The coming Unemployment Insurance bottleneck

The coming Unemployment Insurance bottleneck

By Matthew Yglesias

SlowBoring.com. 

March 25, 2021. 

Man singing at a protest with a sign reading "MITCH BETTER HAVE MY MONEY; extend unemployment insurance; save the $600"

(Photo By Tom Williams/CQ-Roll Call, Inc via Getty Images)

An acquaintance recently told me that he’d been offered his bartending job back now that things are opening up again, but he’s actually making more money thanks to bonus Unemployment Insurance. So what he’s going to do is just do some shifts unpaid and collect tips off the books.

This is not how the system is “supposed to work,” but in practical terms, it’s a win-win. I also heard a story of a restaurant owner who considered offering that deal to a former employee who doesn’t want his old job back, but the owner decided it’s not worth it given the risk of getting caught. So the guy is just going to continue collecting UI for a bit and see what happens.

Anecdotes are not data! And two examples aren’t even a proper trend.

But I do think there’s a real issue here, as a policy that was designed for a world of widespread restrictions on business activity is rapidly colliding with a reality that basically everything will be open all summer. No halfway decent person is going to say “Aha! We’ve got to stick it the unemployed by cutting benefits.” But bonus UI costs a bunch of money, and there are ways to restructure the program to be more helpful to people and to better fit the aid to the actual situation we face.

The weird Unemployment Insurance situation

Unemployment Insurance in the United States is, for bad historical reasons, run as a joint state-federal program for which the feds kick in some of the money, but all the administration and many of the rules are in the hands of the states.

The idea of UI is that if you qualify (which often isn’t the case), the UI system will replace some, but not all, of your wages. Due to decentralization, the replacement rate varies considerably, but it’s always pretty stingy — in part to keep taxes low and in part because conservatives tend to believe that generous UI replacement rates actually make unemployment higher by making people less desperate to accept new jobs.

Flash back to late-February and early-March of 2020 and it’s clear that a ton of people are going to lose their jobs due to the pandemic. Many countries adopt systems where, in effect, the government pays companies to not lay their workers off. American policymakers think it would be a better idea to simply let the UI system address this and make benefits more generous. Due to the nature of the crisis, the concern about the disemployment effect sort of went away, and Democrats (led, I believe, by Ron Wyden, who’s their top guy on the Senate finance committee) moved to both extend the duration of unemployment benefits and also increase the UI replacement rate.

But then state administrators start saying their systems can’t, technically speaking, just kick people up to a 95% or 100% replacement rate. So Democrats come up with the kludge of just adding $600 to everyone’s weekly benefit, calculating that this is both a round number and also creates an approximately 100% replacement rate for the average worker. Republicans briefly freak out that this is too generous; Bernie Sanders turns it into a thing on social media; and Steve Mnuchin basically tells congressional Republicans to shut up and pass the damn CARES Act before the economy collapses.

Doing it as a flat increase creates a weird situation where the people in the bottom half of the earnings distribution are better off losing their jobs than remaining employed. There are a lot of different ways to visualize this, but I like this chart by occupation — a lot of “lockdown” opponents painted themselves as friends to the retail and foodservice workers of America, but the fact is that with the CARES Act in place, these people were better off financially after losing their jobs, to say nothing of health risks.

Graph showing median replacement rate of wages for workers in different common occupations

Now interestingly, the best research available says this did not cause higher unemployment than you’d have seen otherwise. One reason is that the incredible generosity of the CARES Act helped bolster the demand side of the economy. Another reason is that there just wasn’t a lot of hiring of foodservice workers and receptionists during this time — if there aren’t any job openings, it doesn’t really matter if job-seekers are a little under-motivated. Last but by no means least, this CARES bonus UI was set to kind of randomly expire in the summer of 2020 with no clear plan to replace it. So many people still liked the security of a job.

Here we are a year later, and somehow none of the problems with state UI programs have been fixed. So the American Rescue Plan once again includes a UI bonus. This time it’s less generous — only $300/week — but now it’s set to run through early September. But unless something goes horribly awry, there should be plenty of hiring demand in May, June, July, and August. So the risk of a disemployment effect seems bigger than it was last year, even though the program is stingier.

The numbers on $300 a week

I asked Peter Ganong, the lead author of that paper I just referenced, about replacement rates under the new, less generous scheme. He observed that because the Rescue Plan also contains an increase in COBRA subsidies, in some cases it’s more appropriate to look at this as a $400/week bonus rather than $300/week.

This chart shows it both ways.

graph showing share of benefit eligibles in group based on a $300 or $400 supplement

The upshot here is that conservatively, 40% of the jobless are better off unemployed than back at their old jobs. That’s less because the bonus is so generous and more because of the way the pandemic shook out, where the unemployed are disproportionately low-wage workers.

Now I think the people who are getting a very-slightly-above 100% replacement rate would still rather have a job, since the bonus UI is going to expire.

But if you are currently unemployed and taking a job means a 20% or even 40% drop in income, then of course you’re going to hesitate. That’s especially true because working often comes with costs, especially things like childcare. If you’re young, it’s going to be extremely tempting to make this a Party Summer and then figure things out in September. If you’re not-so-young, then you can pocket your UI benefits and save yourself summer childcare burdens and then figure things out in September.

This is not the end of the world or anything, but it could be an undesirable roadblock to economic recovery.

The specter of inflation

To me, the biggest reason to worry about this is that we’re now living through a somewhat ridiculous surge of concern about inflation.

The basic fear here is that Biden and the Fed have done too much to stimulate the economy, and now we’re going to overheat. I don’t buy that. But I do buy that we will face some supply bottlenecks. One thing I’ve really missed during the pandemic is working in my neighborhood coffee shops. But three of those coffee shops have closed down over the past year. If everyone tries to resume our pre-pandemic coffee habits, there’s not going to be anyplace for us to go. If the owners reap extraordinary profits for a little while by charging a price premium, I say they deserve it.

But part of why it’s fine is that it should induce new coffee shops to open. One of the three closed places already has a new vape shop in the former space. But two are just empty, and there are other vacant storefronts around.

We want Jay Powell to say “it’s fine that coffee prices have gone up and nobody can get a seat because a bunch of new places are about to open.”

But if people don’t really want to take new jobs this summer, it will be hard for new businesses to open (and for existing ones to expand their hours), keeping some prices up, which could create a big overheating narrative that in turn slows the recovery and/or discredits expansionary policy.

It would be good to avoid that outcome, but especially good to avoid that outcome without being stingier to low-wage workers.

Some ideas for improvement

My favorite way to improve the current policy is pretty simple.

Right now, everyone is eligible for about 23 weeks of $300/week bonus Unemployment Insurance.

If you get a job tomorrow and don’t need those 23 weeks, you should get a $6,900 check.

If it takes a few months for you to and you only have 10 weeks of eligibility left by the time you get a job, you should then get a check for $3,000.

If you get a job two weeks before the bonus expires, you then get a check for $600.

This works for everyone. Since a job pays more than regular UI, you are always better-off financially saying yes to a job offer. Then, since taking the job triggers a big cash payment, it also works as fiscal stimulus. You could imagine doing this on a bipartisan basis or throwing it into a second reconciliation bill later in the year.

I see basically no downside to this idea and think it’s what Congress should do.

Another way to go would be to stretch the UI bonus out. You could say that starting in mid-June the bonus tapers down to only $200/week, but then instead of ending in September you get a $100/week bonus through the end of the year. That improves the work incentives while also doing more than the current plan to help the people who are most in need. I like the bonus idea more, but I hear from some people on the Hill that there is worry about the politics of such large checks going out to some people while essential workers who’ve been on the job this whole time get neither.

Either option would be a big improvement on the status quo. Also — Congress should fix Unemployment Insurance.

Let’s make this better

March 2020 was a crisis situation, and they did the best they could. There were a lot of administrative problems with people getting signed up, but they mostly ultimately did get signed up.

Still, more than a year later, we are left with fundamental flaws in our system:

The state administration doesn’t help achieve anything, and states seem to administer UI poorly. We should federalize it.

The actual underlying computer systems should not be so clunky and bad as to have made it impossible to simply increase the replacement rate.

Bonus UI as a recession-response is highly effective, but it should be made automatic and conditions-based rather than relying on arbitrary timetables.

To just discuss that last point, we often have a tradeoff in stimulus between targeting (which is good because you want effective stimulus) and universality (which is good for popularity and reaching more struggling people).

Bonus UI is both.

Giving money to the jobless is very good targeting because they have immediate financial needs. But Unemployment Insurance as a system is a universalistic program that everyone can use when they need it. The way this ought to work is that when the ratio of job-seekers to job-openings is high, the replacement rate gets more generous. That’s a time when the economy needs stimulus more than it needs to worry about search effort. Then when the ratio normalizes, it should become less generous (though probably still more generous than average benefits today) as we worry more about incentives and less about stimulus.

Democrats kept considering doing automatic stabilizers and then dropping the idea, worrying about the CBO score if you didn’t have programs arbitrarily phase out. But thanks to a fairly rapid pace of vaccination, it now looks like an automatic program probably would have been cheaper than a schedule-based one. Which is a reminder that while there’s room in life for cynical political calculation, when faced with genuine uncertainty, the fallback of doing the right thing is always a good option.