Pasar al contenido principal

The Tension Between Reducing Tax Rates and Reducing Deficits

Over the past few months, a number of analyses have highlighted the difficulty of cutting income tax rates deeply, producing a significant revenue contribution to deficit reduction (as part of a larger deficit-reduction package), and maintaining the progressivity of the tax code.[1]   Most recently, the Joint Committee on Taxation (JCT) issued an analysis on tax expenditures and tax rates that attracted considerable attention.[2]   Some have interpreted JCT’s findings as confirming the view that sweeping changes in tax expenditures are likely to lead to only modest reductions in income tax rates, particularly if deficit reduction is to be achieved.  Others have argued that JCT’s findings do not preclude tax reform that sharply lowers tax rates while increasing revenues. 

As policymakers assess the import of these analyses, we encourage them to be sure to include one important ingredient:  a healthy dose of political reality.  A finding that it is technically possible to achieve sufficient tax-expenditure savings to pay for sizeable reductions in tax rates is not the same thing as such a course being politically viable.  Policymakers should avoid committing to a specific, lower top income tax rate until they know what measures to shrink tax expenditures Congress can actually pass and how much savings those measures will produce.[3]   Otherwise, the most critical goal of tax reform at this time — producing a significant contribution to deficit reduction (while maintaining or improving the progressivity of the tax code) — will likely be lost.

The Committee for a Responsible Federal Budget (CRFB) examined the JCT analysis and has argued that it substantially understates how much tax rates can be reduced. [4]   CRFB notes that the plans from former Clinton White House Chief of Staff Erskine Bowles and former Senator Alan Simpson, co-chairs of the National Commission on Fiscal Responsibility and Reform, and former Clinton Office of Management and Budget Director Alice Rivlin and former Republican Senate Budget Committee Chair Pete Domenici, co-chairs of a Bipartisan Policy Center debt reduction panel, reduced the top income tax rate more sharply — to under 30 percent — while increasing revenues above current-policy levels for deficit reduction. 

In its analysis, JCT assumed quite dramatic changes in tax expenditures:  eliminating all itemized deductions, including those for mortgage interest, charitable contributions, and state and local taxes; taxing capital gains and dividends at ordinary income tax rates; and repealing tax-free interest on new state and local bonds.  CRFB maintains that the JCT analysis could have gone considerably further because it did not also include changes in various other tax expenditures — particularly what it calls the largest tax expenditure in the code, the tax exclusion for employer-sponsored health insurance; CRFB says that eliminating or substantially reducing these additional tax expenditures could pay for much bigger reductions in tax rates.

CRFB also notes that JCT measured changes in tax rates from current law (the tax rates scheduled to go into effect on January 1, 2013) rather than from the lower current-policy tax rates (the rates now in effect), and that the rates achieved in JCT’s analysis would have been lower if current-policy rates had been the starting point.  A statement that Alan Simpson, Erskine Bowles, Alice Rivlin, and Pete Domenici released as part of CRFB’s “Moment of Truth” project makes some of the same points in abbreviated form.[5]

While the CRFB analysis emphasizes tax-expenditure savings that the JCT study does not include, it tends to gloss over the fact that in key areas — such as the complete elimination of deductions for mortgage interest and charitable contributions — the JCT study assumes savings substantially largerthan those that policymakers are likely to be able to pass.  By maintaining that it is possible to lower tax rates to the levels proposed by Bowles-Simpson or Rivlin-Domenici while raising enough revenues to make a sizeable contribution to deficit reduction, the CRFB analysis tends to obscure the exceedingly difficult nature of this challenge.  As explained below, the Bowles-Simpson and Rivlin-Domenici plans would require enactment of radical tax-expenditure changes that go far beyond the landmark base-broadening of the Tax Reform Act of 1986.  Many tax policy experts and practitioners believe such a sweeping rewrite of the tax code would have virtually no chance of passage. 

Moreover, to hit their targets for both increasing revenue for deficit reduction and lowering tax rates, without making regressive tax changes that favor the wealthy over lower- and middle-income Americans, both Bowles-Simpson and Rivlin-Domenici had to take virtually every possible step to limit tax expenditures.[6]   As a result, there is essentially no margin for slippage in their plans.  If Congress failed to go as far as those plans propose in cutting back or eliminating any significant tax expenditure, particularly in eliminating the preferential treatment for capital gains, tax rates would have to be higher than those plans call for or the contribution of revenues to deficit reduction would have to be lower.  As the description below of the specific tax-expenditure changes in the Bowles-Simpson and Rivlin-Domenici plans indicates, it is virtually certain that lawmakers will not embrace such plans in their entirety or anything close to them.

Another significant problem is that CRFB’s estimates of the savings that can be achieved by eliminating most or all of the employer health exclusion — and then applied to lowering tax rates — is very likely too large.  CRFB appears to assume that all of the resulting increase in tax revenues would be available for this purpose.  But any estimate of the savings from substantially scaling back or eliminating the tax exclusion for employer-sponsored insurance needs to take into account the increase in federal health spending — for Medicaid and subsidies to purchase coverage in the new health insurance exchanges — that would result if most or all of the employer exclusion is eliminated, and the number of employers providing health coverage consequently declines (as almost certainly would occur).  If all of the increase in revenues is used to lower tax rates while federal health care spending rises, the result will be an increase in the deficit.

Tax-Expenditure Changes Likely to Finance Much Smaller Rate Cuts Than Some Hope

In its analysis, JCT measured tax changes from a current-law baseline, which reflects the tax policy that will be in effect starting next year, assuming that various tax cuts (including those enacted in 2001 and 2003) expire at the end of this year as scheduled.  Under this baseline, both tax rates and revenues are higher than under a current-policy baseline, which assumes that most of the expiring tax cuts are permanently extended.  JCT also assumed in its analysis that, over the next ten years, tax reform would produce roughly the same level of revenues as under current law. 

The JCT analysis found that the elimination of all itemized deductions plus other changes could finance only a modest 4 percent reduction in tax rates, lowering the top income tax rate from 39.6 to 38 percent.  CRFB notes that if JCT had started from a current-policy baseline — and assumed that tax reform would maintain revenues at the current-policy level — the top rate could have been reduced to a substantially lower level. 

To be sure, if one undertakes tax reform that is revenue-neutral relative to a current-policy baseline, both rates and revenue levels after tax reform will be substantially lower — and deficits substantially higher — than if one undertakes revenue-neutral tax reform relative to a baseline that assumes that higher tax rates will take effect in January as scheduled.  But regardless of the tax-rate baseline one uses, once political realities are taken into account, changes in tax expenditures are likely to finance much more modest reductions in tax rates than some hope for.

While CRFB argues that the JCT study took into account only a subset of tax expenditures and points to other plans that eliminate or cut heavily additional tax expenditures to finance significantly deeper rate cuts (a proposition discussed in more detail in the next section), it effectively discounts the difficulty of achieving even the policy changes highlighted in the JCT study.  The JCT analysis assumes extensive changes:  the elimination of all itemized deductions (including those for mortgage interest, charitable contributions, and state and local taxes), the taxation of capital gains and dividends at ordinary income tax rates, and repeal of tax-free interest on new state and local bonds.  It is extremely unlikely that all of these changes, which would raise more than $2 trillion over ten years, could ever be achieved in the real world, given the popularity of these provisions and the forces lined up to protect them.  Policymakers will need to focus in a realistic manner on how much savings they can actually secure from broadening the tax base before theytake any action toembrace significantly lower rates as a basic tenet of deficit-reducing tax reform. 

Finally, while JCT started from a current-law baseline that results in higher tax rates than if it had started from a current-policy baseline, the JCT analysis also assumed that all of the tax-expenditure savings would go for rate reduction and none for deficit reduction.  The current-law baseline has substantially lower deficits than the current-policy baseline.  Consequently, applying the JCT study’s approach to a current-policy baseline and assuming that all of the savings go for rate reductions would not only entail relying on politically unrealistic assumptions about eliminating all itemized deductions but also would frustrate the goal of deficit reduction.  If a current-policy baseline — with its much higher deficits — is used, then a very large share of the savings from curtailing tax expenditures will need to go to deficit reduction rather than rate reduction, if a balanced deficit-reduction plan is to be achieved.

Low Top Rate in Bowles-Simpson and Rivlin-Domenici Would Be Very Difficult to Achieve

CRFB points to the Bowles-Simpson and Rivlin-Domenici plans as examples of how reforms that sharply lower rates can raise sufficient revenues both to pay for the rate cuts and to increase revenues above current-policy levels for deficit reduction.  It notes that both plans got the top income tax rate below 30 percent.  The Bowles-Simpson report contained no formal tax reform plan, only an “illustrative” option to show that it was possible (at least on paper) to produce the required tax-expenditure savings.  No commission member voting for the overall deficit reduction plan in December 2010 needed to indicate support for this hypothetical tax-reform option. 

While the Bowles-Simpson illustrative plan and the tax reform component of the Rivlin-Domenici plan hold attraction for many policy analysts (including a number of us at CBPP), the chances that they — or something similarly radical — could actually win approval in Congress are slim.  There is little chance either party will support changes as far-reaching as these, which go far beyond the type of base-broadening measures enacted in the Tax Reform Act of 1986.  For example, both plans:[7] eliminate or drastically scale back nearly all itemized deductions, replacing some key ones (such as the mortgage interest and charitable deductions) with tax credits worth, overall, a fraction of the current deductions, and eliminating other deductions (such as for state and local taxes) without putting anything in their place;[8]

  • tax capital gains and dividends at ordinary income tax rates and require the full taxation of any unrealized capital gains when the individual who holds the assets dies;
  • sharply curtail tax benefits for retirement savings and the investment earnings (or “inside build-up”) of life insurance;
  • phase out the exclusion for employer-provided health insurance and eliminate flexible savings accounts for out-of-pocket health costs; and
  • eliminate dozens of lesser-known tax expenditures such as the exclusion of workers’ compensation benefits and military disability pensions from income. 

We raise this not to disparage these proposals, a number of which have merit, but to illustrate the extreme political difficulty of passing tax-reform changes as extraordinarily sweeping as those in the plans offered by Bowles-Simpson and Rivlin-Domenici. [9]

Many analysts and practitioners share the view that changes in tax expenditures as far-reaching as these are unrealistic.  For instance, the Congressional Research Service (CRS) has advised that it would likely prove very difficult to secure tax expenditure savings beyond $100 to $150 billion a year — an amount insufficient to reduce rates and shrink the deficit on anything close to the scale called for in the Bowles-Simpson illustrative option or the Rivlin-Domenici plan.[10]

Similarly, John Buckley, former chief of staff of JCT and former chief tax counsel for the House Ways and Means Committee, wrote in Tax Notes (about the JCT analysis and other responses to it), “One thing is absolutely clear.  The large numbers [for tax-expenditure savings] that have been tossed around in the debate up to this point vastly overstate the real potential revenue gains from elimination of tax expenditures.”[11]  

Such skepticism should not discourage efforts to enact ambitious tax reform.  But it should keep deficit-reduction advocates from making large reductions in tax rates a condition for raising revenues as part of a deficit-reduction package on the assumption that such reductions in tax rates and deficits will be paid for by dramatic and unprecedented cutbacks in tax expenditures.

Shrinking the Tax Exclusion for Employer-Sponsored Insurance Would Raise Federal Spending

In making its case that the JCT analysis understates the potential for rate reduction, one of CRFB’s main arguments is that eliminating or substantially reducing the employer health exclusion — which it terms the largest tax expenditure in the tax code, and which the JCT analysis didn’t cut back or eliminate — could help finance much deeper reductions in tax rates.  Eliminating or sharply reducing the employer exclusion would certainly produce more tax revenue.  But it also would increase federal spending, by increasing enrollment and costs in Medicaid and in the subsidies established by the Affordable Care Act (ACA) to help near-poor and middle-income families buy health insurance in the new insurance exchanges.  If the tax savings from eliminating much or all of the employer exclusion were devoted entirely to tax-rate reductions, the net effect would be to increase the deficit, because federal spending would rise while overall revenues remained unchanged.

Large Rate Cuts Raise Concerns About the Progressivity of the Tax Code

An important principle of the Bowles-Simpson plan is that deficit reduction should not harm the most vulnerable members of society or increase poverty or inequality.  (The Rivlin-Domenici plan was consistent with that principle.)  Bowles-Simpson set as a goal that tax reform should make the tax code at least as progressive as it will be once the upper-income tax cuts expire, a goal that Rivlin-Domenici achieved.  Unfortunately, the issue of progressivity is often lost in current discussions about how low rates can go in tax reform. 

The Bowles-Simpson illustrative tax reform option and the Rivlin-Domenici plan lowered tax rates sharply and raised significant revenue relative to current-policy levels.  Both plans also proposed very progressive tax expenditure reforms in order to offset the substantial, highly regressive effects of their across-the-board tax rate cuts.  They did so primarily by:

  • protecting refundable tax credits, which are the tax expenditures most targeted to low- and moderate-income working families; and
  • eliminating the most regressive tax preferences and sharply scaling back other tax preferences that disproportionately benefit the most affluent members of society. 

For instance, both plans would eliminate the preferential tax rates for capital gains and dividends and end the forgiveness of capital gains tax at death.  The benefits of both of these tax preferences are highly skewed to people at the top of the income spectrum.  Both plans also sharply curtailed tax preferences for saving, which disproportionately benefit high-income households.  Similarly, both plans converted the deductions for mortgage interest and charitable contributions into tax credits to sharply limit the benefits for high-income families and better target these tax benefits on people with low or moderate incomes.

Progressivity goals are an essential component of tax reform, given growing income inequality and the fact that the tax code is the principal mechanism through which the wealthiest Americans can be asked to make a significant contribution to deficit reduction.  But many who advocate achieving sharply lower tax rates through tax reform have not shown similar enthusiasm for curtailing tax preferences for capital income.  If tax rates are to be substantially lower, it likely will not be possible to keep tax reform from being regressive and favoring the most well-off at the expense of Americans of lesser means unless the capital gains tax break is sharply scaled back or ended (as it was under the 1986 Tax Reform Act), other tax expenditures are scaled back in a progressive manner, and refundable tax credits for lower-income working families are protected.

The Bowles-Simpson illustrative plan and the Rivlin-Domenici proposal were crafted to meet the competing goals of rate reductions, deficit reduction, and progressivity.  These plans leave little room for maneuver, however, given that they leave virtually no tax expenditure (other than refundable tax credits) untouched.  Policymakers ought to keep these realities in mind when assessing the viability of sharply lower tax rates being part of a tax reform plan that also seeks to make a significant contribution to deficit reduction without reducing the progressivity of the tax code.

Employer-sponsored insurance is the primary source of health coverage for Americans; 156 million non-elderly people obtained coverage through employers in 2011.  Elimination or large-scale reduction of the tax exclusion for employer-sponsored insurance would likely lead a number of employers to drop health coverage.  This would likely be especially true among small employers, who face no penalty under the ACA if they don’t offer insurance.[12]   A considerable number of larger employers would likely drop coverage as well, because under the ACA, the penalty they would face for not offering insurance is modest relative to the cost of health coverage for their employees and their families. 

If most or all of the tax exclusion is eliminated, an appreciable number of employers could drop coverage, especially since employers would know that most of their employees and their families would have other ways to secure coverage.  This would increase federal spending in two ways:

  • more low-income employees and their family members would enroll in Medicaid; and
  • more near-poor and middle-income employees and their family members would qualify for premium and cost-sharing subsidies to buy insurance in the exchanges (and, facing the ACA’s mandate to have coverage, would enroll).

There also could be a third effect:  more people would likely become uninsured, especially working-poor individuals in states that do not adopt the ACA’s Medicaid expansion, which the Supreme Court has made a state option.  (In addition, as John Buckley’s recent Tax Notes article explains, eliminating the employer health exclusion would be regressive, raising taxes as a percentage of income much more for middle-income people than for those much higher on the income scale.[13]

The CRFB materials (and the statement from Alan Simpson et al.) imply that all of the savings from eliminating most or all of the employer exclusion could be used to finance the costs of sharply reducing tax rates.  But the savings actually produced would be the net of the increase in revenue minus the increased spending for Medicaid and health insurance subsidies.  Neither the Congressional Budget Office (CBO) nor, to the best of our knowledge, any other such institution has estimated the increase in federal health-care expenditures that would result, and hence the amount of net savings that would be produced.  CBO also hasn’t analyzed the effect that such a measure would have on the number of Americans without health insurance.  In short, such a change could have important impacts that policymakers will need to have information on, and to take into account, in making decisions on these matters.

The reductions in income tax rates that various tax-reform options claim they can finance consequently overstate the savings from scaling back or eliminating the employer exclusion because they do not reflect the resulting increase in federal expenditures.  Judgments about how much tax rates could be lowered by proposals that substantially curtail the employer health exclusion are premature until full information and cost estimates are available on this matter, and until we also have analysis on the impact this would have on the number of Americans who are uninsured.

Importantly, these caveats also apply to proposals to shrink tax expenditures through some form of per-taxpayer limit on such expenditures, such as a proposal by Martin Feldstein, Daniel Feenberg, and CRFB’s Maya MacGuineas to limit the overall amount that a taxpayer can claim for a package of major tax expenditures to 2 percent of the taxpayer’s adjusted gross income.[14]   The estimates from Feldstein et al. show that roughly half of the savings from their tax-expenditure limit would come from sharply reducing the tax exclusion for employer-sponsored insurance.[15]   To secure savings of that magnitude, a large share of the exclusion would have to be eliminated.  Feldstein and his colleagues cite the full revenue increase resulting from their tax-expenditure limit as available for tax-rate reductions and/or deficit reduction, without taking into account the offsetting increase in federal health expenditures that would ensue.[16]


While CRFB acknowledges that it will require making “very hard choices” to achieve the amount of tax-expenditure savings needed under the Bowles-Simpson illustrative option or the Rivlin-Domenici plan to lower rates and reduce the deficit, it understates the challenge.  Policymakers are highly unlikely to enact the far-reachingtax-expenditure reductions those plans would require.  In assessing the JCT study, CRFB is effectively conducting a technical exercise in what is a highly political environment.

We are particularly concerned that CRFB’s analysis risks creating misunderstanding among policymakers grappling with the nation’s deficit problems, by fostering an expectation that sharp tax-rate cuts can be part of revenue-raising tax reform.  We fear this could lead some policymakers to insist, as a condition of agreeing to a framework for a major deficit-reduction package, on dramatic tax-rate reductions that cannot actually be achieved in the real world if the tax-reform part of the package is to make a substantial contribution to shrinking the deficit.  Some policymakers could reject — or block — worthy tax reforms that raise needed revenue in a progressive manner but that fail to lower rates as much as called for in a pre-set framework.  A worse result would be if policymakers push hard for the lower rates and end up giving short shrift to raising meaningful revenue for deficit reduction.

Policymakers who are serious about deficit reduction should be wary of plans that count on financing large reductions in both tax rates and the deficit through sweeping tax-expenditure changes.  It would be especially misguided for policymakers to require — in initial legislation setting a framework and parameters for subsequent tax-reform/deficit-reduction legislation — that tax reform must reduce the top income tax rate to a specified low level before policymakers know the specific tax-expenditure savings they can actually achieve.  Policymakers should not put rate reduction ahead of deficit reduction.

Finally, estimates of savings from substantially scaling back or eliminating the employer health exclusion — through either regular tax-reform legislation or a tax-expenditure limitation — need to take into account the added federal expenditures that such an action would engender, as well as the impact on the number of uninsured (and on the progressivity of the tax code).  It would be especially disheartening if policymakers used large cuts in the employer exclusion to justify deeper tax-rate reductions for wealthy households, and then used the ensuing increases in Medicaid and health insurance subsidy costs to justify deeper cuts in those programs that resulted in more poor and middle-income Americans becoming uninsured or underinsured.

End Notes

[1] For instance, see Jane G. Gravelle and Thomas L. Hungerford, “The Challenge of Individual Income Tax Reform: An Economic Analysis of Tax Base Broadening,” Congressional Research Service, March 22, 2012 and Samuel Brown, William G. Gale, and Adam Looney, “On the Distributional Effects of Base-Broadening Income Tax Reform,” Tax Policy Center, August 1, 2012,


[3] Chuck Marr and Chye-Ching Huang, “How Tax Reform Could Become a Trap,” Center on Budget and Policy Priorities, June 8, 2012,

[4] Committee for a Responsible Federal Budget, “JCT Report Does Not Conflict With Rate-Reducing Tax Reform,” October 12, 2012,; Committee for a Responsible Federal Budget, “Tax Reform: Reducing Tax Rates and the Deficit,” October 15, 2012,

[5] Joint Statement by Alan Simpson, Erskine Bowles, Alice Rivlin, and Pete Domenici, October 12, 2012,

[6] The only tax expenditures they did not propose cutting were refundable tax credits for low-income working families; avoiding cuts in these credits is essential to prevent tax reform from having regressive effects that punish working-poor and near-poor families. 

[7] See The National Commission on Fiscal Responsbility and Reform, “The Moment of Truth,” December 2010; and The Bipartisan Policy Center Debt Reduction Task Force, “Restoring America’s Future,” November 2010.  Also see Tax Policy Center description of the Bowles-Simpson illustrative plan as modelled for distributional estimates, Tax Policy Center Table T10-0258,; Eric Toder, “Distribution Estimates for Bipartisan Policy Center Tax Reform Proposal,” Tax Policy Center, November 16, 2010,

[8] The Bowles-Simpson illustrative option called for converting the mortgage interest deduction into a non-refundable tax credit equal to 12 percent of a taxpayer’s interest costs on a mortgage up to $500,000, while the Rivlin-Domenici plan would create a 15 percent refundable credit on mortgage interest up to $25,000.  Both credits applied only to primary residences.  For charitable deductions, the Bowles-Simpson option called for a 12 percent non-refundable credit for contributions above 2 percent of adjusted gross income; Rivlin-Domenici called for a 15 percent refundable credit.

[9] Note that the original Rivlin-Domenici plan also called for instituting a type of value-added tax to raise revenue.  That proposal has been dropped from the updated Rivlin and Domenici tax reform plan, in part because it lacked political support.

[10] Jane G. Gravelle and Thomas L. Hungerford, “The Challenge of Individual Income Tax Reform,” Congressional Research Service, March 22, 2012,  The $100 to $150 billion savings figure is for 2014.

[11] John Buckley, “Does the JCT Experiment Poke Holes in the Romney Tax Plan?” letter to the editor, Tax Notes Today, October 22, 2012.

[12] Employers with fewer than 50 full-time-equivalent employees (FTEs) are exempt from health reform’s employer responsibility requirement.

[13] Buckley, op. cit.

[14] Martin Feldstein, Daniel Feenberg, and Maya MacGuineas, “Capping Individual Tax Expenditure Benefits,” NBER Working Paper 16921, April 2011,

[15] Feldstein, Feenberg, and MacGuineas estimate that in 2011, if the cap had been in place it would have raised a total of $278 billion, but if employer-sponsored insurance were excluded from the cap it would have raised $140 billion less.

[16] The paper states that with a cap in place, “the additional tax revenue would be $278 billion...”, and it assumes that the effect on the deficit would be the same as the additional income tax revenue, stating, “Our analysis implies that a two percent of AGI cap on tax expenditure benefits would reduce the 2011 fiscal deficit by $278 billion dollars....”  (Emphasis added.) See