Sooner or later, Congress will realize it needs new revenues to help balance the budget, and trimming tax subsidies is the way to get them.  But will it tackle individual preferences, such as the mortgage interest deduction, one at a time? Or, will it try to limit the political bloodshed and go after these tax breaks across-the board? The Tax Policy Center has looked at three ways to limit the benefit of tax subsidies for high-income households. Each would raise billions in new revenues—money that could be used to both reduce the deficit and cut tax rates. And each would be quite progressive. However, they’d work very differently. The first is the Obama Administration’s proposal to cap the value of itemized deductions at 28 percent. The second, called the Effective Minimum Tax, would require high income households to pay a rate of at least 21 percent on their Adjusted Gross Income (AGI) plus municipal bond income. The third is a modified version of a plan offered by Marty Feldstein, Dan Feenberg, and Maya MacGuineas that would limit the value of deductions and credits to 2 percent of AGI. Let’s take a look at each of them. The Obama plan would raise about $165 billion over 10 years (I’m comparing each plan to current policy--that is, the taxes people pay today). According to this proposal, those in the 33 percent and 35 percent brackets could take all the same itemized deductions they do now, but their value would be capped at just 28 percent.   Obama would raise taxes on about 5.4 million housholds. About 95 percent of the new taxes would be paid by the highest-income 5 percent of households, although a handful of middle-class people would  also get clipped. The average tax hike in 2013 would be about $2,800, and the top 0.1 percent would pay an average of about $54,000 more than they do today. The Effective Minimum Tax (EMT) was developed by TPC’s Jim Nunns. It would operate something like the alternative minimum tax but would be imposed on a broader tax base and affect only top-bracket taxpayers. They'd figure their regular tax, calculate their EMT rate, and pay the higher of the two. This levy would raise about $170 billion over 10 years.  Only about 400,000 households--all in the top 5 percent--would pay more tax. And they'd owe an average of $80,000 more. Those in the top 0.1 percent of income would pay much more than under the Obama plan—an average tax increase of more than $400,000. This tax, like the others, assumes the existing Alternative Minimum Tax remains on the books. But in the real world, if the EMT is enacted, I suspect the AMT would be repealed. If so, the plan would either have to impose a higher EMT rate or hit more middle-income taxpayers to generate the same amount of money.    Limit Tax Breaks to 2 Percent of AGI. The third broad cap on tax breaks is based on a plan designed by Harvard economist Marty Feldstein, Dan Feenberg at the National Bureau of Economic Research, and Maya MacGuineas of the New America Foundation. While they would have capped the value of most tax preferences at 2 percent of AGI for all taxpayers, TPC targeted the cap only at high-income households so we could better compare it to the other two plans. Importantly, while Obama would target only deductions, this proposal taxes a much broader base. The 2 percent limit would also apply to the value of employer-sponsored health insurance, the child and dependent care tax credit, and general business tax credits. Even scaled back to include only high-income households, this plan would generate twice as much money as the other two—more than $500 billion over 10 years. About 2.8 million housholds would pay more, and the average tax increase would be about $16,000. The top 0.1 percent would pay about $240,000 more. These are just three examples of how Congress could tackle the $1 trillion in tax expenditures without battling over each one individually.  In an ideal world, Congress would address each on their economic merits. But, this is Washington, and I wouldn’t be surprised to eventually see an across-the-board plan surface. And it is likely to look like one of these.