Thoughts on the May Jobs Figures

On June 5, the BLS reported that the national unemployment rate in May dropped to 13.3 percent, from 14.7 percent in April, and that nonfarm employment increased by 2.5 million. The stock market staged a huge rally and the media expressed astonishment. I guess I don’t read enough economic forecasts these days because I was surprised everybody was so surprised.

On May 21 I tweeted:

How can the PPP have committed over $600 billion to firms expected to maintain payrolls, and yet unemployment claims are over 38 million? Either many of those unemployment claims will be withdrawn once the PPP money flows through, or the PPP program has been a complete failure.

The May jobs figures answered my question. The Paycheck Protection Program (PPP) has been at least a partial success. From March 21 through May 30 initial claims for unemployment insurance totaled an astonishing 42.2 million, a number equivalent to about 27 percent of total employment at the beginning of March. But in the meantime, Congress passed the CARES Act, a central feature of which was the PPP. The PPP was intended to limit unemployment during the pandemic shutdown through, essentially, a federal subsidy of business payrolls, as well as to help small businesses survive by providing the liquidity to pay other fixed costs.

The PPP was designed as a loan program, but since the loans are 100 percent forgivable it is, in effect, a grant program. As initially legislated, PPP loans are forgivable if the borrowing firm used 75 percent of the loan for payroll and restored its pre-pandemic employment level by June 30. Through the end of May, 4.5 million PPP loans were approved, totaling $510 billion. However, the program only began taking applications on April 3 and administrative problems caused delays and frustration. Then the program ran out of money and Congress had to enact a new phase of emergency aid to refill it. Money didn’t start flowing to firms in a big way until late April. By then some 28 million Americans had filed for unemployment benefits.

The two criteria for loan forgiveness induced firms with PPP loans to rehire employees who had been laid off; many of those employees probably had already filed for unemployment. Undoubtedly, many firms also recalled workers simply because they wanted to continue operations and the PPP loans enabled them to do so. So it was to be expected that many of the initial unemployment claims were precautionary and, as my tweet suggested, eventually evaporated.

The timing of the rehires, as it relates to the reported unemployment rate, was anyone’s guess. The dip in the May unemployment rate, the survey for which was taken the week of May 10, indicates that the rehiring was relatively rapid. What exactly happens to the unemployment rate in June and July is even more speculative, insofar as Congress, on June 3, passed legislation extending the forgiveness deadline to December 31 (and also lowered the amount of the loan that must be used for payroll maintenance to 60 percent.) So firms are under less pressure to restore their employment to pre-pandemic levels by the end of June. It should not be a surprise if the unemployment jumps up again in coming months, nor should it be a surprise if it stabilizes.

There were other things going on during that period in addition to the PPP loans. Six counties of the San Francisco Bay area issued stay-at-home orders effective March 17, which were followed by a statewide California order on March 19 and statewide orders in New York and Illinois on March 20, New Jersey on March 21, and Ohio and Massachusetts on March 23. There was a lot going on, and a lot of confusion, in the three weeks leading up to the BLS unemployment survey the week of April 12.

Where I live nearly all restaurants were completely closed during that period, then beginning in mid-April began reopening for take-out service only; by May, nearly all of them were open for take-out. There was similar confusion in the construction industry and other “non-essential” businesses–employers were trying to figure out whether it was safe to operate, and in many cases needed to get a state ruling on whether they were permitted to operate and how. This sorting-out process is clearly evident in the employment data. Jobs in the construction industry dropped by 995,000 from March to April, but in May the number increased by 464,000. The numbers are even more dramatic for the food service industry, which lost 5.4 million jobs in April but gained back 1.4 million of them in May.

One issue that got a lot of attention was the “misclassification error” relating to people who responded that they were employed but absent from work. Typically those workers are out on vacation, jury duty, maternity leave and the like. Because of the pandemic an usually high number of survey respondents reported that they were employed but absent from work. BLS interviewers were instructed to classify people absent from work due to the coronavirus as “unemployed on temporary layoff.” However, apparently a large number were not so classified and if they were, the reported unemployment rate would have been about three percentage points higher. The problem caused a lot of grumbling and conspiracy-theorizing on the web but I’m in the camp that thinks it’s an innocent technical glitch.

In fact, I’d go further and say that people who responded that they’re employed but absent from work due to the pandemic are closer to what we usually think of as employed than unemployed. Evidently, the respondents themselves thought so or they wouldn’t have told the interviewers they were employed. It’s really a Schrodinger’s Cat-like employment status that’s new to all of us and there’s not necessarily a right way to categorize it. But in any case, the problem existed in April as well so a more or less stable unemployment picture remains.

On Twitter and elsewhere I’ve objected to the notion that the American economy was “collapsing” due to the pandemic, that it was “shut down” and needed to be “reopened.” I just don’t think those terms convey accurately the peculiar economic situation we’re in, and how you frame a problem tends to affect what responses to it are considered. In a similar mode Paul Krugman has suggested that the economy is in a condition similar to a “medically-induced coma.” I understand his point but don’t particularly like the way the metaphor implies an inert, unresponsive patient. I prefer seeing the US economy as a 3-engine jetliner that has been forced to cut off power to one of its engines. If the pilots are skillful, it can still fly, maneuver, and land safely.

It’s in that sense that I find the May jobs report most interesting. It’s our first accurate picture of what the U.S. economy looks like when we’ve shut off power to one of its engines. Take food service, for example. Employment in the sector declined from 12.3 million in February to about 7.6 million in May, so on that basis we can estimate that it’s operating at about 60 percent of capacity. Granted, a barely-warm take-out meal isn’t exactly fine dining, but it still gives us a valuable benchmark to help prepare for a second (or third or fourth) wave of the pandemic. May employment in manufacturing was about 91 percent of its February level, and retail employment was a surprising 87 percent. The business and professional services sector was at 90 percent of its February employment, and the finance and insurance industry at 99 percent.

The employment figures for office-using industries show how impressively adaptable the economy is. In May the unemployment rate for workers in management, professional and related occupations had risen to only 6.6 percent, and the unemployment rate for all workers with a Bachelor’s degree or higher was 7.4 percent. Those figures topped out at 4.8 percent and 5.0 percent, respectively, after the 2008 financial crisis, so they are historically bad, but nevertheless somewhat hopeful considering that virtually every office building in America is currently sitting empty. It underscores that one of the most remarkable aspects of the pandemic response is how totally and seamlessly the white-collar labor force transitioned to remote work. The internet held up, and without it either the economy would truly be collapsing or hundreds of thousands more Americans would have died of the virus.

So in the May numbers I see a jetliner that is still aloft and can make it to a safe landing. The question is whether we cut the oxygen to all the passengers by 15 percent or so, or throw 15 percent of them out the plane. And it’s a problem that our pilot, President Trump, only cares about saving himself, and much of the crew wants to fly by ideology rather than by protocol. After the jobs report was released, a spokesman for the Senate Finance Committee said that the data shows that Congress “should not rush to pass expensive legislation paid for with more debt” and another Republican official told the New York Times “Goodbye phase 4.”

That is the worst possible conclusion to draw from the May jobs report. As Neil Irwin wrote in an excellent New York Times analysis, “Other data points to a severe but slower-moving crisis of collapsing demand that will affect many more corners of the economy than those that were forced to close because of the pandemic.” I don’t think we’re quite there yet, demand can still recover if the virus gives us a summer respite, but the danger is high. A recent Bureau of Economic Analysis release showed that personal income in April increased by 10.5 percent–entirely due to increased government transfers–but personal consumption expenditures dropped 13.6 percent. We can’t force people to spend money, especially if they’re afraid to travel or to eat out, but wise economic policy would seek to plug any shortfalls in private aggregate demand with adequate government spending.

That’s why the emerging Republican stance toward further emergency spending is so dangerous. State and local governments have already reduced their payrolls by 1.6 million workers since March, and states and cities are now preparing budgets for their new fiscal years, many of which begin July 1. Mitch McConnell is surely savoring the prospect of Democratic governors and mayors taking political heat for draconian budget cuts, but will that spectacle be so enjoyable that he’s willing to risk a depression in order to see it?