Feldstein Proposal Increases Federal Revenues but the Devil’s in the Details

April 30, 2013

Download (PDF) Fiscal Fact No. 366: Feldstein Proposal Increases Federal Revenues

Professor Martin Feldstein of Harvard has called for limiting the tax savings from itemized deductions, tax-exempt municipal bond interest, and the tax-free status of employer-provided health insurance.[1] Feldstein's plan would not allow the tax savings from these items to exceed 2 percent of adjusted gross income (AGI).

For example, suppose that a couple has an AGI of $125,000. Also suppose the couple has itemized deductions, tax-exempt municipal bond interest, and employer-provided health insurance that together total $30,000. If the couple is in the 25 percent tax bracket, these deductions and tax exemptions lower their tax bill by $7,500. Professor Feldstein's plan would restrict the tax saving to $2,500 (2 percent of $125,000), which would increase the couple's tax bill by $5,000.

Professor Feldstein claims that his plan would be a potent revenue raiser, trimming the federal budget deficit by $220 billion in 2013 if none of the extra revenue is diverted into higher government spending—a big if. As one exception to the cap, Professor Feldstein suggests sparing charitable donations from the limitation. He claims that if the cap ignored charitable contributions and the first $8,000 of employer-provided health insurance, the current year's deficit reduction would be $141 billion and the ten-year deficit reduction would be $2.1 trillion.[2]

To better understand the implications of this proposal, we have used the Tax Foundation's tax simulation model to assess whether these numbers are plausible and to examine how the cap would affect economic growth. In each of the cases discussed below, we began by generating a conventional revenue estimate, which entails the (unrealistic) static assumption that macroeconomic aggregates (output, employment, investment, etc.) are unaffected by taxes. We then produced a dynamic estimate based on the recognition that people would respond to the higher marginal tax rates resulting from the cap by saving and working less.

The Tax Foundation model confirms that the proposed Feldstein limitation would be a major revenue raiser. In order to compare our results to Feldstein's $220 billion revenue estimate, the model was initially run without protecting the charitable deduction from the 2 percent cap. That generated a static revenue estimate of $220 billion, which is virtually identical to Feldstein's number, and a dynamic estimate of $188 billion. It is a bit surprising the agreement is so close because of dissimilarities in the models, such as different base years and different methods of attributing employer-provided health insurance to workers, but apparently the differences cancelled out.[3]

Because Professor Feldstein strongly urges that charitable contributions be shielded from the cap, we made that assumption in all later model runs. As shown in Table 1, if charitable contributions are not limited by the cap, the estimated static revenue gain is $216 billion.[4] The Tax Foundation's dynamic simulation finds, not surprisingly, that the higher marginal tax rates stemming from the cap would depress economic activity. Those results are also in Table 1. The model estimates that once the economy has adjusted to the new tax rules, the private business capital stock would be 1.8 percent smaller, the wage rate and private work hours would drop 0.4 percent and 0.5 percent respectively, and private business GDP would fall by 0.9 percent. The smaller economy would have a negative feedback on tax revenues, taking away an estimated 15 percent of the money the Feldstein limitation would otherwise collect. Still, the plan would be a significant revenue raiser. However, for every $1 of new revenue raised, the plan would lower GDP by 73 cents. In assessing the plan, policymakers should be aware of this cost in terms of lost economic output.

Table 1. Economic And Budget Effects Of The Feldstein Plan Versus 2013 Law (billions of 2012 dollars)

GDP

-0.86%

Private business GDP

-0.93%

Private business stocks

-1.77%

Wage rate

-0.39%

Private business hours of work

-0.55%

Federal revenue (dynamic)($ billions)

$184.9

Federal spending ($ billions)

-$3.6

Federal surplus (+ = lower deficit) ($ billions)

$188.5

Static revenue estimate ($ billions)

$216.1

% Revenue reflow vs. static

-14.5%

$GDP ($ billions)

-$135.6

$GDP/$tax increase (dollars)

-$0.73

Weighted Average service price

Change

% Change

Corporate

0.11%

0.79%

Non-corporate

0.12%

1.03%

All business

0.12%

0.86%

DISTRIBUTION EFFECTS

(billions of 2012 dollars except as noted)

All Returns

Average after-tax income per return

Static

Dynamic

AGI Class

Change

% Change

Change

% Change

< 0

-$1,937

2.00%

-$1,015

1.05%

0 – 5,463

-$2

-0.07%

-$26

-0.91%

5,463 – 10,925

-$3

-0.03%

-$71

-0.87%

10,925 – 21,850

-$29

-0.18%

-$161

-0.99%

21,850 – 32,775

-$130

-0.48%

-$340

-1.26%

32,775 – 43,700

-$319

-0.84%

-$605

-1.59%

43,700 – 54,625

-$605

-1.24%

-$952

-1.94%

54,625 – 81,938

-$1,103

-1.64%

-$1,574

-2.35%

81,938 – 109,250

-$1,838

-1.95%

-$2,476

-2.63%

109,250 – 163,875

-$3,782

-2.88%

-$4,622

-3.52%

163,875 – 218,500

-$6,137

-3.28%

-$7,311

-3.91%

218,500 – 273,125

-$7,808

-3.22%

-$9,210

-3.80%

273,125 – 546,250

-$11,474

-3.13%

-$13,578

-3.70%

546,250 – 1,092,500

-$27,798

-3.76%

-$31,910

-4.32%

> 1,092,500

-$111,189

-3.09%

-$132,505

-3.68%

TOTAL FOR ALL

-$1,384

-2.28%

-$1,783

-2.94%

The distributional analysis indicates that the plan would cause the most pain for upper-income taxpayers. They are more likely than lower-income taxpayers to itemize and have tax-exempt interest. However, taxpayers at all income levels would enjoy less after-tax income because the overall economy would be smaller due to the negative effects of higher marginal tax rates. The estimated drop in after-tax incomes would be slightly less than 1 percent at the bottom of the income scale and 3.5 percent or higher at incomes above $100,000.

Table 2 presents the results if the Feldstein cap were used for a different purpose: financing lower marginal tax rates while maintaining static revenue neutrality. According to the Tax Foundation model, the cap could finance a 17 percent, across-the-board cut in individual income tax rates. For example, the 10 percent bracket could be reduced to 8.3 percent, and the 39.6 percent bracket could be cut to 32.9 percent. The dynamic simulation model indicates the trade would help the economy, with the private capital stock rising 1.6 percent, the wage rate and private work hours increasing 0.16 and 1.05 percent respectively, and private business GDP growing 1.2 percent. Because of the positive economic feedback, the model estimates that federal revenues would actually increase by $38 billion. The dynamic analysis indicates that most lower-income and many middle-class taxpayers would see their after-tax incomes rise by approximately 1 percent due to lower tax rates and a stronger economy. Higher-income taxpayers would either gain slightly or break even, on average, because the lower rates and healthier economy would offset the bite of the cap.


Table 2. Economic And Budget Effects Of Using The Revenues From The Feldstein Plan To Lower Statutory Tax Rates

(billions of 2012 dollars)

GDP

1.09%

Private business GDP

1.21%

Private business stocks

1.61%

Wage rate

0.16%

Private business hours of work

1.05%

Federal revenue (dynamic)($ billions)

$37.9

Federal spending ($ billions)

$3.3

Federal surplus (+ = lower deficit) ($ billions)

$34.7

Static revenue estimate ($ billions)

$1.2

% Revenue reflow vs. static

3128.1%

$GDP ($ billions)

$171.6

$GDP/$tax increase (dollars)

$4.52

Weighted Average service price

Change

% Change

Corporate

0.11%

0.79%

Non-corporate

-0.38%

-3.18%

All business

-0.05%

-0.39%

DISTRIBUTION EFFECTS

(billions of 2012 dollars except as noted)

All Returns

Average after-tax income per return

Static

Dynamic

AGI Class

Change

% Change

Change

% Change

< 0

-$1,931

2.00%

-$3,141

3.25%

0 – 5,463

-$2

-0.06%

$27

0.97%

5,463 – 10,925

$4

0.05%

$91

1.11%

10,925 – 21,850

$27

0.16%

$194

1.20%

21,850 – 32,775

$60

0.22%

$329

1.22%

32,775 – 43,700

$99

0.26%

$467

1.23%

43,700 – 54,625

$74

0.15%

$525

1.07%

54,625 – 81,938

$67

0.10%

$681

1.01%

81,938 – 109,250

$18

0.02%

$851

0.90%

109,250 – 163,875

-$625

-0.48%

$488

0.37%

163,875 – 218,500

-$1,081

-0.58%

$471

0.25%

218,500 – 273,125

-$1,822

-0.75%

$12

0.01%

273,125 – 546,250

-$2,168

-0.59%

$634

0.17%

546,250 – 1,092,500

$2,915

0.39%

$8,622

1.17%

> 1,092,500

$30,846

0.86%

$60,108

1.67%

TOTAL FOR ALL

-$7

-0.01%

$518

0.85%

While revenue is the main draw of the plan, Martin Feldstein argues it is also consistent with good tax principles. In some respects that is true but in many others it is questionable or false.

Economists have recognized for years that the tax-free status of employer-provided health insurance, which was popularized in World War II as a way to circumvent wage controls, distorts the efficient allocation of resources by making employer-provided health insurance appear artificially cheap. A strong economic case can be made for treating it like other forms of labor compensation by listing it on employees’ W-2s and including it in the tax base. However, two criticisms of Professor Feldstein's approach to taxing employer provided health care are that it is complicated and incomplete.

Similarly, a solid economic case can be made that the municipal bond tax exemption is a poorly designed method of extending federal aid to states and localities.[5] (The subsidy occurs because lenders are willing to accept state and local debt issues at lower interest rates than they would otherwise due to the favorable tax treatment.) A cleaner approach would be to end the municipal bond exemption prospectively and, to the extent Congress deems it appropriate, provide aid to states and localities transparently and explicitly through the federal budget process. Here, too, Professor Feldstein’s approach suffers from being complicated and partial. In addition, because the cap would apparently apply to interest on existing municipal bonds as well as new ones, it would amount to a retroactive tax change, which raises many concerns.

Like the health insurance and municipal bond exemptions, itemized deductions are among the many tax provisions that the U.S. Treasury and the Joint Committee on Taxation classify as tax expenditures, by which they mean departures from a normal tax system that have the effect of subsidizing particular products or activities. Although the tax expenditure lists are often presented as cut-and-dried compilations of tax loopholes and subsidies, the reality is that the lists are highly subjective, and while some of their components are clearly subsidies, others are needed to measure income properly or reduce tax biases.[6] Rather than targeting itemized deductions because they are called tax expenditures, a more principled course of action would be to examine them individually and decide on a case-by-case basis which should be allowed and which should not.

Professor Feldstein has a soft spot in his heart for the charitable deduction, perhaps because he knows from his long association with Harvard and the National Bureau of Economic Research about the good work often done by nonprofit institutions. But the exception he would make for the charitable deduction illustrates why it would be bad tax policy to place a blanket cap on all other itemized deductions without carefully examining the merits of each.

A cap is especially inappropriate because itemized deductions are already subject to multiple limitations: several can only be claimed if they exceed a certain percentage of AGI; many are limited if the taxpayer's AGI exceeds certain thresholds; and several are disallowed when computing the alternative minimum tax (AMT). In addition, why does he propose to cap itemized deductions but not the standard deduction? Moreover, if Professor Feldstein seeks a generalized limitation on tax expenditures, why does he skip over the many tax credits on the tax expenditure list?

A means of cushioning the economic damage if the plan were adopted, while still substantially lowering the federal deficit under both static and dynamic scoring, would be to use half the revenue to reduce the deficit and the other half to finance an across-the-board statutory rate cut. Professor Feldstein does not endorse a fifty-fifty split, although he briefly observes, "With that much extra revenue from tax reform, it would be possible to cut some marginal tax rates."[7] The Tax Foundation model simulated this option, in which the 2 percent cap were imposed and all statutory individual rates were cut 8.5 percent. The model estimated that both the static and dynamic revenue gains would be approximately $100 billion, while GDP increased very slightly (0.2 percent) and the capital stock was essentially unchanged. This combination, however, would not lessen the plan's complexity, lack of transparency, and failure to base the proposed changes in the tax base on coherent tax principles.

If the goal is to collect more tax dollars by hook or by crook, Professor Feldstein's plan would cause less harm than many other suggestions. That is something in its favor but hardly a ringing endorsement.


[1] Martin Feldstein, It’s Time To Cap Tax Deductions, Washington Post, Mar. 12, 2013, http://www.washingtonpost.com/opinions/its-time-to-cap-tax-deductions/2013/03/12/af05081c-8a63-11e2-8d72-dc76641cb8d4_story.html; Martin Feldstein, A Simple Route To Major Deficit Reduction, Wall Street Journal, Feb. 20, 2013, http://online.wsj.com/article/SB10001424127887324880504578296920278921676.html; Martin Feldstein, Daniel Feenberg, & Maya MacGuineas, Capping Individual Tax Expenditure Benefits, National Bureau of Economic Research Working Paper 16921 (Apr. 2011), http://www.nber.org/papers/w16921.

[2] See Feldstein, A Simple Route, supra note 1.

[3] The Tax Foundation's income tax calculator used the IRS's 2005 Public Use File, aged to 2008. The dollar amounts were then inflation-adjusted to 2012. We assumed a simple relationship between wages and the value of employer provided health insurance whereas Professor Feldstein and his colleagues assembled highly detailed information on how employer provided health insurance is likely to vary among workers.

[4] Consistent with Professor Feldstein's explanation of his proposal, the charitable deduction is shielded from the limitation, the cap does not apply to taxpayers who claim the standard deduction, and itemizers are allowed to switch if that would prevent their tax bills from rising as much.

[5] For a discussion of some of the problems with subsidizing state and local governments through the municipal bond exemption, see Congressional Budget Office, Frank Sammartino, Federal Support for State and Local Governments Through the Tax Code, Testimony before the Senate Committee on Finance, Apr. 25, 2012, http://www.cbo.gov/sites/default/files/cbofiles/attachments/04-25-TaxCodeTestimony.pdf.

[6] For an insightful discussion of some of the peculiarities and debatable assumptions that underlay the tax expenditure list, see U.S. Office of Management and Budget, Analytical Perspectives, Budget of the United States Government, Fiscal Year 2009, ch. 19, http://www.whitehouse.gov/omb/budget/fy2009/pdf/spec.pdf. Especially enlightening is the explanation of how different the list would be if measured against a comprehensive consumption tax rather than the current, often inconsistent baseline.

[7] See Feldstein, It’s Time to Cap, supra note 1.


Tags


About the Author


Related Research