Monday, September 10, 2012

How not to blow it with phase-outs

Today's Wall Street Journal article, How Not to Blow It With Financial Aid, appparently about financing college education, has important lessons for the ongoing grand fiscal debate.

The article is about college financial aid, especially federally funded, and unwittingly exposes the atrocious incentives of the system.
"Every $10,000 reduction in income is going to improve your aid eligibility by [about] $3,000" if you have one child in college..

Every dollar a child has in assets—that includes bank accounts or trust funds—cuts their possible award by 20 cents. Every dollar a child makes in income above $6,130 (the limit for 2013-14 aid) cuts their possible award by 50 cents...

Every dollar a student gets from a 529 plan owned by other relatives is considered income to the student and reduces potential financial aid by 50 cents if the student is above the income threshold
So, we're looking at 30, 20, and 50 percent additional marginal tax rates on income or savings -- how much you lose if you make or save an extra dollar. The journal is full of useful tips to game the system.

Now, to the larger question. What matters for economics is the total, marginal tax rate. If you earn one extra dollar, and then spend it, how much stuff or services do you actually get -- after all taxes are included, including payroll, federal income, estate, excise, state income and excise, local, sales or property (if a good) or a second round of income and social insurance taxes (if a service). Ideally, we'd add in the burden of taxation incorporated in product prices too -- if the government taxes a business and that raises the price of what you pay, that's just like taxing you for buying the good.

And phaseouts. Phaseouts of deductions and eligibility for benefits count the same as taxes. If earning an extra dollar lowers your eligibility for college financial aid by 30 cents, that that is exactly the same thing as a 30 cent additional marginal tax rate.

In our grand national discussion about taxes, I seldom see any attempt to add this up -- to addess the total marginal rate, including all sources of taxation and deduction and benefit phaseouts. (And all questions of margins and distributions should include benefits in the same breath with taxes. If the tax system were flat and the government writes checks to people with lower income, that's the same progressivity as if the tax system were progressive and spending were not.) The debate over Federal taxes focuses on the headline federal income tax rate, 35 vs. 39%. But that is basically meaningless. The extra 30% marginal rate for people sending kids to college has not been mentioned once. Or the effects on the economy if people follow the Journal's advice to intentional impoverishment.

I wish I got to ask a question at the Presidential debate, of both candidates: Sir, what do you believe is the highest total marginal tax rate any American should pay, including all sources, phaseouts, and means-tested benefits? What do you think that rate is now? What steps will you take to ensure that no American pays a  higher rate?  Just asking the first question should send the fact-checkers on a field day and we'd start talking about how high marginal tax rates are now.

Even the Journal gets this wrong on occasion. Here's a graphic, from today's Hollande-Arnault story

No, dear Journal, this is not even close to the "top marginal rate." To start with, the US has state and local taxes, which Europe does not. And Europe has fewer phaseouts.

And now the conundrum of tax policy: As you can imagine, I'm a big fan of lowering rates and paying for it by eliminating deductions and tax expenditures. If I were in charge, the mortgage-interest and charitable contribution deductions would be gone, as would the deductibility of employer-provided group health plans. Now you know why I'm not in charge.

In this weekend's Meet The Press interview, Gov. Romney said he wanted to "limit deductions and exemptions for people at the high end" only.   Well, phasing out deductions is the same as a marginal tax rate.  If earning an extra dollar lowers, say, the deductions you can take on your existing income by 50 cents, that counts as a 50% marginal tax rate every bit as much as  if we just take the money.

I haven't done the numbers, but that's going to make it  harder to lower the economically meaningful marginal rates. It would seem far better to give the lower-rate-and-no-deductions offer to everyone.

The extraordinary complexity of the tax system is also curious. Perhaps the hope is that, since no economist seems to be able to calculate the true marginal rate, and people seem not to talk about it much, that nobody notices and so ignore the disincentives. Alas, there is an army of tax lawyers who are very good at this sort of thing, and even Wall Street Journal articles can advise that the year before your kid goes to college might be a good one to take that round the world tour rather than make any money.


Casey Mulligan wrote to let me know his upcoming book "The redistribution recession (cover at left, link to Amazon) undertakes a lot of marginal tax rate calculations, finding a maximum over 400 percent,  "over a pretty wide range of income, although applicable to a small percentage of the population." It will be on top of my stack as soon as it comes out.

A physician wrote with a comment, on my statement that economists don't calculate total marginal rates often enough:

..I certainly did. As a solo general surgeon in private practice, in 2004, with a gross business income before taxes of roughly $500K, I figured that the 39.6% Federal + 9.98% state top income tax rates + 6% [state] sales + Medicare which no longer peaked out, + property taxes, medical license fees, malpractice fees which were already at $100K for me and headed higher, and no scholarship help for the 4 out of 10 kids in college at the time, my marginal rate was somewhere north of 70%. Once I 'retired' from surgery and became a biology professor, making around $50K, my gross income was one tenth as much, but now one of my kids got a full-ride scholarship at [University], another got a half-ride scholarship, and another got a couple thousand that would not have been given under my earlier circumstances. By my 'going Galt', I figure that the .gov took at least a $200K hit (I remember previously paying $161K in fed. income taxes alone), whereas my disposable income was only about half of what it had been before. So you can bet that we non-economists, with all the individually detailed information at our disposal do indeed make these kinds of calculations, even if they're tough for economists to do in aggregate. At least for me, the argument that a simple 36% federal income tax is below the Laffer curve hump is lame, given other factors


  1. "If I were in charge, the mortgage-interest and charitable contribution deductions would be gone, as would the deductibility of employer-provided group health plans."

    I would vote you in.

    These types of policies simply distort the incentive structure and lead to unintended consequences.

    1. "deductibility of employer-provided group health plans"

      Eliminate deductibility and (in my humble opinion) you have single payor within five years.

  2. I think your point that increasing the margional rate and phasing out deductions being the same is incorrect. Increasing the rate makes you less willing to work, or save or take chances in investments. Phasing out deductions discourages expendures the deduction for which is phased out(for example, a large house with a big mortgage, or the willingness to live in a high tax state).
    You comment re the WSJ table of tax rates, remember, not all states have state income taxes. I live in Texas (you may want to look into teaching at SMU, I will invite you over for dinner). Europe does not have state income taxes, but has high VAT taxes that are much broader than US sales taxes. In any event, I would take WSJ article with grain of salt, as cross country comps are not valid as base is different (journalist,and most economists, rarely get details right about taxes).
    The studant loan phase out, like plans to "means test" SS, raises the question as to how much we want to punish good behavior, like hard work and savings, and how much less good behavior you will get if you keep punishing it. And how will that impact the economy.
    As a tax lawyer I could give you my answer as to why we have complex tax laws, but that would make this post too long.

  3. The right generally will argue that a 35% marginal tax rate on a banker making $10 Million per year is a crippling disincentive to work but that an effective marginal rate on a
    low income single mother of 80% is just an economic incentive to marry wisely.

    I suggest that an over-arching principle in the design of all programs is that the total effective marginal rate be capped at 50%. That will be a difficult goal to reach.

  4. I don't know if you can say a priori what the highest marginal tax rate should be (especially without stating how many people it applies to). I'm more comfortable stating what things I am or am not comfortable with the government doing (with the latter taking up most of what it currently does). In a situation like Danny Boyle's "Sunshine" where all the fissile material of the earth has to be mined to destroy (or knock off course) an asteroid before it hits the earth, I'd tolerate Laffer-maximum taxation. Although the assumption that we should tax income rather than consumption or the unimproved value of land is annoying.

    David Beckworth has a response to your reading of Woodford. He also links to a good number of other economists/bloggers with opinions on your Woodford post. Looking forward to a rejoinder.

  5. The top marginal rate won't be affected by the phase-outs.

    And high marginal rates for low incomes does not shock any decent public finance economist. Neither their existence, nor that they can make sense. It is a primary feature of the optimal tax literature Diamond and Saez (2011 JEP) built on or Piketty and Saez surveyed for the forthcoming volume of Handbook of Public Finance. They discuss phase-out and transfer (monetary or in-kind) quite a lot. You would enjoy their treatment.

  6. Have you read:
    It analyzes some of the marginal tax rates (with phaseouts), and finds some rates greater than 100%.

  7. 10 kids? That seems a bit high.

  8. Grumpy: love your idea of radical tax simplification by eliminating all deductions and phaseouts. It is correct on every level. I have been advocating it for years. Every tax system I curse the IRS and go on a rant about it.

    Unfortunately, another unmentioned reason why politicians won't do it is because it makes them too accountable. With a simplified tax code, the only lever Congress critters have is the tax rate curve. Their constituents can then easily see if they voted to raise taxes. They want to avoid all such accountability. They would much rather have a complex tax system in which they have multiple levers they can pull, so that they can selectively tell their constituents the good bits, while hiding the fact that overall they raising their taxes.


  9. Your point about government phaseouts constituting effective taxes, and the perverse incentives it leads to, can be expanded in all kinds of interesting ways.

    Consider the possible ways to fund the costs of highways: gas taxes, tolls, car license/registration fees, or general taxes (e.g. sales taxes and income taxes).

    An ideal tax policy is one that is really a user fee, namely, you tax the exact activity that generates a cost, in the amount that pays the cost. Plus it should be cheap to collect, and is not easily cheated. Really bad taxes are the opposite. In particular, they are disconnected from the tax generating activity. One then has no incentive to reduce the cost generating activity, since that activity is being subsidized by all the people who have to do the broad based general activity that is being taxed.

    Of those road funding taxes mentioned before, only gas taxes are good; in fact, they are almost ideal. Gas taxes are extremely quick, cheap, and anonymous to collect. Plus vehicles that are bigger cause far more wear and tear on roads, and generate far more pollution. Since they generally get worse gas mileage, they correspondingly pay higher gas taxes per mile traveled, which compensates for their greater damage.

    Toll roads are horrendous: they waste tons of money on the facilities and people to run them, cause huge waste of time and gas due to artificial traffic jams they create, lend themselves to surveilance, and tax all vehicles the same, ignoring the fact that different vehicles cause different damage.

    Lastly, funding roads thru general sales and/or income taxes is obviously horrible, as it disconnects the tax from the cost generating activity.

    But the problem in the US is that our gas taxes are ridiculously small. They do not remotely cover the known direct costs of roads, such as road construction/maintenance. Toss in the indirect costs, like deaths from pollution, which are much harder to estimate (and may be somewhat subjective) but are thought to be even higher, and you see that total road costs in the US are actually being subsidized by tases on the general economy.

    In other words, the govt is effectively giving gas guzzler owners a welfare subsidy. They are leaching off the general population to fund their wasteful activities. Consequently, they have no incentive to economize.

    I a surprised that more of my fellow conservatives are not leaping to raise gas taxes (along with corresponding elimination of toll roads, and reductions in income taxes). They must not be thinking, and must be having a knee jerk reaction to gas tax proposals from the Democrats, who simply want to raise taxes on everything...


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.