Unsanitized: Jobs Reports Struggle to Catch Up to Today’s Reality

Plus, the case count/death count divergence at the state level. This is The COVID-19 Daily Report for July 2, 2020.

by

First Response

We got a rare Thursday jobs day thanks to the holiday, leading to a simultaneous monthly report for June and a first-time jobless claim report for last week. And they’re telling two different stories.

The monthly report, for which the reference week was June 8, shows the height of the reopening across the country. Put it this way, the jobs report correlates pretty well with where infections are at in the past week or two—elevated. The report shows 4.8 million jobs gained and an unemployment rate down to 11.1 percent. Leisure and hospitality led the way with +1.5 million jobs, about 40 percent of the total. Retail added another 740,000. Even government jobs went up slightly, though this is likely due to a complicated seasonal adjustment error.

The reporting error that showed up in the last jobs report was more muted this time, as questioners were better trained at classifying those temporarily laid off as laid off rather than employed and absent from work. The Bureau of Labor Statistics said in a note that this resulted only in a 1 percent undercount, much smaller than the 3 percent undercount in last month’s report. So even if you apply that, the unemployment rate is down to 12.1 percent from 16.3 percent last month.

By contrast, first-time jobless claims remained high, hovering at around 1.4 million for last week. And continuing claims increased slightly. Over 31.1 million Americans are taking jobless benefits right now, either regular benefits or the special “Pandemic Unemployment Assistance” for gig workers and freelancers

Read all of our Unsanitized reports

The difference is clear: the June jobs report references a moment when state governors were allowing restaurants and bars and retail shops to reopen, and the first-time jobless report references last week, as governors in many states put a pause on that process. This only worsened this week. It shows the difficulty of using a monthly barometer like the jobs report to supply information during a fast-moving pandemic, which can rebound and change the economy much more quickly than a month at a time.

In the face of this, Senate Democrats just unveiled a plan to tie expanded unemployment to the monthly jobless rate.

That seems… wrong for this particular crisis? The bill, from Sens. Ron Wyden (D-OR) and Chuck Schumer (D-NY) would start phasing out the $600/week boost when unemployment in a particular state falls below 11 percent. The state numbers come later in the month, but if the national rate is tentatively at 11.1 percent, many states are likely below that. And if you only get a monthly snapshot, you’re not going to catch up to increases in cases necessitating lockdowns. You could have a worker’s unemployment check drop for a month at the precise moment when it’s more difficult to get a job. If we had a vibrant public job sector, maybe this would be fine, but we don’t.

In fairness, the Wyden/Schumer calculation is based on a rolling three-month average, which would smooth out the kinks a little bit. And the decrease is only $100/week for each percentage point, until it fully phases out once unemployment hits 6 percent. The three-month lag, of course, could mean that “good” months linger in the calculation when the state of things has radically changed. And an individual seeking work is disconnected to everyone around her seeking work, yet her unemployment benefits will rely on the level of unemployed, which is a little odd.

The proposal also maintains gig/freelancer benefits until next March, and extends benefit weeks, both with triggers tied to the state unemployment rate. The biggest impediment to “automatic stabilizers” like this has been the Congressional Budget Office, so good for Wyden and Schumer for not waiting for, or worrying about, a score. And in general, I prefer automatic stabilizers too. But they’re inevitably tied to a lagging indicator when the virus moves quickly. In the current environment, with unemployment replacing so much lost income, it might create a yo-yo effect, with lowered benefits leading to tighter consumer spending leading to higher job loss leading to higher benefits.

Automatic stabilizers that rescue people in moments of crisis happen to have solid support across the country, with 79 percent support according to a Groundwork Collaborative poll. And three-quarters of voters want expanded benefits to continue right now. But Senate Republicans want expanded benefits to die, and the outdated June jobs report will give them ammunition. Even though we haven’t even climbed more than one-third of the way out of the hole created by pandemic job loss, the direction of the gains will create a narrative of success that will spur Republicans to cut off benefits.

The narrative is that expanded benefits are a disincentive to work, which doesn’t at all square with 7 million people going back to work in May and June. Ultimately, the virus, not the level of unemployment benefits, is dictating the state of the job market. We need a policy response to fit those particular circumstances.

The Missing Rise

Case counts reached 50,000 in the U.S. on Wednesday; death counts remained at 697, lower than a week ago. This anomaly has been covered in this space for a couple weeks, and I wanted to do a slightly deeper dive on it.

So I went to the COVID Tracking site and pulled out the data for the fifteen states with the most intense rises in cases right now, to see if the national aggregate was hiding any state spikes in deaths. Of those fifteen states, only three—Arizona, Texas, and South Carolina—showed anything close to an increase in deaths. In the other twelve, deaths are sloping down, despite cases rising for a month. For South Carolina the increase is barely visible. Arizona and Texas do seem to be having more trouble, though in Arizona that’s driven by two outlier days (one yesterday) with a lot of deaths, and in Texas really by yesterday’s total (otherwise you’d see a plateau). Meanwhile, case rises and even hospitalizations all happened a month ago.

If hospitals can’t handle patients in these states more deaths will occur by default, which is why it’s imperative to get a handle on cases. And the real death toll is just obviously higher than the official number; at best we’re looking at trends.

Everyone is bracing for the worst, which is appropriate. But it’s really the case that treatment has improved after six months of dealing with the disease, that medical interventions are happening earlier with better testing, that older people are being more wary of infection and higher caseloads are seen in the young. This pandemic is bad news if only 100 people die per day, and we’re far from that. But I just wouldn’t expect the kind of numbers we saw in April.

Days Without a Bailout Oversight Chair

97. But I’m not alone! Public Citizen and 27 other groups have written to Congressional leadership, demanding a chair to the Congressional Oversight Commission without further delay. The lack of a chair means no staff, no hearings, and oversight that’s muted at best. Given that this was a rallying point for Democrats on the CARES Act, it should be embarrassing to their leadership that this position remains unfilled.

Meanwhile, the Wall Street Journal blares that junk bonds are suddenly underperforming (they’ve been in the same range for a month). Yet collateralized loan obligations tied to junk bonds are soaring, thanks to the Fed backstop. This could be a subject of inquiry for the bailout oversight chair!

Today I Learned

You can count on the Prospect, can we count on you?

There's no paywall here. Your donations power our newsroom as we report on ideas, politics and power — and what’s really at stake as we navigate another presidential election year. Please, become a member, or make a one-time donation, today. Thank you!