People rarely stop to think about just how bizarre our unemployment insurance system actually is. Imagine if automobile insurance worked as follows. After an accident, the insurance company paid you $400 dollars for each week that your car was out of commission, up to 26 weeks.
That system would obviously encourage people to delay repairing the car. This is why insurance companies pay the insured a lump sum, right after the accident. I’ve long advocated the same system for unemployment insurance, in order to reduce work disincentives.
Instead, the disincentive to work was actually increased by the UI reforms in the recent CARES Act. Some of the changes may have been beneficial, such as extending UI to independent contractors. But it’s hard to justify paying people who don’t work more than they received on their previous job, even during times when jobs are hard to come by.
A new study by Attila Lindner and Balázs Reizer suggests that paying UI as a lump sum after the “accident” of job loss could be a win-win proposition. Here’s the abstract:
We estimate the effect of front-loading unemployment benefit payments on nonemployment duration and reemployment wages. Exploiting a sharp change in the path of benefits for those who claimed unemployment benefits after November 1, 2005 in Hungary, we show that nonemployment duration fell by two weeks, while reemployment wages rose by 1.4 percent as a result of front-loading. We show that these behavioral responses were large enough to offset the mechanical cost increase of the unemployment insurance. We argue that our results indicate that benefit front-loading was a Pareto improving policy reform as both unemployed and employed workers were made better off.
READER COMMENTS
VP
Jun 30 2020 at 4:27pm
I misunderstood your analogy at first. I thought that would be a great incentive for the insurance companies to fix your car quickly, especially in those instances when insurance companies are squabbling over who is at fault in an accident. But then I grasped the point you were making. I suppose it depends on who pays the premiums and who is responsible for making sure the repair happens.
I like the lump sum idea. I’m going to find a job right away if I know I can pocket my benefit.
Thomas Hutcheson
Jun 30 2020 at 8:01pm
To begin with, I doubt that if accident insurance paid you a fraction of the monthly value of your cars’s services, I do not think people would delay getting the car repaired in order to receive a greater payment. Now if unemployment insurance paid more (including heath insurance) maybe people would delay.
Lump sum payments could work if you got an amount based on an estimate of how long unemployment would last in a given state or locality, x1, and if the person was still unemployed at the end of that period he received another payment based on another estimate based on statistic of how long people unemployed for x weeks would remain unemployed, x2, etc. This would mean the person would always receive more from not working than working
It seems much easier to just replace 80-90 % of lost income plus health insurance; that would insure that the person never receive more than they lost. Also, I’d finance the scheme with something besides a wage tax.
Scott Sumner
Jul 1 2020 at 1:07pm
You said:
“To begin with, I doubt that if accident insurance paid you a fraction of the monthly value of your cars’s services, I do not think people would delay getting the car repaired in order to receive a greater payment.”
You don’t believe that demand curves slope downward?
Thomas Hutcheson
Jul 2 2020 at 11:21am
This ignores the cost of not having the services of the car. So long as the periodic payment is less than the periodic cost of not having the use of the car, there is no incentive to delay the repair, just like unemployment insurance.
Thomas Hutcheson
Jul 1 2020 at 8:09am
I do not understand how paying out the insurance benefits for an accident on a daily or weekly basis would lead people to delay getting the car repaired (except if they would have to save up to afford the repair. People want to have the services of their car/have a job regardless of whether they are insured.
Mark Z
Jul 1 2020 at 3:43pm
Because they get more money the longer they wait to repair their car.
Dylan
Jul 1 2020 at 7:05pm
I think Thomas is pointing out that for many people, they wouldn’t actually get more money by waiting to fix the car (assuming it isn’t usable in the interim) because they depend on the car to get to jobs which they might lose if they don’t have access to the car, or have to rely on more expensive and less convenient forms of public transportation, etc… More generally, if we are to assume that the person was rational in spending say $20,000 to buy a car based on the benefits it gave them, then it seems like they would be irrational to take longer than needed to get the car fixed when there are much lower sums at stake.
Of course the persons circumstances could have changed from when they bought the car to the time it was damaged, so no doubt on the margin the time it would take for repairs to be done would go up, but likely the majority would continue to make the calculation that the cost of going without is larger than the benefit they get from delaying repairing the vehicle.
I don’t think the analogy really holds up for UI though. There’s not much of a downside to having a functional car. Work on the other hand is something that lots of people don’t really like much, if you’re able to change the calculus a bit to make not working more financially attractive, plenty of people will take that offer, even if they are making a lot less than they would by working.
I wish the bloggers here would stop making this statement as if the actually was some big surprise or revelation. The CARES Act was specifically designed to make it so that not working was more attractive than working for lots of people, because they didn’t want tons of people out looking for work and spreading the virus. This might have been misguided, but it was pretty explicit based on the statements of the politicians that were working on this, so it isn’t just some unintended consequence.
Scott Sumner
Jul 2 2020 at 1:51pm
I don’t use my car for getting to a job, and there are millions of other people like me.
Dylan
Jul 4 2020 at 8:25am
But you bought your car for some reason, and presumably that cost was more than the $50 a week or whatever you would get from spreading out an insurance claim over 26 weeks? Given that the average new car price is around $35,000, and average used cars go for about $20,00. So the average car buyer seems to value the services they get from their car quite highly, whether they need it for going to work, getting to the grocery store, taking kids to daycare, or the millions of other things people use a car for. As long as the value that people get from the car is greater than the value of any spread out insurance payments, we would expect the delay in spreading out payments to not really impact the time it takes to get the car back in working order.
So, as I said in my original post, I expect it would have impact on the margin, but not much on the majority of people making claims. Even for those that don’t have the upfront cash available to make a claim, they should be able to get a loan based on the sum of the payments that are due to them.
robc
Jul 1 2020 at 8:17am
I was unemployed for two weeks last fall. I only received 1 week of benefits because of the way my state calculates it. Getting a lump sump would have been really nice, especially since I had 3 months of severance coming too.
My situation was weird, as I found out in May that I was getting laid off as of Sept 30. I only got the severance if I worked until then, I was offered my new job in late August and set it up to start in mid October so I would have time to move and get settled (I actually drove a packed u-haul to my last day of work, I signed my paperwork, went to the unemployment office and filed, and headed out for a new state).
Dylan
Jul 1 2020 at 2:58pm
I’m surprised you bothered filing for UI under those circumstances, and somewhat surprised that you were granted it. I know regulations differ per state, but I had to sign a statement that I hadn’t been offered a job yet each time I made my weekly claim while unemployed last year.
robc
Jul 2 2020 at 7:55am
I was surprised I got it too, but apparently KY allows it. I for my two weeks unemployed, I had to put down that I was searching for a job each week. I used two of my interviews from August (one being the job I got), apparently they worked for them.
Because my job went overseas (my former company sent about 80% of the IT department to India) and my position didn’t exist otherwise locally, I qualified for TAA (Trade Adjustment Assistance) relocation benefits*, but IIRC, had to file for unemployment in order to apply for them.
*there were other TAA benefits available, but getting the new job early meant I didn’t qualify for any of them.
David W
Jul 1 2020 at 8:34am
How do you avoid moral hazard here? The automotive insurance company has the advantage that cars are in a liquid market; thus they know the value of your car, and can slap a deductible on, then give you a check that almost makes you whole. This makes it unprofitable to deliberately cause accidents, except in rare cases that require contortions that can be investigated.
In the employment market, though, it’s very hard to tell whether you have been caused a 26-week full-salary hit, or a 2-week vacation, particularly from the government’s perspective. Everyone is different.
If you set the payments at a single lump sum, it will compensate many people more than the damage they have been done. Thus it will tempt many people to cause their own firing. Rather than the tradition of the two-week notice, we might end up with the tradition of the unexpected sabotage, destroying value.
The point of the week-by-week low payout structure is to avoid moral hazard in the face of that lack of knowledge. Unemployment softens the blow, but it very rarely makes you better off than if you had kept your job. Thus the government doesn’t have to investigate fault. Your payments will be proportional to the difficulty you have getting reemployed, and can begin immediately. It’s not perfect, as you point out, but better a system than incentivizes a bit of slow searching among a couple percent of workers than one that incentivizes active damage by all workers.
robc
Jul 1 2020 at 12:23pm
I guess we would need to look more into the Hungary system, as they don’t seem to have the problems you suggest. But maybe they do?
Also, moving UI from a government program to private insurance would lead to a solution to those problems. The private insurance companies would figure out the system that works. Or fail trying.
David W
Jul 1 2020 at 12:52pm
Sumner is extrapolating a different system than the one in Hungary which was studied. The Hungarian system change is
Note this is not a lump-sum plan, and does not provide data on responses to a lump-sum plan.
robc
Jul 1 2020 at 2:50pm
While it isn’t a perfect example, it is sometimes as good as you can get in the social sciences.
it isn’t as extreme as a lump sum, but it is part way there by shifting a large chunk of the payments into the first 90 days.
Scott Sumner
Jul 1 2020 at 1:09pm
That’s why we do academic studies, such as the one I quoted here. To answer those questions.
David W
Jul 1 2020 at 1:53pm
You are extrapolating from the study, which does not include any data on your lump sum proposal. See my response to robc above.
Scott Sumner
Jul 2 2020 at 1:53pm
Right, but my proposal is a more extreme version of front-loading.
nobody.really
Jul 1 2020 at 12:40pm
Oh, it’s bizarre–and for reasons not addressed in this post. I find little objectionable about a private citizen taking out insurance on her own income stream (AFLAC!). But what interest does government have in this matter?
Sure, government may desire to help people living in poverty–and thus may have a social insurance system. But why should government privilege someone who recently lost an income over someone who has long struggled to secure an income? Arguably, unemployment insurance exists to promote STABILITY over EQUALITY–that is, to help maintain a person’s income status, whether high or low. It looks like a regressive system to reinforce social stratification.
So what’s the point? Arguably, government wants to encourage people to make long-term investments (such as taking on debt to buy a house), which arguably makes society better off in the long run. A system to make incomes more stable promotes that objective. I guess.
The objections we might raise regarding unemployment insurance generally may not apply to the CARES Act. As I understand it, the point of the CARES Act was to enable people who could not earn a living from home to nevertheless stay home to reduce the spread of coronavirus. Thus, this disincentive to work was not a bug; it was a feature.
Scott Sumner
Jul 1 2020 at 1:10pm
I agree that a private system makes more sense. And we do have private UI of various sorts.
robc
Jul 1 2020 at 2:48pm
I was looking at this recently and the two companies that were offering private UI have both shut it down. One is entirely out of the business and the other is not accepting new policies.
The government monopoly is making it hard for the private insurers.
Michael Sandifer
Jul 2 2020 at 1:51am
I prefer a negative income tax with a UBI component. When unemployed, one would be able to draw against future UBI payments up to the net present value of those benefits, calculated on the basis of an average life span. Could make it no fault, so people could use it to help start businesses.
Comments are closed.