonald told us that he expects big regional variations to show up.

The jobs gains might still seem at odds with the breathtaking number of Americans who have filed for unemployment over the past few months. Millions of Americans were still applying for payments last week, which was part of the reason that many forecasters expected the jobs report to be so bleak. But there were some signals even in these reports that the labor market might be improving. For one thing, continued claims — that is, the total number of people whose claims were paid out in a given week — finally started to decline during the week of May 16. And Bunker pointed out that, even though unemployment claims are reported more quickly than the figures in the monthly jobs report, those initial claims may not be a more accurate indicator of where the labor market stands, in part because there is a lot of variation in who is eligible for unemployment in different states and in how states report their data.

The slower decline in unemployment claims could also be due to the enormous delays in processing times that have been reported across the country, since workers can still receive unemployment payments for the time they spent off the job even if they returned to work in the meantime. Freelancers and independent contractors may also make up a significant chunk of the later claims — they were able to apply for unemployment insurance through a special, pandemic-specific program that rolled out in many states in late April and May.

There was evidence in this month’s jobs report that the small decline in the unemployment rate and uptick in payroll jobs were due to temporarily furloughed workers returning to their positions. According to the Bureau of Labor Statistics, the number of workers on temporary layoff dropped by 2.7 million in May, to 15.3 million. The bureau also reported that the number of unemployed people who were jobless for fewer than five weeks dropped to 3.9 million in May, a decrease of 10.4 million from April. (The number of people jobless for five or more weeks, meanwhile, increased by 8.8 million in May.) This may also speak to the backlog in processing unemployment insurance claims on a week-to-week basis, with many workers showing up in the claims data even if, in reality, they had already returned to work.

All of that might seem like good news — a sign that the economy will, in fact, be able to bounce back quickly as businesses start to reopen. That’s certainly how some policymakers interpreted it. Stephen Moore, a White House economic adviser, told reporters that the report removes the urgent need for another round of fiscal stimulus, which Democrats have been advocating for weeks.

But there’s plenty of bad news tucked into this report too — and several economists told us that just because the labor market data was less catastrophic than we had thought, it didn’t mean we should pump the brakes on government assistance to businesses or jobless workers. For one thing, it’s important not to lose sight of the fact that, again, a 13 percent unemployment rate is still stunningly high — and the researchers at the Bureau of Labor Statistics still think it may not be including some workers who said they were still employed but were absent from work during the entire week referenced in the survey. If those workers were included, the researchers wrote in the report, the unemployment rate would be 3 percentage points higher.

One of the biggest themes of the April jobs report was that the vast majority of the lost jobs (90 percent, excluding temporary jobs that were expected to be completed) were temporary layoffs. In May, that was still true — to an extent. Eighty-seven percent of job losses were still classified as temporary, though the number of permanent job losses increased by 295,000 from April to May.

Several economists told us they were especially focused on the nature of the layoffs in the May report because it’s an important indicator of how painful the recovery could be. “One of the things that’s been really unusual about this recession is how many temporary job losses we’ve had,” Bunker said. “The more permanent job losses we see, the more this will look like a normal recession, and that’s not good news because in the last two recessions, the recovery has been very stubborn and slow.” He said he was not encouraged by the fact that the number of permanent job losses had increased.

And while the coronavirus recession may have left tens of millions of workers jobless across the entire economy, it’s low-wage workers, workers of color and female workers who are continuing to bear the brunt of the job losses.

In April and May, the groups most disproportionately affected by job losses (relative to their share of the labor force) have been black and — especially — Hispanic workers. Despite making up 17.8 percent of the civilian labor force, Hispanic and Latino workers suffered 22.8 percent of the job losses in April and May. They have the highest unemployment rate of any ethnic group — 17.6 percent — and their rate has risen by a staggering 13.2 percentage points since February.

The unemployment rate among black Americans (16.8 percent) is significantly higher than the unemployment rate among white Americans (12.4 percent), too. Not only did unemployment among black workers start from a much higher point (5.8 percent in February, versus 3.1 percent for white workers), but it has also increased by more percentage points since before the coronavirus crisis (11 percentage points, versus 9.3 points for white workers). And while the unemployment rate for other racial groups fell in May, the black unemployment rate actually rose by 0.1 percentage points this past month, including by 0.6 points for black women.

Martha Gimbel, an economist with the philanthropic initiative Schmidt Futures, also pointed out that many black workers are still employed because they’re essential workers in low-wage jobs — which means they may be working under hazardous conditions where they’re more likely to be exposed to the virus, and they’re likely making less money than they would on unemployment.

Women also continue to be hit by job losses harder than men: 51.4 percent of total job losses since April have been women, even though they make up 47 percent of the civilian labor force.

A major factor is that women are more concentrated in the sectors that were disproportionately hit by job losses. In 2019, women made up 53 percent of workers in accommodation and food services, 70 percent of workers in educational services and 78 percent of workers in health care and social assistance. (They were also 82 percent of the workforce in dental offices.) Although all these areas of the economy saw a rebound in May, all have lost significant numbers of net jobs since the coronavirus crisis began.

Americans can be thankful that this month’s jobs report wasn’t as apocalyptic as many feared. But it’s premature to celebrate the dip in unemployment, even if it did defy expectations. “We shouldn’t let one month and one data point totally change how we’re thinking about the economy,” Gimbel said. “The level of unemployment is still really bad, even if the trends are starting to look more positive.”

And while there are signs of a recovery in the data, there is also plenty of evidence that the rebound did not reach all corners of the economy, particularly the most vulnerable ones. For now, the best grade anyone can give the job market is an incomplete — with even more emphasis placed on the next jobs report, which will speak volumes about how to interpret this one.