During the upcoming fiscal cliff negotiations, most of the attention is likely to be paid to two threshold questions: (1) whether or not Congress should agree to new tax revenue, period; and (2) what magnitude of spending cuts should be required for Congress to accept new revenue. Nearly as important, however, is how the revenue is to be collected. In this case, it is quite obvious that revenue should come from limitation on deductions rather than increases to marginal tax rates. However, there are key differences between the various approaches to limiting deductions. If revenue is necessary to avoid the fiscal cliff or achieve sizeable debt reduction, Congress should ensure that deductions are limited in the way that creates the fewest distortions.
It is generally agreed that existing deductions would be only limited for “the rich,” whatever income threshold is used to define that category of households. Yet, if current deductions are eliminated in a manner similar to existing income phase-outs for tax preferences, the results would be large increases in marginal tax rates. The economic literature is quite clear that increases in marginal tax rates cause households to reduce their supply of investment and labor and result in a smaller economy, with most of the adjustment borne by second earners who scale back work or drop out of the labor force entirely. These economic effects do not recede simply because the marginal rate increase arises due to the phase-out of subsidies rather than an increase in statutory rates. Indeed, as Greg Mankiw writes, the Affordable Care Act amounts to one of the largest marginal tax increases in history precisely because each additional dollar earned between 100% and 400% of the poverty line leads to a proportional reduction in insurance coverage subsidies.
In the tax code, phase-outs can generate implied marginal tax rates higher than those faced by households with over $1 million of income. For example, households with incomes below $30,000 face marginal tax rates above the top statutory rate of 35% because each additional dollar they earn reduces the earned income tax credit (EITC). For higher income households, the interaction of various phase-outs can also generate higher marginal tax rates than those faced by ultra-high income earners.
For example, the child tax credit begins to phase-out at $75,000 of income. (For married taxpayers filing a joint return, the phase-out begins at $110,000.) In the phase-out range, the child tax credit is reduced by $50 for each $1,000 of income above these thresholds, which are not indexed for inflation. Hence, a married household earning $130,000 that receives an additional 3% cost-of-living adjustment ($3,900) would see its child tax credit drop by $195 even as its purchasing power remains stagnant. Assuming this household is in the 25% marginal tax bracket, the $3,900 of income would generate $937.50 of additional income tax and $187.50 of reduced tax credits, resulting in an effective marginal tax rate of 30% ($1,170/$3,900). If this household also has qualified tuition expenses of $4,000 or more, it would also be impacted by the phase-out for this deduction, which would add an additional $500 of new taxes for an effective marginal tax rate of 42.8%.
The risk is that Congress would simply take this model and graft it on to the new budget negotiations. For example, instead of allowing current statutory tax rates to expire, it could instead establish a phase-out range for the mortgage interest tax deduction that would start at $250,000 of adjusted gross income (married filing jointly). In this case, the effective marginal tax would have two components: (1) the statutory marginal tax rate; and (2) the additional, implicit tax that comes from exposing more income to taxation as a result of the phase-out. This is hardly more efficient than simply increasing the tax rate itself.
Instead, some may be tempted to embrace the Obama Administration proposal to reduce the tax rate at which deductions are credited against taxes. If this sounds complicated, it’s because it is; and that is the problem. Under current law, an eligible deduction reduces taxable income on a dollar-for-dollar basis. The actual cash value of the deduction is therefore a function of the tax filer’s marginal tax rate. If the taxpayer is in the 35% bracket, a $100 reduction in taxable income is worth $35; if the taxpayer is in the 15% bracket, the same reduction has a cash value of only $15. The itemized deduction is treated like an eligible “business expense” that reduces the actual take-home income of the household and should therefore reduce the amount of income subject to taxation.
In recent budgets, the Obama Administration has proposed limiting the tax value of deductions to 28%. That is, a $100 deduction for a taxpayer in the 35% bracket would reduce income taxes owed by only $28 instead of $35. The problem here is not only added complexity, but also the monomaniacal focus on distributive justice. That the tax benefits of deductions increase with rates seems to be inherently unfair to many analysts who promote tax credits instead of deductions. Credits are “below the line” in tax parlance because they represent actual cash value: a $1,000 tax credit reduces the amount of taxes owed by $1,000 quite irrespective of tax bracket. Credits eliminate the advantages that deductions provide to higher income earners. But what about those households who don’t have any tax liability? Well, promoters of tax credits argue that they should simply be made refundable. This transforms the tax code from a system of collecting revenues to operate the government to a quasi-welfare state inside of a welfare state. The result is a significant percentage of taxfilers who have zero or even negative income tax liability – i.e., their interactions with the IRS result in cash payments.
Table 1: Average Deductions
|Income Level|| |
Avg Deductions ($K)
Total Deductions ($K)
$75K Limitation ($K)
Increase in Taxable Income ($K)
|$200,000 -$500,000|| |
|$500,000 -$1,000,000|| |
|$1,000,000 -$1,500,000|| |
|$1,500,000 - $2,000,000|| |
|$2,000,000 - $5,000,000|| |
|$5,000,000- $10,000,000|| |
|$10,000,000 or more|| |
Instead of phase-outs for existing deductions above certain income thresholds that raise marginal rates, or reductions in the tax value of deductions, Congress should consider caps on the deductions themselves or an across-the-board limitation on deductions. According to the most recent IRS data, households in the $200.000 to $500,000 income range take $51,000 in itemized deductions per year, on average; for households in the $500,000 to $1 million income range, deductions are $101,500 on average (see Table 1). The establishment of a limitation on deductions in the range of $75,000 would likely be sufficient to raise $35 billion in 2013 – close to the $42 billion raised by the expiration of the top two rates – with virtually no effect on households with incomes of under $250,000 and no impact on marginal tax rates. As shown in Table 1, the bulk of the adjustment would be borne by taxpayers with incomes of over $5 million who have average deductions in excess of $900,000 per year. The limitations may have an impact on tax-related decisions to incur added mortgage interest or live in a high-tax state, but they would not impact decisions to reinvest in a business or change calculations for a household’s second earner. Again, best of all households with incomes of under $500,000 would be untouched, on average.
Obviously, questions about whether to raise revenue and the spending reductions required to get Congress to agree to new revenues are of paramount importance. But Congress should also focus on how to raise that revenue most efficiently. The answer is not simply by limiting deductions; it is limiting those deductions in a way that does not bias economic decisions or slow growth.