Set Phase-Outs to Stun
-- Posted by Neil H. Buchanan
Continuing my discussion of the Ways and Means Committee's hearing last Wednesday on tax simplification, previously blogged here and here, I now take up the question of "tax phase-outs." After describing the purpose and mechanics of phase-outs, I will discuss critically the Republican majority's desire to eliminate phase-outs from the tax code.
A phase-out is a mechanism by which a tax provision's effect is reduced, and ultimately eliminated, in a gradual manner, rather than by setting a strict cut-off rule. In our federal income tax system, provisions are usually phased out over a range of income. For example, you could have a tax benefit that is fully available for anyone with income up to $25,000, not available at all to anyone with income above $50,000, and available in decreasing degree as income rises from $25,000 to $50,000. Usually, this is done as a simple linear progression, so that taxpayers with $40,000 in income would be eligible for 2/5 of the benefit in question, because their incomes are 3/5 of the way through the phase-out range. A $1000 tax credit, for example, thus would be reduced to $400.
As I pointed out in my testimony, the mechanics of a phase-out are no different from the mechanics of tax computations themselves. Although you will hear politicians talk a lot about how a single rate of tax would simplify matters ("Just multiply your taxable income by 17%!"), the same fear of math that causes people not to be able to calculate sales taxes or tips is at play in the federal income tax as well. A study a few years ago showed that people would want to be given tax tables, even if there were only one tax rate. Find your income on the table, and write down the number associated with it for tax liability. That is a simple process that works for any system of tax rates.
The same mechanism works for phase-outs. If you are willing and able to do simple arithmetic, the process is easy. If you are not, the IRS can do it for you. Again, the broader point is that the arithmetic is NOT what makes the tax code complicated. Determining whether a taxpayer is eligible for certain provisions at all (for example, various child credits) is what makes life complicated and worrying for taxpayers.
I confess to being surprised by the Republicans' interest in eliminating phase-outs, because it simply never occurred to me that this was a big issue for anyone. The conservative economist who testified at the hearing, however, focused in his prepared testimony on the supposed problems caused by the existence of a collection of different phase-out rules for different provisions in the tax code. I conceded that he had a very minor point, which is that an accumulation of different phase-out rules for different tax provisions can become somewhat time-consuming. Even that was not enough for one member of the committee, who wasted much of his time trying to get me to explain why phase-outs were not inherently complicated. When I finally said, "Twenty different phase-outs are complicated, while one phase-out is not," he backed down (not at all happily, by all appearances).
Even under the current set of rules, a person can look at phase-outs to determine whether it is even worth it to engage with the specifics of a tax provision. After I graduated from law school (as a middle-aged person), for example, I noted that there is a tax credit for higher education expenses. I then saw that the provision phased-out entirely at an income level much lower than mine, making it unnecessary to proceed further. Drawing on that experience, I offered to the Ways and Means Committee a simple solution to dealing with multiple phase-out rules: replace all existing (and future) phase-outs with a single phase-out. (I have not seen this proposal elsewhere, but I would not be surprised to learn that I had reinvented the wheel). This would allow higher-income people to know that all phased-out provisions are simply unavailable to them, and it would tell moderate-income people the fraction of any benefits for which they are eligible. Lower-income people would know that their incomes are low enough to qualify for full benefits.
The other witnesses at the hearing were unable to make a case for repeal of phase-outs. The CPA pointed out that the complications of tax provisions affect planning as well as filling out tax forms. This is true, but it has nothing to do with the issues here. If I know that, say, I am eligible for 32% of any tax benefit for which I am otherwise eligible, then I can plan appropriately. Will I do what is necessary to qualify for a $1000 credit, if I know that it will be worth $320 to me? I can make that decision, based on information that is knowable in advance -- or, to be more precise, information that is just as knowable as all other information for planning purposes.
The argument that phase-outs complicate tax planning, therefore, is simply false. The conservative economist on the panel, however, offered a different argument. He suggested that Congress needed a good reason to exclude certain people from being eligible for the "incentives" that tax provisions provide. If, for example, we want to give people a tax benefit for child care, then we apparently need to explain why that credit is not available to people earning relatively high incomes.
Although this argument is clearly better than the argument that phase-outs complicate tax planning (in that it is not simply a false assertion), it is quite easy to answer, both in terms of efficiency and equity. In terms of efficiency, we have good reason to be confident that some decisions are "marginal" for lower-income taxpayers, but "infra-marginal" for higher-income taxpayers. Higher-income people are likely to spend for higher education, for example, whether there is a tax benefit or not. Spending for child care, similarly, is highly unlikely to depend on tax benefits for higher-income people, but is very likely to be responsive to tax benefits for lower- and moderate-income taxpayers.
As a matter of equity, the case is even easier to make. The society at large has to make decisions about whether to subsidize certain activities; and it makes precious little sense to give wealthier people money to do things that they can afford to do on their own. Even to ask the question: "Why shouldn't this tax subsidy be extended to the wealthiest taxpayers?" is to expose the question as disingenuous.
The effect of eliminating phase-outs, after all, would be to make tax benefits more expensive. Without phase-outs, we would be left with three choices: (1) Pay the higher cost of making a benefit available to upper-income taxpayers, (2) Cut the benefit off without a phase-out, or (3) Eliminate the tax benefit. Choosing (1) would ultimately result in the need to impose higher taxes to pay for the benefits -- higher taxes that, the committee's Republicans would be quick to say, must not be imposed on the beleaguered upper class. This would ultimately mean that extending the tax benefit to the wealthy would be paid for by increasing taxes on the non-wealthy. Choosing (2) would create serious efficiency and equity problems, as the "cliff effect" problem would kick in. That is, taxpayers could lose an entire benefit by earning $1 more in income, if they are at the cutoff point.
That leaves (3), eliminating tax benefits for non-wealthy people. This is being sold as a way to simplify the tax code, but its effect (and, I suspect, the goal of the people pushing this idea) is to eliminate a way to make the tax code responsive to relative economic need. Indeed, the conservative economist went out of his way in his testimony to point out that phase-outs are an indirect way of raising effective marginal tax rates on the wealthy. Worse still, he said, this is not transparent, which is worse than simply raising statutory rates on higher incomes.
It is difficult to take such claims seriously. Saying that Congress could raise rates on upper-income people openly, if it wanted, is simply a way of saying: "We've won the framing war, making tax increases politically unacceptable. You now have to agree to have all progressive tax provisions framed as we want them to be framed, so that they will look to everybody like tax increases."
The sudden focus on phase-outs, therefore, is nothing more than an opportunistic effort to piggyback yet another regressive tax change onto the admirable goal of tax simplification. My proposal demonstrates that it is possible to maintain the progressivity of phase-outs without creating complexity in tax planning or filing. Anything beyond that is nothing more than a cynical attempt to reduce taxes on the wealthy.
Continuing my discussion of the Ways and Means Committee's hearing last Wednesday on tax simplification, previously blogged here and here, I now take up the question of "tax phase-outs." After describing the purpose and mechanics of phase-outs, I will discuss critically the Republican majority's desire to eliminate phase-outs from the tax code.
A phase-out is a mechanism by which a tax provision's effect is reduced, and ultimately eliminated, in a gradual manner, rather than by setting a strict cut-off rule. In our federal income tax system, provisions are usually phased out over a range of income. For example, you could have a tax benefit that is fully available for anyone with income up to $25,000, not available at all to anyone with income above $50,000, and available in decreasing degree as income rises from $25,000 to $50,000. Usually, this is done as a simple linear progression, so that taxpayers with $40,000 in income would be eligible for 2/5 of the benefit in question, because their incomes are 3/5 of the way through the phase-out range. A $1000 tax credit, for example, thus would be reduced to $400.
As I pointed out in my testimony, the mechanics of a phase-out are no different from the mechanics of tax computations themselves. Although you will hear politicians talk a lot about how a single rate of tax would simplify matters ("Just multiply your taxable income by 17%!"), the same fear of math that causes people not to be able to calculate sales taxes or tips is at play in the federal income tax as well. A study a few years ago showed that people would want to be given tax tables, even if there were only one tax rate. Find your income on the table, and write down the number associated with it for tax liability. That is a simple process that works for any system of tax rates.
The same mechanism works for phase-outs. If you are willing and able to do simple arithmetic, the process is easy. If you are not, the IRS can do it for you. Again, the broader point is that the arithmetic is NOT what makes the tax code complicated. Determining whether a taxpayer is eligible for certain provisions at all (for example, various child credits) is what makes life complicated and worrying for taxpayers.
I confess to being surprised by the Republicans' interest in eliminating phase-outs, because it simply never occurred to me that this was a big issue for anyone. The conservative economist who testified at the hearing, however, focused in his prepared testimony on the supposed problems caused by the existence of a collection of different phase-out rules for different provisions in the tax code. I conceded that he had a very minor point, which is that an accumulation of different phase-out rules for different tax provisions can become somewhat time-consuming. Even that was not enough for one member of the committee, who wasted much of his time trying to get me to explain why phase-outs were not inherently complicated. When I finally said, "Twenty different phase-outs are complicated, while one phase-out is not," he backed down (not at all happily, by all appearances).
Even under the current set of rules, a person can look at phase-outs to determine whether it is even worth it to engage with the specifics of a tax provision. After I graduated from law school (as a middle-aged person), for example, I noted that there is a tax credit for higher education expenses. I then saw that the provision phased-out entirely at an income level much lower than mine, making it unnecessary to proceed further. Drawing on that experience, I offered to the Ways and Means Committee a simple solution to dealing with multiple phase-out rules: replace all existing (and future) phase-outs with a single phase-out. (I have not seen this proposal elsewhere, but I would not be surprised to learn that I had reinvented the wheel). This would allow higher-income people to know that all phased-out provisions are simply unavailable to them, and it would tell moderate-income people the fraction of any benefits for which they are eligible. Lower-income people would know that their incomes are low enough to qualify for full benefits.
The other witnesses at the hearing were unable to make a case for repeal of phase-outs. The CPA pointed out that the complications of tax provisions affect planning as well as filling out tax forms. This is true, but it has nothing to do with the issues here. If I know that, say, I am eligible for 32% of any tax benefit for which I am otherwise eligible, then I can plan appropriately. Will I do what is necessary to qualify for a $1000 credit, if I know that it will be worth $320 to me? I can make that decision, based on information that is knowable in advance -- or, to be more precise, information that is just as knowable as all other information for planning purposes.
The argument that phase-outs complicate tax planning, therefore, is simply false. The conservative economist on the panel, however, offered a different argument. He suggested that Congress needed a good reason to exclude certain people from being eligible for the "incentives" that tax provisions provide. If, for example, we want to give people a tax benefit for child care, then we apparently need to explain why that credit is not available to people earning relatively high incomes.
Although this argument is clearly better than the argument that phase-outs complicate tax planning (in that it is not simply a false assertion), it is quite easy to answer, both in terms of efficiency and equity. In terms of efficiency, we have good reason to be confident that some decisions are "marginal" for lower-income taxpayers, but "infra-marginal" for higher-income taxpayers. Higher-income people are likely to spend for higher education, for example, whether there is a tax benefit or not. Spending for child care, similarly, is highly unlikely to depend on tax benefits for higher-income people, but is very likely to be responsive to tax benefits for lower- and moderate-income taxpayers.
As a matter of equity, the case is even easier to make. The society at large has to make decisions about whether to subsidize certain activities; and it makes precious little sense to give wealthier people money to do things that they can afford to do on their own. Even to ask the question: "Why shouldn't this tax subsidy be extended to the wealthiest taxpayers?" is to expose the question as disingenuous.
The effect of eliminating phase-outs, after all, would be to make tax benefits more expensive. Without phase-outs, we would be left with three choices: (1) Pay the higher cost of making a benefit available to upper-income taxpayers, (2) Cut the benefit off without a phase-out, or (3) Eliminate the tax benefit. Choosing (1) would ultimately result in the need to impose higher taxes to pay for the benefits -- higher taxes that, the committee's Republicans would be quick to say, must not be imposed on the beleaguered upper class. This would ultimately mean that extending the tax benefit to the wealthy would be paid for by increasing taxes on the non-wealthy. Choosing (2) would create serious efficiency and equity problems, as the "cliff effect" problem would kick in. That is, taxpayers could lose an entire benefit by earning $1 more in income, if they are at the cutoff point.
That leaves (3), eliminating tax benefits for non-wealthy people. This is being sold as a way to simplify the tax code, but its effect (and, I suspect, the goal of the people pushing this idea) is to eliminate a way to make the tax code responsive to relative economic need. Indeed, the conservative economist went out of his way in his testimony to point out that phase-outs are an indirect way of raising effective marginal tax rates on the wealthy. Worse still, he said, this is not transparent, which is worse than simply raising statutory rates on higher incomes.
It is difficult to take such claims seriously. Saying that Congress could raise rates on upper-income people openly, if it wanted, is simply a way of saying: "We've won the framing war, making tax increases politically unacceptable. You now have to agree to have all progressive tax provisions framed as we want them to be framed, so that they will look to everybody like tax increases."
The sudden focus on phase-outs, therefore, is nothing more than an opportunistic effort to piggyback yet another regressive tax change onto the admirable goal of tax simplification. My proposal demonstrates that it is possible to maintain the progressivity of phase-outs without creating complexity in tax planning or filing. Anything beyond that is nothing more than a cynical attempt to reduce taxes on the wealthy.