Economists, political aides and business leaders had been bracing for another horrific month of job cuts and swelling unemployment. In April, the coronavirus shutdown had caused 20 million-plus job losses. Mounting applications for unemployment benefits had suggested that the misery continued through May.

It didn’t. The gap between what was expected and what happened when the Labor Department issued the jobs report Friday morning was so vast that it raised some doubts about its accuracy. But as analysts dug into the numbers, they found the numbers to be correct and suggested that the pessimistic forecasts might have mainly reflected how hard it is to gauge economic performance during a pandemic.

Here are five major takeaways from a jobs report that showed the economy faring better than believed, even if the overall picture remains bleak, with many millions still jobless and unemployment well into double digits.

Is the unemployment rate really 13.3%?

The drop in the unemployment rate was a shocker. Economists had expected the rate to approach 20%, driven up from 14.7% by job losses topping 8 million. Their forecasts woefully missed the mark. Part of the explanation is the difficulty of assessing data when the situation is changing so quickly.

But it also reflects an acknowledged difficulties by the Labor Department’s Bureau of Labor Statistics in its information-gathering. Millions of people appeared to be erroneously classified by the survey as not working but employed. These people should have been classified as on temporary layoff and therefore unemployed. Had they been counted correctly, the jobless rate would have been roughly 3 points higher — 16% — the government said.

The same issue marred the April jobs report. In that case, the unemployment rate would have been roughly 5 points higher than the 14.7% reported.

The jobs report is drawn from a pair of surveys. A survey of households establishes the unemployment rate. A separate survey of employers determines how many jobs were added or lost. Response rates for these surveys were lower than usual in May because of the viral outbreak. But the government still gathered enough responses to produce the jobs report.

“The household survey response rate, at 67 percent, was about 15 percentage points lower than in months prior to the pandemic,” the report noted.

The Labor Department also includes a broader measure of unemployment. This measure includes not only people who are out of work and looking for a job but also people who stopped looking or who were reduced to part-time hours. That rate was 21.2% in May.

Persistent racial inequality

The past 10 days have achingly illustrated the persistence of systemic racism in the United States. Protests emerged in nearly every major U.S. city after George Floyd’s killing at the hands of Minneapolis police officers. The May jobs report pointed to the multitude of disadvantages that African-Americans endure.

Their unemployment rate ticked up last month from 16.7% to 16.8%. By contrast, unemployment for whites fell to 12.4% from 14.2% in April.

Why did the rate rise for African-Americans? More of them began searching for work in May and weren’t necessarily hired. This meant that the number of officially unemployed African-Americans seeking jobs rose by 87,000 to 3.3 million.

But white people who began job hunts were mostly hired: The number of unemployed white people fell by more than 2 million.

African-American women have suffered the brunt of layoffs. Before the pandemic, their unemployment rate was lower than it was for black men. But that relationship inverted during the past two months. African-American women are now more likely to be jobless.

The racial disparities run counter to the odd assertion made Friday by President Donald Trump that Floyd was “looking down” from heaven with admiration for the May jobs report.

“It’s a great day for him, it’s a great day for everybody, this is a great, great day in terms of equality,” Trump said, ignoring the figures that show otherwise.

Temporary layoffs

It’s not just the job totals that matter — it’s where the gains came from.

There were 2.7 million people who had been temporarily laid off and returned to their jobs in May. This was essentially the lowest-hanging fruit of any recovery. They are people who could quickly return to offices, restaurants, stores and factories that had gradually resumed operations. This left 15.3 million people who were temporarily laid off and still awaiting to return. Even if May’s pace of job growth could be sustained — something most economists doubt — it would take at least six months to bring them all back to work.

More part-timers

The May jobs report revealed a curious increase in people saying they working part-time for “non-economic” reasons. The number of part-timers in this category grew by 2 million in May, a major driver of the overall gains.

This category is important because it excludes part-timers who can’t work as many hours as they’d like because of poor business conditions. Instead, these workers are working fewer hours because of issues involving child care, school and other family obligations. It hints at a likely economic drag if schools and child care centers can’t fully re-open.

There was also an increase in “self-employed” workers. This number rose by 401,000 in May to 7.9 million. That figure is still about 1 million jobs below its pre-COVID-19 levels. But it still suggests that parts of the the economy began to pick up last month.

Still worse than the Great Recession

It was an uplifting jobs report after so many bleak numbers. But it’s worth putting those numbers into context next to the Great Recession, the severe downturn that started in late 2007 and ended in mid-2009. The Great Recession was caused by a housing bust and financial market collapse, a far more gradual decline than what the coronavirus caused.

But even with the solid hiring in May, the current jobs picture is still worse than in the Great Recession. During that downturn, unemployment peaked at 10% in October 2009. That is 3.3 points lower than the current rate. With the Great Recession, it took until 2016 for the unemployment rate to return to pre-downturn levels.

In the aftermath of the pandemic, U.S. employers would have to replicate the strong gains from the May jobs report for the next seven months to return the country to where it was in February. This means the results of the June jobs report, to be released early next month, will be a critical test of the potential recovery’s strength.