Normative Narratives

State of the Labor Market

Most recent Unemployment Insurance (UI) report (weekly, Thursday at 8:30am, last update 12/23). Government Accountability Office report on shortcomings of UI data.

BLS Jobs Report (monthly, First Friday of the month, 8:30 am, last update 12/4. State level data typically come out two weeks later, and metropolitan area data two weeks after that).

The number of people claiming UI as of the week ending 12/5 remained steady at over 20 million. The lead indicator, initial UI claims, are down a bit from last week, a positive sign. Congress also passed a new round of COVID relief, which is now awaiting the President’s signature. So things are looking good, right?

Not really. I think most economists would sum this bill up as “too little too late, but better than nothing”. For one thing it only extends enhanced UI payments and pandemic UI programs for 11 weeks, at which point only a small portion of the population will have been vaccinated and the job market is likely to look similar to what it does now. If we had a functioning Congress then no big deal, they’d come back to the table and pass more bill(s) as needed. Does anyone really think things will go that smoothly? As Paul Krugman put it:

“So while the new legislation provides a sort of bridge to the post-Covid future, it’s a bridge that spans only part of the chasm ahead. And the way the bill was passed offers few reasons to be optimistic about Republican willingness to let the Biden administration finish the project.”

The number of people claiming benefits for longer-term unemployment (Pandemic Emergency Unemployment Coverage and Extended Benefits programs) remains very high (the correct PEUC number for the week ending 12/5 is 4,793,230). This corresponds with BLS data showing the longer-term unemployed making up a larger share of the total unemployed, which makes sense as the pandemic drags into it’s 10th month. The program that serves those unemployed longest, the Extended Benefits (EB) program, is available based on a state’s unemployment rate; currently it is only available in 21 states, Washington DC, Puerto Rico, and the Virgin Islands. Given the obvious need there is absolutely no reason, in a bill of this size, that emergency funding for EB in every U.S. state and territory should not have been provided.


This brings me to a point I was making earlier in the pandemic–what is the “true unemployment rate”? Former BLS commissioner Erica Groshen provided analysis looking at over the year changes in a number of labor force metrics–some captured in traditional BLS labor underutilization measures and some not–compared to an adjusted labor force figure (the prior years labor force for that month, adjusted for population growth). The resulting number is not an unemployment rate per se, but rather a measure of the percent of the labor force disrupted by the pandemic.

I have taken Ms. Groshen’s analysis and ran with it. I made a minor adjustment to the misclassified section (“employed people, not at work”), excluding some sub-sections she included, and including some the BLS does not in their impact statement. For example, I don’t think less people on vacation should decrease the number of people disrupted–vacation disruptions are a labor force disruption caused by the pandemic (excluded in my number, included in Ms. Groshen’s who includes all the sections in the impact statement). On the other hand, I do think that changes in people not at work due to illness and childcare problems should be included, as both are directly tied to the pandemic and resulting school closures (which Ms. Groshen’s figure does include, but the BLS does not). The BLS only includes the “other reasons” category, which makes sense given that they are trying to isolate people who might truly be unemployed, which is a narrower group than those “displaced”).

See below for the figure I come up with, which I plan to update monthly going forward. It shows that as of November, 8.2% of the labor force has been disrupted by the pandemic. This is not comparable to any unemployment rate, as it only considers over-the-year changes from last year. It’s purpose is to isolate the effects of the pandemic. Other unemployment rates include these figures in full, keeping their definitions constant in order to be comparable over time.

Another important contribution of Ms. Groshen’s work is that is talks about both the “pandemic” and “recessionary” forces impacting the economy. Up until this point, pandemic forces have been strongest, as the pandemic rages and CARES act funding has helped bouy the economy. Once a vaccine is widely available recessionary forces will become more prominent (for example, the drag caused by layoffs to state and government workers should insufficient aid be provided to them, as experienced after the Great Recession).

Thinking about the economy this way is particularly illuminating right now. The pandemic has entered arguably it’s deadliest phase, which will cause people to stay inside and away from work (regardless of whether states enact stay at home orders or not). In addition, the expiration of UI benefits would cause people to have less buying power, a recessionary impact. In other words, should Congress fail to act, we could see a compounding of these two forces in a way we have yet to fully experienced, which could absolutely cause a feared double-dip recession.