David R. Henderson  

Taxation of Unemployment Benefits

Separation of Health and State... How to Evaluate Government Pay...

On the comments yesterday on my post on the effect of unemployment benefits on unemployment, Noah Yetter wrote:

You have to pay taxes on unemployement benefits? Am I the only one who finds this to be unalloyed lunacy?
I actually find taxation of unemployment benefits to be "alloyed" sanity. A little background, and I'm summarizing what you can find here.

In 1979, the feds began to tax unemployment benefits for people with more than a certain level of income. To avoid the hundreds of thousands of percent marginal tax rate on one dollar past that threshold that I highlighted about Megan McArdle's proposal, the feds phased in the taxation. In tax economics jargon, they avoided a "notch" and instead created a "kink." In the 1982 tax law, the thresholds were lowered and in the 1986 Tax Reform Act, all unemployment benefits became subject to taxation, thus avoiding not only a notch but also a kink.

Why do I find this sensible? For the same reason the Joint Committee on Taxation did: to reduce the incentive that unemployment insurance gives people to remain unemployed. It was Martin Feldstein in the early 1970s, by the way, who popularized this thinking.

My guess is that Noah finds this crazy because he's implicitly assuming that people who get unemployment insurance are poor or very low-income. Not true. When I was at the Labor Department in 1982, I learned that a large percent of the people who get unemployment benefits are high-income and that for most people who get unemployment benefits, the benefits are substantially less than half of all family income. This is for two reasons. First, most unemployment is short term (although longer-term during recessions and even longer-term during this recession, due, in part, to the length of time for which you can receive benefits.) Second, other family members have income.

Because the Tax Reform Act of 1986 raised the amount of income you could have without paying any tax, someone whose sole income was unemployment benefits and who received such benefits for many months would pay zero or little federal income tax even with unemployment benefits counting toward taxable income.

So why do I say it's "alloyed" sanity? Because I think we shouldn't have a government-run unemployment insurance system. Moreover, the term "unemployment insurance" is a misnomer. It's really a tax-and-subsidy scheme.

Comments and Sharing

CATEGORIES: Labor Market , Taxation

COMMENTS (15 to date)
John Jenkins writes:

I am betting he is thinking that it doesn't make a lot of sense to fund unemployment from tax revenues and at the same time tax the benefits when you could just reduce the benefits commensurately to get the same economic effect without the compliance costs.

PrometheeFeu writes:

Well, I think that the reason why it sounds crazy is because the government is paying you the unemployment benefit and then taking part of it away as taxes. That does sounds kind of stupid if you don't think about multiple incomes etc.

But honestly, I find your arguments quite weak. First, the idea that taxing unemployment benefits reduces the incentive to be unemployed. Now THAT is just a case of giving you stuff just so you can take it away. If that is the goal, just reduce the unemployment benefits instead of tagging more bureaucracy on top.

Second, the concept that "we shouldn't have a government-run unemployment insurance system." That is a valid belief to hold. But that is unlikely to have much to do with why things are the way they are now. I mean, we obviously have government-run unemployment insurance (the only kind possible since unemployment is otherwise uninsurable) and so the design of the insurance plan probably does not reflect the idea that the same insurance plan should not exist. If it does, then it IS insanity.

David R. Henderson writes:

John Jenkins and PrometheeFeu both missed the point. I recommend that both reread my post.

Noah Yetter writes:

No, John Jenkins has it exactly right.

One option is for the government to pay you $100 and tax back $15 of it. The other option is to pay you $85.

My contention is that the first makes absolutely no sense whatsoever and is obviously highly wasteful.

It doesn't have anything to do with your income level or incentives. It's purely an issue of efficiency (technical, not economic).

Fralupo writes:

Hi Noah,

You are assuming that the government knows ahead of time how much of the unemployment benefits will be paid in income taxes. That would require all sorts of coordination and foresight that would only complicate the government's job. Since the vast majority of people will have to file their taxes regardless of whether they receive unemployment benefits or not, it seems more efficient for the individual to do it.


bdm writes:

I don't understand why people aren't getting David's argument. Unless I'm misunderstanding it, his point is that tax rates are a function of income. Unless everyone had the same income, then reducing unemployment benefits by X dollars would create different payments - and therefore different incentives - than taxing unemployment benefits. Taxing unemployment benefits is a way of means-testing unemployment benefits without creating absurdly high implicit marginal tax rates.

David, why don't you think we should have government-run unemployment insurance? This seems like a legitimately insurable good (I want protection against the risk that my employer goes out of business, or there is a recession, etc.) that suffers from adverse selection (because an insurer wouldn't know - or would only know after very expensive research - who are the lazy, prone-to-be-fired and who are the hard-working).

Nick C. writes:

Instead of having a flat rate subject to taxes, couldn't we just scale the payout as a function of income or of tax burden? The payout structure could mirror the tax structure such as to avoid your "notches", and it may involve less red tape and bureaucrats.

It seems to me that all this talk about incentives and high marginal tax rates is largely irrelevant to the original point, though. It's possible to have a tax-free variable payout which delivers the same results that we have now, or a taxed payout which is a very high percentage of original income and destroys incentives. One can fiddle with the details all day and still not touch on which alternative is more efficient.

Assuming incentives and such can be dealt with, I suppose the question becomes: is it more efficient to pay f(income) once, or pay (p*income) and then send it through the existing tax system?

David R. Henderson writes:

Thanks, Fralupo and bdm,
You guys nailed it.

Now to bdm's question. Yes, you can have adverse selection, although that's a possibility, not a certainty. But a compulsory government program doesn't solve that problem--it just ignores it. What you will get is not insurance; it's just a tax and subsidy scheme. Also, you get a lot of moral hazard: I won't take this job because I can do better if I wait.

Tom West writes:

Because I think we shouldn't have a government-run unemployment insurance system.

I know it shouldn't matter, but it is tough to hear that from somebody who holds one of the most protected jobs in the country.

Is there *any* job more secure than a university professor?

Tom West writes:

Whoops. For some reason I thought this was Bryan's posting. My apologies.

David R. Henderson writes:

Tom West,
I have tenure so if it applies to Bryan, it applies to me. But you're right that it shouldn't matter. The argument, as you well know, is independent of the person making it. For what it's worth, I'm not an advocate of tenure.

Robert Johnson writes:

Great post! The post and the discussion together really elucidated this issue for me. I think you're right, David.

Paul Wilson writes:

I can understand both sides of the argument. I have been on one side at one time or another. I understand that by taxation of benefits you encourage ones who receive these benefits to find employment. I work for a unionized company that is driven by seniority and when the guys get laid off during the early months of the year they stay home and collect benefits instead of work. WHY? Because they can make more money that way instead of finding another job making 300-400 LESS weekly.
Now on the flip side, those working who collect unemployment whom collect substantially less amounts because of the lack of work at the same company (economy) throughout the previous year are forced to collect a meager amount and forced to try to make it work. I KNOW. I HAVE BEEN THERE. Taxation at this point is a very difficult thing to justify to that individual.
Where does the personal responsibility come in at? I think the problem comes from the "I deserve something because I'm me" attitude that's present in today's culture. If you can work, do it and collect what's leftover in benefits. Otherwise, if you make under a certain benefit amount DONT TAX IT.


John Jenkins writes:


I understood what you said. Maybe you should re-read what I wrote, where I was describing SOMEONE ELSE's position. Was I really unclear?

David R. Henderson writes:

John Jenkins,
Yes, I thought you were unclear. In describing his position, which I, and I think most readers, thought to be his position, I thought you were saying it made sense. My error.

Comments for this entry have been closed
Return to top