Yesterday Mitt Romney finally hinted at how he'd pay for his 20 percent, across-the-board income tax cut. He wants to cap "tax expenditures" — the deductions and exemptions in the tax code that promote employer-based health insurance, home-ownership, charitable contributions and so forth. And he wants to cap them at $17,000.

Mitt Romney speaks at a campaign rally in Fairfax, Virginia, September 13, 2012. (REUTERS/Jim Young)

What exactly he means by this is a bit unclear. Does the $17,000 cap mean that deductions can't reduce a person's taxes by more than $17,000? That would raise a good chunk of revenue, mostly from high earners. Or is Romney proposing that taxpayers can't reduce the income subject to tax by more than $17,000? That would raise even more money, but it would hit a lot of middle-class households. The typical household gets about $15,000 chopped off its taxable income from the employer health-care exemption alone, so many families would hit the $17,000 limit very quickly.

But as Ezra argued yesterday, capping deductions is a potentially promising avenue for tax reform. It promises to raise a whole lot of revenue without attacking specific industries like mortgage lenders and health insurers that have a vested interest in the current system. And like most good ideas, Romney's proposal isn't exactly new. Here's a list of five plans for capping tax deductions that are already out there — and what they mean for taxpayers.

1) The 2 percent solution

Perhaps the most-discussed capping plan was proposed by Harvard's Martin Feldstein, NBER's Daniel Feenberg and Maya MacGuineas, president of the bipartisan Committee for a Responsible Federal Budget. They propose capping the amount that the itemized deduction, the employer health exemption and a few tax credits can help a given taxpayer at no more than 2 percent of income.

To illustrate, suppose I'm making $45,000, have $15,000 in deductions and exemptions and am in the bottom 10 percent income tax bracket. My deductions and exemptions give me a $15,000 * 10% = $1,500 benefit. That's 3.3 percent of income, above the cap. So under the Feldstein/Feenberg/MacGuineas plan, my benefit would be reduced to $45,000 * 2% = $900. That's a total tax increase of $600.

If that sounds like a big middle-class tax hike, it is. Basically everyone would see a tax increase of around 2.5 to 3.5 percent of income under the plan:

On the other hand, the plan raises a ton of revenue -- $278 billion in 2011 alone. Most of that comes from limiting the employer health-care exemption. That alone raises about $140 billion. There would be an additional $46 billion from limiting the mortgage interest deduction. What's more, the limit on the health exemption hits the middle-class hardest. Poor families are hit hard  by limits to the child tax credit, while the rich lose out on the state tax and charitable deductions:

One could limit the plan's impact on the middle class by phasing it in only for those making more than $200,000 a year. The Tax Policy Center estimated that this change would raise $47.6 billion in 2015, or $519.7 billion over 10 years. That's about a sixth of what the full Feldstein/Feenberg/MacGuineas plan would raise. But it would hit high earners considerably harder, with the top 0.1 percent (those making more than $2.6 million) seeing a tax increase of about $240,000.

2) Obama's plan

President Obama's latest budget proposes a different approach to limiting tax expenditures. Right now, the benefit that a taxpayer gets from a deduction or exemption depends on what tax bracket he or she is in. For a millionaire in the 35 percent tax bracket, a $10,000 deduction is worth $3,500. For a person in the 10 percent bracket, that same deduction is worth only $1,000. 

Obama's budget would change that by treating deductions for high earners (those in the 33 percent, 36 percent and 39.6 percent brackets under the Obama plan) as if they were earned at the 28 percent bracket. So a millionaire, who normally would see a $3,960 benefit from a $10,000 deduction under Obama's rates, would see only a $2,800 benefit under this plan.

The plan meets Obama's goal of not raising taxes on people making under $200,000, but it consequently raises much less revenue than the 2 percent plan described above. The Tax Policy Center estimates that Obama's proposal raises just $164.2 billion over 10 years, compared to $519.7 billion under the 2 percent cap for high earners (and about $2.8 trillion under the full Feinstein/Feenberg/MacGuineas plan). It also hits high earners much less, with the top 0.1 percent getting a $53,632 tax hike.

3) Minimum taxes

This plan comes courtesy of the University of Chicago's Richard Thaler. He proposes taxing anyone making more than $1 million per year at 28 percent, period. There would be no deductions, credits, exemptions or capital gains preference -- just a flat 28 percent rate. Given that millionaires currently pay an average of 20.9 percent in income taxes, that's a big revenue raiser.

The legislative proposal to which this bears the strongest resemblance is the Buffett rule, which proposes a 30 percent minimum rate for millionaires. The full 30 percent rate comes into effect only for people making $2 million or more, with a phase-in between $1 million and $2 million. The rate also exempts employer-provided health insurance and charitable contributions. That proposal brings in about $263.7 billion over 10 years and raises taxes on the top 0.1 percent by an average of $555,856.

Both proposals are quite similar to the Alternative Minimum Tax. The AMT was implemented in 1969 to solve the same problem that the Buffett rule seeks to solve: millionaires who use tax breaks to reduce their burden to a very low level. Currently, it works as a 28 percent flat tax (26 percent for singles) for high earners, with no personal exemption or standard deduction and sharply limited other deductions (though the health-care exemption persists).

But because incomes keep rising and the AMT rates aren't tied to wage growth or inflation, more and more people are affected. That's why Obama's budget proposes replacing it with the Buffett rule and Mitt Romney wants to eliminate it entirely,.

An alternative would be building a better AMT. The Tax Policy Center considered what would happen if you implemented an "effective minimum tax" of 27 percent for couples making more than $250,000 on top of the current system (the rate would have to be higher, or the eligibility cutoff lower, to eliminate the AMT as well). The EMT would tax currently exempt bonds but leave employer-provided health care untaxed.

The TPC found that the EMT would raise $169 billion over 10 years -- less than the 2 percent cap or the Buffett rule but more than the 28 percent limit in Obama's budget. But it hits the top 0.1 percent with a $401,513 proposal -- higher than any plan besides the Buffett rule. In a way, an EMT would be like the Buffett rule with a bigger base (because charitable contributions and certain bonds are taxed) and lower rate. Both are more progressive than the 2 percent deduction cap idea, as Thaler's 28 percent tax idea would, in all likelihood, be.

4) Easy Pease-y

The current tax code includes another way to limit tax deductions for high earners: the Pease limit. Named after its author, former Democratic congressman Donald J. (Don) Pease (Ohio), the limit establishes a cutoff ($177,550 in 2013) and then reduces deductions by 3 percent of the amount by which a household's income exceeds that cutoff, up to a maximum reduction of 80 percent.

For example, suppose I make $300,000 and have $50,000 in deductions. The Pease limit would reduce my deductions by ($300,000-$177,550) * 3% = $3,673.50. That's less than 80 percent of $50,000, so my deductions are reduced by that amount. But if I make $10 million and have $50,000 in deductions, it would reduce my deductions by ($1 million-$177,550) * 3% = $294,673.5. That's way more than 80 percent of my deductions, so they're reduced by only 80% * $50,000 = $40,000.

The Pease limit was eliminated in the Bush tax cuts of 2001, but it's set to come back if those cuts expire next year. Obama's budget also assumes that the Pease limit is reinstated. The Treasury Department estimates (Excel doc) that that would raise $129 billion over 10 years relative to current policy.

Most of the above proposals would replace or work on top of Pease, but you could imagine a reform that just extends it, by eliminating the 80 percent limit, or increasing the reduction from 3 percent to 5 or 10 percent, or by reducing the cutoff.

5) Burn them all down!

Perhaps the simplest of all reforms would be to replace all itemized deduction with an increased standard deduction. This would have the advantage not just of eliminating special preferences and simplifying the code but of providing a tax break for low- and middle-income earners. The most notable plan with this element is the Armey-Shelby flat tax that was in vogue among conservatives in the 1990s. It would institute a 17 percent flat rate on all income above a standard deduction ($35,400 for a family of four; presumably it'd be larger in 2012 dollars).

Other plans increase the standard deduction without eliminating others, which effectively simplifies the code and eliminates preferences for low- and middle-income earners. The Wyden-Coats tax reform plan does this, as would former OMB director Peter Orszag's plan for the fiscal cliff. This wouldn't raise revenue, but it would eliminate the distorting influence of tax expenditures for most taxpayers.