Wednesday, October 9, 2013

Mulligan on Obamacare Marginal Tax Rates

Casey Mulligan wrote a nice Wall Street Journal Oped last week, summarizing his recent NBER Working Paper (also here on Casey's webpage) on marginal tax rates.

What do I mean, tax, you might ask. Obamacare is about giving people stuff, not taxing. Sadly, no. Obamacare gives subsidies that are dependent on income. As you earn more, you receive fewer subsidies for health care, reducing the incentive to earn more. Casey tots this sort of thing up, along with the actual taxes people will pay.

Economists use the word "tax" here and we know what we mean, but it would be better to call it "disincentives" so it's clearer what the problem is, and just how painful we make it for poor people in this country to rise out of that poverty.

As you can see, the average marginal "tax" rate went up 10 percentage points since 2007, and about 5 percentage points due to Obamacare alone.

Going back to the working paper, I think this is actually an understatement. (Probably the first time Casey or I have ever been accused of that!)

First, not even Casey can add everything up, and it all adds in one direction. State and local taxes, and vast number of state, county, city and other income or asset-based transfers all add. I haven't read all his papers or the whole book yet, but did Casey get them all? For example, I just got in the mail notice for a little program offered by the state of Illinois to lower your property taxes if you earn less than $100,000 per year. Nice, but one more little incentive not to earn more than $100,000 per year, and not in Casey's calculation.

In email correspondence, Casey pulled me back from these thoughts in a way that is revealing about the calculation. I wanted to add sales tax. After all, if you earn a dollar, but you have to pay 10% sales tax to do anything with it, that's another 10% distortion, no? Casey responds no, because he wants to measure the income-compensated distortion to labor, period. If you don't work, and somehow you also get income, you still have to pay the 10% sales tax. So the sales tax does not distort that pure work-no work decision. Casey's right, but I think this clearly illuminates the conservative nature of his calculation and what it means. There are a lot more margins, wedges, and distortions out there, and he's not trying to measure them all. He's also not trying to measure the wedges and disincentives for employers to hire people, towards non-market activities, and certainly not the effects of the regulatory tangle.

Another note of conservatisim: "The results account for the fact that many people will not participate in programs for which they are eligible." This is an important issue, that at least had not sunk in for me until reading a recent CBO report. People don't sign up for all the benefits to which they are eligible. If they did, marginal tax rates would be astronomical. It also sends a warning: Just how long will it be before people in an increasingly stagnant economy figure out all the programs they are eligible for?

Drawing a single line can also be unduly calming. You might say, "well 50% isn't so bad. Europeans still work, sort of, paying 50% marginal tax rates."  But as Casey reminds us, the spread in marginal tax rates across people is enormous. For example, the paper has a nice example (p. 13, table 2) of how a typical earner will come out ahead by choosing to work part time and receive subsidies, rather than work full time. This is a case of a 100% marginal tax rate.

It's likely that the effect of marginal taxes is nonlinear. Much of the labor decision comes in chunks: work or don't work; work part time or full time; apply for benefits or don't, with transactions costs and irreversibilities.  Suppose half the population feels a 100% marginal tax rate and half feels zero. Half the population works, half does not. That is likely a much larger effect than if the whole population felt a 50% marginal tax rate.

In sum, it's probably worse than even Casey's graph. But Casey is doing the right thing in putting up a carefully documented graph and paper rather than speculating. Speculation is for blogs, not papers and not for good opeds.

And Casey's big point remains the additional effect of Obamacare and other changes to Federal programs. Whatever you think the level is, it's now 10 percentage points more than what it used to be. On average.


  1. Ive always been on the fence with the "disincentive to work" argument over [relatively] smooth tax profiles. We work, not only for remuneration, but also out of ambition and to uphold social norms. Also, as you note, supply decisions are "chunky". I believe the literature supports relatively small supply elasticities. One of my main gripes is with the structure of subsidies. At certain points (say, 400% of the poverty level) there is a discrete drop in after-tax/transfer income. I would think that this is where you get bunching, and it would not take rocket surgery to plan around. Regardless of how you feel about ACA, this has got to be disappointing.

    1. "We work, not only for remuneration, but also out of ambition and to uphold social norms. "

      I agree; however, this warrants close examination.

      First, as regards "ambition", I think one's "ambition" decreases in proportion to the diminishment of income one is allowed to retain from that work effort. But, OK, I accept that you mean "ambition irrespective of remunerative reward". I agree that people are naturally competitive and hierarchical.

      However, this brings us to upholding "social norms". As the rewards for work decrease and the rewards from receiving entitlements increase, does that not change the prevailing "social norms"? If "everyone else" (all my friends, etc) is collecting welfare the "social norm" of working is no longer necessarily the "norm". And, as the norm changes, one's "ambition" in the non-remunerative sense also changes.

      I think this aspect of incentives is understated and under-appreciated by economists. Higher "disincentive rates" for work not only discourage work but also tend to change social norms and "ambition" independently of the raw arithmetic. It is a sort of reinforcing or reiterative cycle. Critics would likely be quick to say this is a "moral judgement", but this isn't a moral judgement at all---it is merely an observation of collective human responses that I suppose could properly be confirmed by sociologists.

      As an example, I've always thought that the reduction in smoking is not so much attributable per se to the high excise taxes meant to discourage it, but to the fact that the "social norm" changed such that smoking became somewhat unacceptable socially. Ostracization is a strong social force. A number of people quit smoking due to the expense involved, I'm sure; however, the diminishment of the number of smokers in turn changes the "norm" in a fashion that reinforces the trend. Conversely, my casual observation is that various types of drug usage, legal or illegal, have become increasingly more prevalent as social taboos diminish.

  2. "First, not even Casey can add everything up, and it all adds in one direction."

    This observation had to have been lifted off my "to do list": I was about to contact Mr Mulligan and plea with him to see if he could take the analysis further. I would like someone to include the effects of other major public (and private) redistributive schemes, including not only those in the category of state and local incentives (some cited above), but also the effect of income-based higher education support, the implied redistribution in social security benefits (ignoring for now the impact of proposals to means-test, and reduce to zero, these benefits), etc.

    Whether all of these implied marginal tax rates can be ceritified to disincentivize away from income producing work anyone affected, all of the time is fatuous. It must do, for most, some all the time. Also, a comprehensive view of how "progressively" success is treated in America would be a useful couterview to conventional wisdom.

  3. "just how painful we make it for poor people in this country to rise out of that poverty."

    I'm not sure I understand how disincentivising rising out of poverty makes it more "painful". These programs make poverty less painful, and so diminish the marginal benefit of exiting poverty, but they don't lower anyone's standard of living except by choice.

    1. I totally agree. That line made no sense.

  4. "Obamacare gives subsidies that are dependent on income. As you earn more, you receive fewer subsidies for health care"

    Sounds familiar.

    "From each according to his ability, to each according to his need".

  5. If various welfare programs are disincentives to rising out of poverty, then this seems to suggests that incentives to not be in poverty in the first place, prior to the establishment of social welfare programs should have been adequate for eliminating poverty.

    But they weren't. We've always had poverty, and probably always will. When you say, Dr. Cochrane, that these programs are "disincentives", what you really mean is that they allow a low income person to live a more middle-class life. And I, for one, don't see any problem with that - it means that they're successful!

    Imho, one of, if not *the* biggest contributor to poverty, or income disparity at all, is *innate* intellectual capacity, ie, agents have little control over how capable their potential earnings are. Am I really willing to tell them that because of this, they can't live a more middle-class life? Morally, no. Morally, I'm not sure how you or anyone else can for that matter.

    1. "the biggest contributor to poverty, or income disparity at all, is *innate* intellectual capacity"
      Do you realize that you take poverty and income disparity as synonyms!!!!
      Like Margareth Thatcher said: "You rather the poor be poorer provided that the rich are less rich" That is a profound philosophical difference.

    2. Of course, that Thatcher quote sidesteps my argument by assuming that "Disincentives" to work are the very thing that prevents working class folk from rising to the top whereas, in reality, it has *far* more to do with intelligence.

    3. "Rising to the top"
      You see, my point is that your biggest concern is income distribution. Poverty is an absolute concept, not a relative one. Its true that free market capitalism creates great income disparities, but its undeniable the best system for the poor. If not, then why is it that so many millions are so eager to migrate to the USA? Only someone who has lived his whole life in the USA can't see this. Why do you care so much about rich people? If Steve Jobs creates Apple and becomes super rich in which way is that bad for you or the poor?

    4. Actually, poverty is most definitely relative. Indeed, even the Tatcher quote (apart from being a total strawman) emphasizes that very relativeness.

      As to Dr. Cochrane blog-post, I believe Jefftopia's reply is perfect. Show me how statelessness in Somalia has so perfectly eliminated poverty... (I disagree with Jeff, though, on the innate intelligence stuff but that's another topic entirely).

      I would only add that very few of us can actually vary our supply of labour according to incentives:

  6. While policies over the last few years may cause disincentives to work through marginal tax rates, haven't there been some trends over the past couple of decades that have increased the incentives to work? For example:

    More and more people have credit card debt. Thus, increasing your income can be especially affective at leaving more money in your pocket.

    Having more liquidity by working more reduces the impact of overdraft and other fees.

    I would be interested to know the total value to earning an extra $1 today relative to 20 years ago. Even if that dollar is taxed at a slightly higher rate, it might be particularly valuable today to earn that extra $1.

  7. "Casey responds no, because he wants to measure the income-compensated distortion to labor, period. If you don't work, and somehow you also get income, you still have to pay the 10% sales tax. So the sales tax does not distort that pure work-no work decision. Casey's right,"
    I think Casey is wrong and you were right. The sales tax means you get less consumption for giving up an hour of leisure, distorting your work-leisure choice. Say we only had the sales tax and no other taxes - is Casey saying it would be non-distorting?
    Say we put on a sales tax and compensated. the point is that it would cost more to compensate for the tax than we would raise in revenue in the compensated equilibrium. that is the deadweight loss from the tax and it is distorting.

    By Casey's logic the income tax is not distorting, as you would still pay it on your income compensation

  8. It would be a big deal if labor supply elasticities were large. But they are not.

  9. You make a wonderful case for Medicare for All, which would remove the means testing and the associated marginal rate distortions that are almost certainly worse than those caused by a hypothetical expanded Medicare tax.

  10. I am not an economist, just an independent working guy whose health insurance was cancelled under ACA and wound up with the cheapest bronze plan with a 40% higher premium, a 50% higher deductible and a wife unhappy because she can't use her primary care doc of 28 years. I can't follow Dr. Mulligan's calculus. But I can go to to calculate subsidies. There I find that for a family of 3 wanting a silver plan, for example, if your Adjusted Gross Income is $45,960 you get just over $1000.00 per month in subsidy. If your AGI is $45,961, your subsidy is $0.00. If that were my income level, I'd say "no" to any raise my boss wanted to give me less than $12,000+ (to make up for the extra income taxes on the higher wage). And a boss is certainly unlikely to hand out a raise that's so large--25% of current wages. I don't think any further proof of a disincentive to work harder is really needed.


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.