Tax Simplification: Issues and Options

William G. Gale
William G. Gale The Arjay and Frances Fearing Miller Chair in Federal Economic Policy, Senior Fellow - Economic Studies, Co-Director - Urban-Brookings Tax Policy Center

July 17, 2001

Mr. Chairman and Members of the Committee:

Thank you for providing me with the opportunity to present my views on issues and options related to simplification of the tax code. My testimony is divided into two sections. The first provides a summary of my principal conclusions; the second provides the economic analysis that supports these views.


Basic Issues

  • The notion that taxes should be simpler is one of the very few propositions in tax policy that generates universal agreement. The fundamental paradox of tax simplification is that, although everyone thinks taxes should be simpler, almost every year tax policy becomes more complex. This suggests that pleas for simplification need to be buttressed by an understanding of the causes of complexity and the likely outcome of simplification efforts.

  • Simpler taxes have numerous benefits. They would reduce taxpayers’ costs of complying with the tax system in terms of time, money, and mental anguish. They would likely raise the use of tax subsidies-say, for education-reduce unintentional tax evasion, and increase the likelihood that taxpayers would see the tax system as fair.

  • But simpler taxes also have costs. In particular, they reduce the ability of policy makers to achieve other goals of tax policy. Features of the tax code that are designed increase tax equity, police intentional tax evasion, or encourage some particular activity often increase complexity.

  • These considerations of the costs and benefits of simplification have several implications for thinking about tax complexity:

    —Tax complexity arises in large part as the result of a trade-off between simplicity and other goals.

    —As a result, the fundamental issue is not the overall level of complexity; rather it is whether particular tax provisions, tax systems (or alternative means of providing government services, such as spending or regulations) provide good value for the complexity they create. This depends on the magnitude and incidence of the costs and benefits of complexity, where the benefits include the extent to which complexity aids in achieving other policy goals.

    —Moreoever, the factors that generate complex tax systems-policy trade-offs, politics, and taxpayers’ desire to reduce their own tax burdens-are not features of particular tax policies per se. They will likely remain in force even if the tax system were reformed or replaced. As a result, an analysis of the extent to which policy changes can affect tax complexity should incorporate these factors.

Simplification and EGTRRA

  • The new tax law provided a few simplifying measures (with respect to the EITC, the repeal of limitations on itemized deductions and personal exemptions, and the reduction in marginal tax rates).

  • On net, however, the new law made taxes much more complex and made tax planning much more difficult. This is a result of the “sunset” provisions, the long and variable phase-ins and abrupt phase-outs of numerous provisions, the failure to address the long-term AMT problem and the willingness to increase the severity of that problem, complicated provisions regarding the estate tax, and an increase in targeted subsidies in education and retirement saving.

  • The new law also substantially reduced future prospects for simplification because it allocated such a large share of projected budget surpluses toward other uses, and because politically viable simplification often requires reduced tax revenues to ensure that there are few losers from the reform.

Directions for simplifying taxes

  • The key to tax simplification is to make fewer distinctions across economic activities and personal characteristics. Taxes should be imposed on a broad base at relatively low rates that do not vary by income source or expenditure type. Progressivity should be embodied in the rate structure and the tax base, not in the design of specific provisions. Universal exemptions, deductions, or credits are much simpler than targeted ones.

  • The following types of reforms could make taxes simpler as well as fairer and more conducive to economic growth: addressing the uncertainty created by sunset and phase-out provisions of EGTRRA; reforming the individual AMT; eliminating (or at least coordinating) phase-outs of tax credits; coordinating and consolidating provisions with similar purposes (including provisions aimed at low-income families with children, retirement saving and education); and reducing the top tax rates in conjunction with taxing capital gains as ordinary income.

  • Filing and record keeping could be enhanced by consideration of “return-free” tax systems, and by significantly raising the standard deduction.

Estimates of complexity in the current system

  • Reliable estimates of the costs of compliance, administration, and enforcement of the income tax vary widely, due in part to inadequate data. The best estimate is that, in 1995, those costs ranged between $75 billion and $130 billion, or between 10 and 17 percent of revenues. These costs are distributed mainly to taxpayers in higher income groups.

  • There is wide disagreement on the compliance costs of the estate tax, but the most reliable estimates place those costs at about 10 percent of revenues.

Complexity and fundamental tax reform

  • Some have turned to new tax systems-such as a flat tax or a national retail sales tax-as an alternative way to simplify taxes. In their idealized form, these taxes are extremely simple. But a crucial caveat is that no country has successfully enacted or administered a high-rate national retail sales tax or a flat tax. The features of tax systems that exist in the real world have been forged through a combination of revenue requirements, political pressures, responses to taxpayer avoidance and evasion, lobbying, and other processes that any operating tax system would eventually have to face. Notably, all of these factors tend to raise complexity. In contrast, tax systems that exist only on paper-such as the NRST and the flat tax-appear simpler in significant part because they have not had to face real-world tests.

I. Tax Complexity: Some basics

A. Measuring complexity

Tax complexity has many dimensions and could plausibly be defined in different ways. Following Slemrod (1984), we define the complexity of a tax system as the sum of compliance costs-which are incurred directly by individuals and businesses-and administrative costs-which are incurred by government. Compliance costs include the time taxpayers spend preparing and filing tax forms, learning about the law, and maintaining record-keeping for tax purposes.

The costs also include expenditures of time and money by taxpayers to avoid or evade taxes, to have their taxes prepared by others, and to respond to audits, as well as any costs imposed on any third-parties, such as employers. Administrative costs, although incurred by government, are ultimately borne by individuals. These costs include the budget of the tax collection agency, and the tax-related budgets of other agencies that help administer tax programs.

Defining complexity as the total resource cost provides a quantitative measure by which different tax systems can be compared, and by which the administrative aspects of a particular tax system can be evaluated relative to its impacts on equity, efficiency, and revenue. But the definition is not ideal. Slemrod (1989a) points out that a particular subsidy could be so complicated that few taxpayers use it. If it were simplified, and enough additional people used the subsidy, total resource costs would rise, even though the subsidy itself had become less complicated.

A number of issues arise in efforts to measure tax complexity: First, permanent and transitory costs may differ. A new tax provision may raise compliance costs temporarily, as people learn about the change, even if it reduces costs in the long-term. Likewise, for administrative costs, the capital cost of upgrading IRS computers might appear as a current-year budget expenditure rather than being amortized over time. Second, only the incremental costs due to taxes should be included. Even with no taxes, firms would need to keep track of income and expenses to calculate profits, and individuals would engage in financial planning. This activity should be omitted from compliance cost measures. Third, an analysis of tax complexity alone may generate misleading conclusions. Governments can impose policies via taxes, spending, regulations, or mandates. Any tax provision can be made simpler by eliminating it, but if it then is recreated as a spending program, the overall complexity of government may rise.

B. Benefits of simpler taxes

Simpler taxes would be beneficial in a number of ways. First, simpler taxes would reduce taxpayers’ of complying with the tax system in terms of time, money, and mental anguish. By reducing these costs, simplification would reduce the overall burden of taxation.

Second, tax provisions that are simpler are more likely to be used. Provisions aimed at encouraging certain activities-such as saving for college-will be less likely to be used and hence less effective if people cannot understand how they work.

Third, making taxes simpler would probably raise compliance rates (i.e., reduce illegal tax evasion). To some (uncertain) extent, people do not pay taxes because they do not understand the tax law. Clarifying and simplifying tax rules can only help to make people understand the law better, and would likel

Finally, simpler taxes would generate more public support and thus should be an essential part of any effort to improve the delivery of government services. The biggest complaint about the tax system for many people is not the amount of taxes they pay but rather the sheer, and seemingly needless, complexity of what appear to be everyday tax situations (Graetz 1997).

C. Which features of the tax code generate complexity?

The level of complexity can be influenced by structural elements-such as the tax base, the tax rate structure, and the allowable deductions, exemptions, and credits-as well as by administrative features of the tax code. The three most discussed tax bases are income, wages, and consumption. Holding the other features of the tax system constant, income is the most difficult of the three bases to tax. Income may be decomposed into its sources-wages and capital income-or its uses-consumption and saving. For a wide variety of measurement and timing reasons, it is generally easier to tax wages than capital, and easier to tax consumption than saving.

Tax rates are typically either graduated, like the current income tax, or flat, like the payroll tax. Flat-rate taxes can have lower compliance costs than graduated taxes. The presence of graduated rates gives taxpayers incentives to avoid taxes by shifting income over time or across people. And flat-rate taxes allow more efficient administrative structures to function. Taxes imposed at flat rates can be easily collected at source, since the rate does not vary across taxpayers.

Exemptions, deductions or credits that are universal create little complexity. However, targeted provisions require clear definitions of eligible taxpayers and activities, and can create compliance headaches. Finally, different ways of administering taxes may affect complexity. For example, withholding taxes at source or eliminating the requirement to file a tax return could reduce compliance costs for individuals.

The discussion above suggests that, other things equal, the simplest system would tax consumption at a flat rate with universal deductions, credits or exemptions, and with withholding at source. Yet, the U.S. and many other countries tax income on a graduated basis, with numerous targeted credits and deductions, and with withholding at source only for certain types of income. Given the prevalence of these alternative systems, and absence of any country that taxes only in the simplest way described above, it is instructive to ask why existing systems deviate so strongly from the simplest structure.

D. Why are taxes complex?

Any plea for simpler taxes has to start by addressing a basic problem: If everyone thinks taxes should be simple, why are taxes so complicated? At least four factors help explain why taxes become complicated and suggest keys to making taxes simpler.

The first, and most important, is conflict among the consensus goals of tax policy. Although almost everyone agrees that taxes should be simple, most people also agree that taxes should be fair, conducive to economic prosperity, and enforceable. Even if all parties agree on these goals, they do not typically agree on the relative importance of each goal. As a result, policy outcomes usually represent efforts to balance one or more goals against the others. That is, sometimes a certain amount of complexity is created or permitted in order to help achieve other policy goals. For example, attempts to make taxes fairer often conflict with attempts to make taxes simpler. Most countries tailor tax burdens to the characteristics of individual taxpayers. This may improve tax equity, but it also creates complexity. It requires tracing income or consumption from the business sector to the individual. It requires reporting and documenting individual characteristics such as marital status, number of dependents, and age, as well as the composition of expenditures or income. It allows tax rates that vary with individual characteristics, creating opportunities for tax avoidance.

In this context, tax complexity is like air pollution: it is an unfortunate and undesirable consequence of products or services that we, as a society, desire. Just as the optimal level or air pollution is not zero-since that would mean that many of the goods and services society cherishes could not be produced-the optimal level of tax complexity is not zero. And just as we should seek the most efficient ways to reduce air pollution, we should also seek the most effective ways to make taxes simpler.

The second factor that generates tax complexity is the political process. Politicians and interest groups have interests in targeted subsidies that reduce taxes for particular groups or activities. But targeted subsidies inevitably make taxes more complex by creating more distinctions among taxpayers and among sources and uses of income.

Third, some complexity is necessary to deter tax avoidance. Taxpayers have every right to reduce their taxes by any legal means. But this activity inevitably raises questions about whether particular activities or expenditures qualify for tax-preferred status. The Treasury Department responds with complex rules designed to limit avoidance. Taxpayers in turn respond by inventing complex transactions to skirt the new rules. This can create a vicious cycle that leads to more and more complex rules and increasingly sophisticated and complex avoidance strategies.

Fourth, many complicated provisions were enacted to raise revenue or limit revenue losses during times of rampant budget deficits. For example, the landmark Tax Reform Act of 1986 (TRA)-a remarkable accomplishment in many respects-fell short of its goal of simplicity to meet the requirement of “revenue neutrality.” TRA created several complicated phase-outs and hidden taxes in order to raise revenue and meet distributional targets. Insofar as complexity has arisen from efforts to limit revenue loss, the surplus that existed at the beginning of this year and the political consensus in favor of some sort of tax cuts created an opportunity to simplify taxes. In that regard, and as discussed further below, the recent tax act is not only a missed opportunity for simplification, but may also have used up whatever funds would otherwise have been available to support simplification efforts.

E. Implications for thinking about tax simplification

Recognition of these factors has several important implications for the study of tax complexity.

  • First, the fundamental question is not the overall level of complexity, but whether particular tax provisions, tax systems (or alternative means of providing government services, such as spending or regulations) provide good value for the complexity they create. This depends on the magnitude and incidence of the costs and benefits of complexity, where the benefits include the extent to which complexity aids in achieving other policy goals.

  • Second, the factors that generate complex tax systems-policy trade-offs, politics, and taxpayers’ desire to reduce their own tax burdens-are not features of tax policies per se. They will likely remain in force even if the tax system were reformed or replaced. As a result, an analysis of the extent to which policy changes can affect tax complexity should incorporate these factors.

  • Third, there is an important distinction between private and social gains or costs. Suppose everyone had to fill out five extra lines of the tax form to receive a $1,000 tax cut. Each person might regard that as “good complexity,” worth the cost of providing extra information. But, holding tax revenues constant, the revenue would still have to be raised from somewhere, so the net tax cut would be zero—that is, everyone’s tax “cut” would be from a higher initial tax liability and net taxes would be the same. Thus, from a social perspective, the sum of all individuals’ “good complexity” could be zero or negative.

II. Simplification and the new tax law

A. Provisions

The Economic Growth and Tax Relief Reconciliation Act (EGTRRA) was signed into law by President Bush on June 7, 2001.

Both the most important and most novel aspect of EGTRRA is the general provision that the entire bill “sunsets” at the end of 2010. All provisions of the bill are eliminated and the tax code at that point reverts to what it would have been had the tax bill never been passed.

The act also contains numerous specific provisions. Some of these are listed in table 1 and described here along with their effective phase-in and phase-out dates. They are listed in order of the tax cut provided when fully phased in.

  • Reduce marginal income tax rates of 28 percent or higher:

    The 28, 31, and 36 percent tax rates (which apply to married households with taxable income above $45,200, $109,250, and $166,500, respectively) will each fall by 3 percentage points, and the 39.6 percent top rate (which applies to married households with taxable income above $297,350) will fall to 35 percent. Each of these rates declines by 0.5 percentage points for tax years beginning after December 31, 2000, with further declines in 2004 and 2006.

  • Eliminate the estate tax:

    The effective exemption in the estate tax is raised from $675,000 currently to $1 million in 2002, and then gradually to $3.5 million in 2009. The top effective marginal tax rate is reduced from 60 percent currently to 50 percent in 2002 and then gradually to 45 percent in 2009. The credit for state-level estate taxes is gradually phased out between 2002 and 2005, after which it is replaced by a deduction. This change finances about one-quarter of the cost of the entire reduction in federal estate taxes. In 2010, the estate and generation-skipping transfer taxes are repealed, the highest gift tax rate is set equal to the top individual income tax rate, and the step-up in basis for inherited assets that have capital gains is repealed.

  • Create a new 10 percent income tax bracket:

    A new tax bracket of 10 percent is carved out of the first $6,000 of taxable income for singles, and the first $12,000 of taxable income for married couples. This income is currently taxed at a 15 percent rate. Starting in 2002, the 10 percent bracket is implemented by changing the tax rate and withholding schedules. In 2001, the 10 percent bracket is implemented by providing an advance credit for 2001 taxes. The advance credit is a one-time payment of the minimum of the taxpayer’s 2000 income tax payment (the payment due on April 15, 2001) or $300 ($600) for singles (married couples). This payment is intended to substitute for the 10 percent tax bracket in 2001, but for some taxpayers it will serve more as a rebate of the previous years’ taxes because taxpayers who do not owe taxes in 2001 but did owe them in 2000 will not have to repay the rebate they receive.

  • Increase and expand the child credit:

    The child credit is gradually increased, from $500 currently to $1,000 by 2010. The child credit is also made refundable to the extent of 10 percent of a taxpayers earned income above $10,000 for 2001-4 and 15 percent for subsequent years, with the $10,000 amount indexed for inflation. Refundability improves the access to, and amount of, child credit benefits for low-earning households.

  • Partially address the marriage penalty:

    The standard deduction for married couples gradually rises from 174 percent to 200 percent of the standard deduction for singles in the years 2005 to 2009. The top income level in the 15 percent bracket for married couples gradually rises from 180 percent to 200 percent of the similar level for singles from 2005 to 2008. The beginning and ending of the EITC phase-out will gradually increase by $3,000 by 2008, and will be indexed for inflation thereafter.

  • Repeal of limitations on itemized deductions and phase-outs of personal exemptions:

    The repeals are phased in between 2005 and 2009.

  • Pension and IRA provisions

    Contribution limits for Individual Retirement Accounts and Roth IRAs will rise to $5,000 by 2008 and be indexed for inflation thereafter. Contribution limits to 401(k)s and related plans will rise gradually to $15,000 in 2006 and then be indexed for inflation. Additional so-called “catch-up” contributions of up to $5,000 for anyone over the age of 50 will be permitted. Roth 401(k) plans can be established starting in 2006. A non-refundable credit for retirement saving for low-income taxpayers will be available between 2002 and 2006.

  • Education provisions

    Taxpayers may take an above-the-line deduction for qualified higher education expenses, but only for the years 2002 to 2005. Effective in 2002, the contribution limit on education IRAs rises to $2,000 from $500. The definition of qualified expenses from education IRAs is expanded. Pre-paid tuition programs will now benefit from tax-free withdrawals as long as the funds are used for education. Deductions for student loans are made more generous.

  • Temporarily, limited AMT relief

    Between 2001 and 2004, the exemption amount in the individual AMT is increased by $2,000 for single taxpayers and $4,000 for married taxpayers. This provision is abolished at the end of 2004.

B. Effects of EGTRRA on tax complexity

It would be an understatement to say that simplification was not one of the goals of EGTRRA. In fact, the overall net impact of the new tax law will be to make taxes more complicated over time.

There are three bright spots for simplification. First, for the earned income credit, the bill simplifies the definition of earned income, the definition of a qualifying child, and calculation of the credit. This is an important set of changes since it allows benefits to be provided to low-income household

Second, the bill repeals the limitations on itemized deductions and the phase-out of personal exemptions will simplify taxes for high-income taxpayers. These provisions are hidden taxes that serve no purpose that could not be generated by rate adjustments. In fact, the repeal was implemented in exchange for a smaller reduction in marginal tax rates for the highest income taxpayers than would otherwise occur. This trade-off-giving up explicit rate reductions in exchange for provisions that simplify the tax system-could provide a useful model for dealing with the problems created by the alternative minimum tax in the future.

Third, the reduction in income tax rates will indirectly help to simplify tax planning. Increasing the number of tax brackets does not generally make compliance more difficult; taxpayers will continue to look up their tax liability in a tax table. But lower tax rates simplify tax compliance indirectly by reducing the incentive to avoid taxes or find tax shelters.

Despite these changes, however, the overwhelming net effect of the bill will be to make tax filing and tax planning more complex.

  • Complexity due to increased uncertainty: Sunsets and phase-outs

    As noted above, the most novel feature of the bill is the sunsetting of all provisions as of December 31, 2010. In addition, various features of the bill phase-in and phase-out at different times. Taken at face value, these provisions make tax planning more complex, since the tax rules will be changing on a near constant basis, giving taxpayers incentives to shift the level, form and timing of their income and deductions. The good news is that few people take the sunset provisions at face value. The bad news, though, is that not taking them at face value makes tax planning even more complex, since it is not yet known what will replace the sunset and phase-out provisions, or when such provisions will be altered. The prevailing sentiment may be best summed up by Washington Post columnist Al Crenshaw (2001) who noted that “The new tax law doesn’t make planning unnecessary, it just makes it impossible.”

    While sunsets and phase-outs create planning difficulties for any situation, they appear to have particularly egregious effects in at least two areas: estate planning and pension choices. Taxpayers may end up having to make their wills and estate plans contingent on the year in which they die, because the provisions are legislated to change so massively on a year-to-year basis. For pensions, a key goal is to raise employer sponsorship of plans. But employers will naturally be reluctant to incur the fixed costs of creating new plans and educating their employees about the plan, if there is a chance that the plan, or the particular provisions that made the plan worth offering, may not be in existence after a few years.

  • Complexity due to increased number of choices

    Complex rules or documentation procedures are a common source of tax complexity. However, a new and increasing source of complexity might be termed “choice” complexity. This occurs when taxpayers are given numerous subsidies but may only use one or a few of them. This type of complexity has proliferated with regard to retirement saving, where taxpayers have been able to choose to allocate contributions among traditional, Roth, and education IRAs for several years. Under the tax bill, they will soon be able to choose to allocate 401(k) contributions between traditional and Roth plans as well. Similar issues apply to the variety of education subsidies that exist today, and which were expanded in EGTRRA.

    In economic models that feature fully informed consumers who make choices without incurring transactions costs, having more options is always preferable to having fewer options. However, in designing tax policy it is not necessarily the case that more options are always worth the added costs. First, the differences in benefits to a household between choosing one of a set of options versus another in the same set may be smaller than the costs of determining which the best option. But of course the household does not know that until it has undertaken the cost. Second, having more choices, for example with respect to retirement saving, requires more record-keeping by the taxpayer and the government.

  • Alternative minimum tax

    The AMT is a parallel tax system that was created to prevent high-income taxpayers from aggressively using tax shelters and deductions to eliminate their tax burdens. Taxpayers must calculate the AMT if their regular income tax liability is less than their AMT liability. The AMT is quite complex and requires tax filers to make many detailed calculations. Currently, fewer than 2 million taxpayers face the AMT.

    There are (at least) two “AMT problems” facing the tax code currently. The first is that, even without the new tax law, the number of taxpayers facing the AMT is scheduled to rise to about 20 million by 2011 (JCT 2001a). This occurs primarily because the AMT exemption amounts are not indexed for inflation. In addition, the overwhelming reason why these taxpayers will end up facing the AMT is that the personal exemptions and state tax deductions that they take in the regular income tax are not allowed in the AMT. Thus, the AMT will increasingly be capturing more people, and from the perspective of curtailing tax sheltering, the wrong people over time. While the new tax law does not make this problem worse, it does not do anything to fix it, either.

    The second problem is created by the new tax law. By 2010, when the law is fully phased in, JCT estimates that about 35 million taxpayers will face the AMT (JCT 2001a). This occurs because the tax law reduces regular income tax but not (in years after 2004) AMT.

    The bill offers only temporary, partial relief against the AMT, but that provision sunsets after four years. As a result, any gains in simplicity arising from lower income tax rates would be offset several times over after 2004 because lower rates would subject millions of taxpayers to the individual alternative minimum tax.

  • The estate tax

    Abolition of the estate tax sounds, on the surface, like a simplifying measure, but in the tax bill it is not. The bill stipulates three stages for estate taxes: from 2002 to 2009, the tax is modified in many ways. In 2010, the estate tax is abolished and step-up in asset value for inherited assets is repealed. In 2011, the estate tax is reinstated, as is the step-up in asset value for inherited assets.

    This creates several sources of complexity. The first is the sunset provision, as noted above. The second is the transition period before the estate tax is abolished. The estate tax phase-out is slow and involves several changes between now and 2009: the exemptions are raised, the tax rates are reduced, the credit for state taxes is abolished and replaced with a deduction, and gift tax limits are dramatically changed. Both the sunset and the transition make effective estate planning quite complex between 2002 and 2011.

    The third issue is the repeal of basis step-up at death. Under current law, when an heir receives an asset from an estate, the basis price is “stepped up” The new bill features “basis carryover:” heirs inherit an asset’s original basis price. Implementing carryover raises vexing issues. For example, some families would have to keep records for generations to keep track of asset purchase prices and improvements. Carryover basis would raise taxes on many heirs compared with current law unless modest gains are excluded from the new rule. But exempting a portion of capital gains would create a great deal of complexity. For example, under current law, it is easy for a parent to split an estate equally among his or her children. Under basis carryover, the estate would have to decide how to allocate a capital gain exclusion among the children. The assets inherited by children who received equal bequests, but different exclusion amounts, would be worth different amounts on an after-tax basis. A carryover basis provision was enacted in the late 1970s, but was repealed before it took effect because taxpayers complained about the new complexities and problems in implementation. There is no reason to think these issues would be any easier to deal with now.

  • Expansion of targeted subsidies (mainly in education and retirement saving)

    Targeted subsidies complicate taxes. Each program require precise definitions of eligible taxpayers, income levels, and qualifying expenses. Many of the proposed incentives would require separate worksheets or tax forms. The possibility of honest mistakes or fraud would rise commensurately. The government would need to spend more on monitoring or auditing taxpayers, and the programs would likely send more lower- and middle-income households to paid tax preparers. The main culprits along these lines in the tax bill are the education subsidies. As one example, one provision of the bill will let people buy computers, educational software, and internet access for their school-age children with tax-preferred funds.

C. Effects of EGTRRA on prospects for tax simplification

Besides directly complicating the tax code, EGTRRA has substantially dimmed prospects for tax simplification in the future, because the tax act allocates revenues that could otherwise have been used for simplification.

Significant tax simplification almost has to be associated with net tax cuts. The Tax Reform Act of 1986, for example, substantially simplified individual income taxes but also cut the revenue collected from such taxes. The overall act was deemed revenue-neutral because net taxes collected at the corporate level were slated to increase.

Simplification has proven difficult in the past because eliminating loopholes and preferences in a revenue-neutral package of individual income tax changes means that taxes on some people and some activities will rise, while taxes on others will fall. This naturally raises difficult political issues. Achieving simplification in a tax cut package, however, could avoid the politically difficult offsetting revenue increases, giving everyone lower and simpler taxes.

D. Effects of new tax proposals on complexity

In the aftermath of EGTRRA, the Ways and Means Committee has approved HR 7, which among other things would allow households who do not itemize their deductions to take an above-the-line deduction for charitable contributions. Perhaps the most notable feature of this proposal is the tiny contribution limits involved: the provision would allow for up to $25 per person for this deduction in 2002, rising to $100 per person in 2010.

This proposal could simplify matters for the 2 percent of taxpayers who currently itemize, but whose deductions other than charity are less than the standard deduction. But for the roughly 70 percent of taxpayers who take the standard deduction, the change would add complexity. They would need to keep records of contributions, which might be difficult if the contributions are small or in cash. A similar deduction in the early 1980s created serious compliance problems, with many taxpayers claiming undocumented deductions. Both the cap on non-itemized charitable deductions and the interaction of this provision with the phase-out of itemized deductions for high-income taxpayers would complicate choices for some taxpayers and require more auditing and monitoring by the IRS. It is hard to see how the complexity engendered by these provisions would be worth the costs, and lawmakers might consider simply raising the standard deduction instead of providing an above-the-line deduction for charitable giving.

III. Simplifying the existing tax system

Despite the setback that EGTRRA represents for actual and prospective simplification efforts, there are a number of options available to policy makers who are interested in simplifying the existing tax system.

A. Principles

The key to tax simplification is to make fewer distinctions across economic activities and personal characteristics. Taxes should be imposed on a broad base at relatively low rates that do not vary by income source or expenditure type. Progressivity should be embodied in the rate structure and the tax base, not in the design of specific provisions. Universal exemptions, deductions, or credits are much simpler than targeted ones. Broadening the base by eliminating targeted preferences and taxing capital gains as ordinary income directly removes major sources of complexity. Using the revenue raised to increase standard deductions removes people from the tax system, and using the revenue to reduce tax rates reduces the value of sheltering and cheating. Increasing the number of people that face the same “basic” rate facilitates withholding of taxes at the source, which further simplifies taxes and raises compliance. In short, broadening the base and reducing the rates, which in general may be considered efficiency-enhancing, would also simplify taxes (see Pechman 1990, Slemrod 1996, Slemrod and Bakija 1996, Gale 1997, 1998).

Slemrod (1996) refers to such plans as “populist simplification.” That is, they make taxes simpler for a large number of taxpayers, but the overall saving in compliance costs may not be very large. Not all structural reforms, of course, have the same impact on compliance costs. Slemrod (1989b) found no significant saving from changing to a single-rate tax structure. In contrast, eliminating the system of itemized deductions would result in a substantial reduction in expenditures on professional assistance; the impact on total compliance costs, though, varied depending on the model used.

B. Specific Reforms

The following reforms could make taxes simpler as well as fairer and more conducive to economic growth.

  • Address the Uncertainty Created by Sunset and Phase-out Provisions of EGTRRA

    The most urgent simplification need is to clean up the tax planning problems, complexities and uncertainties created by EGTRRA with regard to seemingly capricious phase-in and phase-outs of provisions and the sunsetting of the entire bill. Either the provisions should be made permanent or they should be abolished. Having numerous tax provisions dangle for an indefinite period does not simplify the tax code. (On a related note, it would also make sense to decide whether to keep permanently or to abolish the entire set of temporary tax provisions that existed even before EGTRRA.)

  • Reform the Individual AMT

    To spare middle-income people who were never its target, the AMT should be indexed for inflation, deductions should be allowed for dependents and state and local taxes, and all personal credits should be available against the AMT. Any proposal that cuts regular income tax liabilities should be required to make conforming adjustments to the AMT so that more taxpayers are not subjected to the alternative tax. Some would argue that the AMT should be eliminated altogether. But a reformed AMT would prevent the very wealthy from eliminating their tax liability, and legislators will probably want to be spared the embarrassment of seeing how successfully the well-advised can exploit loopholes.

  • Eliminate (or at least Co-ordinate) Credit Phase-Outs

    A number of credits phase out across different income ranges. Each credit requires separate worksheets and tax calculations. The phase-outs create hidden taxes over the phase-out range, and diminish the effectiveness of the credits in encouraging the activities they are designed to spur.

  • Coordinate and Consolidate Provisions with Similar Purposes

    In a number of areas, numerous provisions-each with slightly different rules-apply to the same general activity. Coordinating or consolidating the following provisions would simplify taxes, often with little or no forgone revenue:

    EITC, Dependent Exemption and Child Credit Several recent proposals would combine features of the tax code that deal with families with children. Coordinating the three tax subsidies-and adopting a common definition of “qualifying child”-could make taxes much simpler for low-income households.

    Education Subsidies Choosing among the alternative tax subsidies for college education requires college algebra and a lawyer’s attention to detail. These choices could be made far simpler through consolidation into two subsidies, one focusing on saving incentives for education, and one on either deductions or credits for current educational expenditures.

    Saving Incentives Independent of employer-provided accounts, households may save in Individual Retirement Accounts (IRAs), Roth IRAs, educational IRAs, and Keogh plans. Rules concerning contribution limits and withdrawal patterns vary by program. Consolidating these options into one or two non-overlapping options with simple and broad rules on eligibility, contribution, and withdrawal rules would simplify tax planning for retirement.

    Capital Gains Capital gains will eventually be taxed at up to eleven different rates, depending on the asset, the owner’s income, when the asset was purchased, and how long it was held. It would be much simpler to replace this confusing hodge podge with an exclusion of a set fraction of capital gains from taxable income-say 50 percent-as was done prior to 1987.

  • Reduce the top tax rates and tax capital gains as ordinary income

    This was the cornerstone of the deal struck in 1986 that allowed substantial simplification of the individual income tax. It would massively reduce incentives to shelter funds and the need to engage in complex tax planning.

C. Simplify Filing and Record-Keeping

Thirty-six countries administer some sort of “return-free” tax system. Under such a system, the taxpayer or the taxpayer’s employer supplies a few information items to the tax authorities, who calculate the tax due and bill the taxpayer. Up to 52 million taxpayers (and many more if the standard deduction were significantly increased) could be placed on a return-free system with relatively minor changes in the structure of the income tax. These include filers who have income only from wages, pensions, Social Security, interest, dividends, and unemployment compensation; who do not itemize deductions or claim credits other than the EITC or the child credit; and who are in the zero or 15 -percent tax bracket (Gale and Holtzblatt 1997).

Nevertheless, the net cost savings may not be great. Over 80 percent of the affected taxpayers currently file the relatively simple 1040A and 1040EZ returns and the others file 1040s but have relatively simple returns. Taxpayers subject to a return-free system would still have to provide information to tax authorities on a regular basis. Some administrative costs would merely be shifted from taxpayers to employers, other payers, and the IRS. And if state income taxes were not similarly altered, many taxpayers would still need to calculate almost all of the information currently needed on the federal return.

Another way to reduce the costs of filing and record-keeping would be to expand the standard deduction significantly. This would curb administrative costs by reducing the number of households that itemize their deductions. It would also provide a tax cut for many low- to middle-income households. Estimates suggest that if the number of personal exemptions each household was granted were reduced by one and the standard deduction was raised by $4,000, the number of itemizers would fall by one-third, revenues would be maintained, and progressivity would be enhanced (Aaron and Gale 1996).

D. Procedural changes

Procedural changes in the tax policy process might indirectly help to simplify taxes by raising the visibility and explicit consideration of simplicity and enforcement issues. For example, the recent IRS restructuring legislation requires the IRS to report to Congress each year regarding sources of complexity in the administration of Federal taxes. The Joint Committee on Taxation (JCT) is required to prepare complexity analysis of new legislation that impacts individuals or small businesses.

Another way to increase the visibility of simplification issues is for the Treasury or a Congressional agency to release an annual list of simplification proposals. A Treasury “blue book” released in 1997 contained over 50 proposals for simplification, two of which were enacted later that year. The IRS restructuring act requires the JCT to include simplification proposals in biennial reports on the state of the Federal tax system.

E. Fund simplification experiments

Finally, a serious commitment to tax simplification could be established if Congress and the Administration would devote a small amount, say 0.5 percent, of the size of any proposed tax cut to conduct experiments and trial runs to show what type of simplification taxpayers would like and how best to establish such procedures. Given the magnitude of tax cuts recently enacted, 0.5 percent would go toward funding a very large amount of new efforts to make taxes simpler.

IV. Complexity in the current tax system

A. Compliance costs in the income tax

The complexity, or total resource costs, of the current tax system can be divided into several components: the amount of time it takes individuals and businesses to comply with the tax system, the valuation of that time, the out-of-pocket costs incurred by taxpayers, and the administrative costs borne by government.

Three surveys, conducted during the 1980s and described in table 2, provide data on the time taxpayers needed to comply with federal taxes. Slemrod and Sorum (1984) surveyed 2,000 taxpayers in Minnesota in 1983. Weighting the responses to reflect national averages, they estimated that taxpayers spent 2.1 billion hours filling out their 1982 federal and state income tax returns. Blumenthal and Slemrod (1992) repeated the survey in 1990 and found that time requirements for 1989 returns had increased to 3.0 billion hours. Unlike the earlier survey, the latter survey’s estimates include time spent arranging financial affairs to minimize taxes.

The largest survey, commissioned by the IRS and conducted by Arthur D. Little (ADL, 1988), asked 6,200 taxpayers by mail about time spent preparing 1983 federal income tax returns. ADL also surveyed 4,000 partnerships and corporations and their paid preparers. ADL used the results to develop models that could be used with readily available data to estimate compliance costs in future years. To develop the models, the time for each activity (e.g., learning about tax law) associated with each form was assumed to be a linear function of the number of items on the form, the number of words of instructions and references to the IRC and regulations, or the number of pages in the form. Based on these models, ADL estimated that taxpayers spent 1.6 billion hours on 1983 individual income tax returns and 1.8 billion hours on 1985 returns. For partnerships and corporations, the estimates were 2.7 billion hours for 1983, and 3.6 billion hours for 1985.

The IRS currently uses the ADL models to estimate the time required to complete forms and schedules. These estimates are published with the tax forms as part of the “Paperwork Reduction Act Notice.” For FY 1997, OMB (1998) estimates that taxpayers needed 5.3 billion hours to comply with the requirements of all tax forms and IRS regulations. This estimate applies to businesses and individuals, and includes all federal taxes, not just income taxes.

Several features of the ADL/IRS model are problematic, however. Most obviously, complexity can be related to many factors other than the number of lines or words on a form. When complex FIND ME

Given an estimate of the number of hours individuals and businesses spend complying with the tax system, the next component of compliance cost requires placing a value on taxpayers’ time. The surveys above did not inquire about this issue. Instead, analysts have generally imputed some measure of opportunity cost to individuals. Different methodologies result in widely varying estimates of the value of taxpayers’ time (table 3). Vaillancourt (1986) uses the taxpayer’s pre-tax wage, on the grounds that this is the cost to society. Slemrod (1996) argues that taxpayers are more likely to forgo leisure than work to complete a tax return, and so uses after-tax wages. Payne (1993) and Hall (1995, 1996) value individual and business taxpayers’ time by averaging the hourly labor costs of one of the major accounting firms and the IRS. This approach undoubtedly overstates the appropriate costs for individual taxpayers. The implicit assumption that a taxpayer and tax professional operate at the same level of efficiency when completing a tax return is doubtful, and ignores the expertise the tax professional has developed. And the vast majority of taxpayers do not face tax situations anywhere near as complicated as those seen by an accountant at a major firm or an IRS examiner.

Estimates of the total resource costs of operating the income tax vary widely (table 2). Payne (1993) estimates costs of $277 billion (1995 dollars) for 1985.9 Hall (1996) estimates costs of about $141 billion in 1995. Slemrod (1996) estimates costs of $75 billion in 1995. The differences between these estimates are driven largely by two factors: whether to use the results from the ADL business model or the business survey, and how to value the time spent by businesses and individuals. Both Payne and Hall use the results from the ADL model, which appears to overstate the relevant costs. Slemrod uses the results from the survey. Both Hall and Payne value taxpayer time at the cost of tax professionals’ time, which is problematic for reasons stated above. Slemrod values taxpayers’ time at the after-tax wage.

Given the existing data, it is possible to suggest a range of plausible estimates of the annual costs of operating the income tax. Slemrod’s $75 billion estimate provides a realistic lower bound. An upper-bound estimate of $130 billion is obtained by adjusting Hall’s estimate for the value of time (using Slemrod’s estimate of $15 an hour rather than Hall’s estimate of $39.60), and adding individuals’ out-of-pocket expenditures ($8 billion that Slemrod and Payne include) and tax administrative costs ($5 to $7 billion).

All of these estimates are based on taxpayer surveys. However, although they may provide the best available information to date, the survey results should be interpreted with caution. All of the surveys have low response rates. They do not distinguish between permanent and transitory costs. The surveys omit compliance costs imposed on taxpayers after returns are filed (except for Payne, who provides only a rough estimate of audit costs). It is unclear whether survey respondents have netted out the cost of non-tax activities, or distinguished the costs of one tax from other taxes. In addition, the surveys were undertaken in the 1980s and are now dated. Several major and minor tax bills have become law over the last 15 to 20 years. It is not evident that the IRS methodology captures these changes. Over the same period, technological change has generally worked to reduce compliance costs. For example, when the IRS initiated the first pilot of electronic filing in 1986, a handful of professional tax preparers electronically transmitted 25,000 returns. By 2000, over 35 million taxpayers filed electronically. In many cases, they filed from home by telephone or personal computer. The cost savings from electronic filing are not reflected in the compliance cost estimates. All of these considerations suggest the need for a new, comprehensive survey of taxpayer compliance costs.

B. The distribution of tax complexity

Measures of resource costs indicate the total administrative burden of taxes, but provide no information about which taxpayers bear the biggest burdens. Just as the distribution of tax payments is central to policy discussions, the distribution of the burden of tax complexity is also worth considering.

For many taxpayers, direct contact with the income tax is relatively simple. In 1998, 17 percent of taxpayers filed the 1040EZ, a very simplified version of the standard 1040 form. An additional 21 percent of taxpayers filed the 1040A. Relative to the 1040EZ, the 1040A requires more information and contains several more complicated provisions, but it is still fairly simple.

The remaining taxpayers filed the standard 1040 form. About 8 percent of taxpayers filed the 1040 but were eligible to file a 1040A or 1040EZ. An additional 6 percent did not itemize their deductions, did not claim capital gains or losses, and did not have business income (defined to include business net income or loss, rents, royalties, farm net income, farm rental income, partnerships, S-corporations, estates and trusts). The figures above show that in 1998, over half of taxpayers either filed a simplified form or filed the 1040 but did not itemize deductions, have business income or report net capital gains. Thus, for most taxpayers, filling out an income tax form is relatively straightforward.

Survey estimates support these findings. Blumenthal and Slemrod (1992) found that, while the average taxpayer reported spending 27.4 hours on filing income tax returns and related activities, 30 percent spent less than 5 hours, and 15 percent spent between 5 and 10 hours. At the high end, 11 percent spent 50-100 hours and 5 percent spent more than 100 hours. Out-of-pocket costs averaged $66 (in 1989 dollars), but 49 percent of filers had no such costs and another 17 percent had costs below $50. Slightly over 7 percent spent more than $200. Expenditures of time and money were highest among high-income and self-employed taxpayers.

Information on the use of paid preparers may provide additional evidence on how complex individuals find the system to be. In 1998, 53 percent of tax filers used paid preparers. Among those who filed the 1040, 64 percent used preparers. Even among 1040A and 1040EZ filers, 35 percent used preparers. At first glance, these figures suggest that most taxpayers do not believe they have simple tax situations. However, it is not entirely clear how to interpret the figures. Some individuals use preparers to obtain quicker refunds through electronic filing. Also, with relatively high income and often little leisure, many families pay others to clean their homes, plan their retirement nest egg, etc.; that they have turned to professional tax preparers as well may not provide any evidence about complexity.

C. Complexity and corporate taxes

The factors most likely to create high compliance costs for large corporations include depreciation rules, the measurement and taxation of international income, the corporate alternative minimum tax, and co-ordinating federal and state income taxes (Slemrod and Blumenthal, 1996). In addition, the largest firms are almost continually audited, and final resolution of corporation tax returns can stretch over several years. Nevertheless, the magnitude of compliance costs and the impact of tax complexity on firm operations is controversial.

At one end of the spectrum, company representatives have testified in Congress that it cost Mobil $10 million in 1993 to prepare its U.S. tax return, which comprised a year’s worth of work for 57 people. These costs sound astonishingly high at first glance, but closer examination suggests otherwise. In 1993, Mobil operated in over 100 countries and had world wide revenues of $65 billion and profits of about $4 billion. Mobil’s revenues exceeded the GDP of 137 countries and 22 of the states in the United States. Mobil’s self-reported costs of compliance were about 0.015 percent of revenues and 0.25 percent of profits. Viewed in this context, the burden imposed by compliance with the U.S. income tax appears relatively small.

In contrast, a recent study of the Hewlett-Packard corporation concluded that “[a] large U.S. multinational company can complete an accurate corporate tax return with the functional equivalent of three full-time tax professionals” (Seltzer 1997, p. 493). It would be interesting to know why Mobil’s return required so many more resources than Hewlett-Packard’s. To the extent that the problem lies in the tax system, it would be useful to know which features caused the problems.

D. Compliance costs and complexity in the estate tax

Estimates of the compliance cost of the estate tax vary enormously, partly because the methodologies are suspect. Munnell (1988) is cited as claiming that “the costs of complying with the estate tax laws are roughly the same magnitude as the revenue raised” (Joint Economic Committee 1998). But Munnell actually wrote that compliance costs “may well approach the revenue yield.” Even this more modest conclusion, however, is based on a number of rough calculations and more or less informed guesses, rather than hard evidence.

Munnell noted that, at the time, the American Bar Association reported that 16,000 lawyers cited trust, probate, and estate law as their area of concentration. Valuing their time at $150,000 per year on average and assuming they spend half time on estate taxes yields $1.2 billion in avoidance costs, compared to estate tax revenues of $7.7 billion in 1987. To get from $1.2 billion to close to $7.7 billion, Munnell refers to “accountants eager to gain an increasing share of the estate planning market,” financial planners and insurance agents who devote a considerable amount of their energies to minimizing estate taxes, and the efforts of the individuals themselves, and concludes that the avoidance costs “must amount to billions of dollars annually.” It is also worth noting that Munnell’s estimates are now out-of-date and that estate tax revenues have risen dramatically during the intervening period. Thus, even if compliance costs at that point were almost equal to revenues, they may not be today.

Davenport and Soled (1999) estimate tax planning costs by surveying tax professionals about average charges for typical estate planning in six different estate size classes and applying these estimates to the number of returns filed in 1996. This yields estimated costs for planning of $290 million. Using fairly ad hoc but not implausible adjustments for such factors as the number of nontaxable decedents that do tax planning and tax planning that has to be repeated when tax laws change, they estimate planning costs of $1.047 billion in 1999. They add $628 million for estate administration costs, based on taking one-half of the total lawyers’ fees and other costs reported on estate tax returns, and reducing that number by 45 percent to reflect the tax deductibility of the costs. (Note that the last reduction is inappropriate for measuring the social, rather than private, costs of the activity.) The sum of their estimates for planning and estate administration comes to $1.675 billion in 1999, or about 6.4 percent of expected receipts. They allocate another 0.6 percent of revenues for the administrative costs of IRS estate tax activities, for an estimated total cost of collection of 7.0 percent of revenues.

The Davenport-Soled (DS) estimate is more recent and more detailed than Munnell’s. Although both estimates require some arbitrary assumptions, it is difficult to see how the basic DS methodology could be re-done with an alternative set of reasonable assumptions to yield an estimate that avoidance costs are anywhere close to 100 percent of revenues.

The estimates above are based on suppliers of estate tax avoidance techniques. Another approach would be to survey the demanders of the service, the wealth owners. This approach has been employed with some success for the U.S. individual income tax (Slemrod and Sorum 1984, Blumenthal and Slemrod 1992), and the corporation income tax (Blumenthal and Slemrod 1996). As a point of comparison, based on such studies, Slemrod (1996) concludes that collection costs for the U.S. individual and corporate income tax is about 10 percent of the revenue collected.

Unfortunately, no reliable and comprehensive survey research has been carried out for the estate tax. What does exist applies only to businesses, and may be considered suspect. Astrachan and Aronoff (1995) surveyed businesses in the distribution, sale, and service of construction, mining, and forestry equipment industry, and separately surveyed businesses owned by African-Americans. Each of these are very special and small subsamples of the estate tax population, and the methodology employed is worrisome on a number of dimensions. For example, the authors include as a cost of avoidance the amount spent on insurance premiums to provide liquidity for paying the estate tax. This expense is properly thought of as pre-paying the tax liability, and to consider it as a cost in addition to the tax liability itself is surely inappropriate double counting.

Astrachan and Tutterow (1996) survey 983 family businesses in a variety of industries and find that family business owners have average expenditures of over $33,000 on accountants, attorneys, and financial planners working on estate planning issues; family members averaged about 167 hours spent on estate planning issues over the previous six years (the time frame for the dollar expenditures is not made clear). However, these estimates include life insurance fees that represent prepayment of estate tax liabilities. In addition, an unknown fraction of the costs is due to estate planning, inter alia about intergenerational succession of the business, that is unrelated to taxation. Repetti (2000), while corroborating in surveys of estate tax attorneys the broad magnitude of the Astrachan and Tutterow results, argues that a significant portion of these costs would be incurred even in the absence of estate taxes.

In sum, there is some evidence on the costs of estate planning for small businesses, but the estimates are marked by conceptual problems and disagreement about the fraction of costs due to the estate and gift tax as opposed to non-tax factors or other taxes. For the broader population, there is no informative evidence from surveys of wealth owners.

V. Simplification and the national retail sales tax

A national retail sales tax has been proposed recently by Congressmen Dan Schaefer (R-CO) and Billy Tauzin (R-LA) and by a group called Americans for Fair Taxation (AFT). The sales tax base would include almost all goods or services purchased in the United States by households for consumption purposes. The imputed value of financial intermediation services would also be taxed. To tax households’ consumption of goods and services provided by government, all federal, state, and local government outlays would be subject to federal sales tax. The tax would exempt expenditures abroad, half of foreign travel expenditures by U.S. citizens, state sales tax, college tuition (on the grounds that it is an investment), and food produced and consumed on farms (for administrative reasons).

The sales tax would provide a demogrant to each household equal to the sales tax rate times the poverty guideline, the annual income level below which a family of a given size is considered in poverty. States would collect the sales tax, and businesses and states would be reimbursed for tax collection efforts. The IRS would monitor tax collection for businesses with retail sales in numerous states.

The required tax rate in a national retail sales tax merits attention. Tax rates can be described in two ways. For example, suppose a good costs $100, not including taxes, and there is a $30 sales tax placed on the item. The “tax-exclusive” rate is 30 percent, since the tax is 30 percent of the selling price, excluding the tax. This rate is calculated as T/P, where T is the total tax payment and P is the pre-sales-tax price. The “tax-inclusive” rate would be about 23 percent, since the tax is 23 percent of the total payment, including the tax. This rate is calculated as T/(P+T). Sales taxes are typically quoted in tax-exclusive rates; this corresponds to the percentage “mark-up” at the cash register. Income taxes, however, are typi

The AFT proposal assumes a 23 percent tax-inclusive rate (30 percent tax-exclusive). The Schaefer-Tauzin proposal assumes a 15 percent tax-inclusive rate (17.6 percent tax-exclusive). The difference in rates in the two proposals is due to the different taxes slated for abolition. Both proposals would abolish taxes on individual income, corporate income and estates. The AFT would also eliminate payroll taxes, which raise considerable sums currently, while the Schaefer-Tauzin proposal would eliminate excise taxes, which raise little revenue.

The actual required rates would be much higher, however, for several reasons (Gale 1999). First, the plans stipulate that government must pay sales tax to itself on its own purchases but fail to allow for an increase in the real cost of maintaining government services. Fixing this problem alone raises the required rate in the AFT proposal to 35 percent on a tax-inclusive basis and 54 percent on a tax-exclusive basis (Gale 1999). Second, the plans do not allow for any avoidance or evasion, though it is universally acknowledged that both will occur. Third, the plans propose to tax an extremely broad measure of consumption, but political and administrative factors would very likely require a narrower base. Conservative adjustments for these factors raise the required tax-inclusive rate to 48 percent and the tax-exclusive rate to 94 percent in the AFT proposal, and 35 percent and 54 percent, respectively, in the Schaefer-Tauzin proposal (Gale 1999). Related analysis by the Joint Committee on Taxation (2000) reaches similar conclusions.

A. Sources of complexity

As a flat-rate consumption tax with a universal demogrant, the sales tax contains many of the features that generate simpler taxes. In principle at least, the simplicity gains could be impressive. Most individuals would no longer need to keep tax records, know the tax law, or file returns. The number of taxpayers who would have to file would decline significantly, and would include only those sole proprietorships, partnerships, and S- and C-corporations that made retail sales. The complexity of filing a return would decline dramatically as well.

Nevertheless, a NRST could create new areas of complexity. The demogrant is based on the HHS poverty guidelines, which rise less than proportionally with the number of family members. For example, in 1998, the poverty level was $8,050 for a single individual, plus $2,800 for each additional family member. Thus, the poverty level for a family of four was $16,450, just over twice the level for an individual. This structure will create incentives in many households for citizens to try to claim the demogrant as individuals rather than families. It is also not obvious from AFT descriptions how the demogrants would be administered, or even which agencies would be responsible for determining eligibility and monitoring taxpayers. Thus, the compliance and administrative costs of ensuring that the appropriate demogrant is paid could be significant.

Another area of potential complexity stems from tax avoidance and evasion behavior. The primary way to avoid sales taxes would be to combine business activity with personal consumption. For example, individuals may seek to register as firms, individuals may seek to purchase their own consumption goods using a business certificate, and employers might buy goods for their workers in lieu of wage compensation (GAO 1998). Ensuring that all business purchases are not taxed and all consumer purchases are taxed would require record-keeping by all businesses, even though only retailers would have to remit taxes in a pure retail sales tax. The AFT proposal deviates from a pure retail sales tax by requiring that taxes be paid on many input purchases and that vendors file explicit claims to receive rebates on their business purchases. This would raise compliance costs further.

A second source of tax avoidance and evasion concerns the importation of goods and services from abroad. Imported purchases of up to $2,000 per year per taxpayer would be exempt from the sales tax. This feature is likely to be exploited fully by many taxpayers, not because they travel abroad but because it would be very simple for firms to set up off-shore affiliates, warehouses, or mail order houses and ship goods to domestic customers. Moreover, it would be very difficult to monitor such arrangements and it seems quite likely that taxpayers could end up importing more than $2,000 per person on a tax-exempt basis. Some related evidence on the potential extent of these problems comes from the experience with state-level “use” taxes under which taxpayers voluntarily make tax payments on goods purchased in other states. Enforcement of such taxes has been “dismal at best” (Murray, 1997). The development of electronic commerce could raise many additional avoidance and evasion problems for the sales tax.

B. Compliance cost estimates

There are no rigorous estimates of the compliance and administrative costs associated with a high-rate NRST. Some evidence is available with respect to state sales taxes. Slemrod and Bakija (1996) report that administrative costs were between 0.4 and 1.0 percent of sales tax revenues in a sample of 8 states, and compliance costs were between 2.0-3.8 percent of revenues in 7 states. Hall (1996) cites a Price-Waterhouse study that found that retailers spent $4.4 billion complying with state retail sales taxes in 1990. Adjusting for increased retail sales between 1990 and 1995, he asserts an NRST with no demogrant would have administrative costs of $4.9 billion.

Unfortunately, as Slemrod and Bakija (1996) note, compliance cost estimates from state sales taxes are likely to vastly understate the analogous costs in a NRST for several reasons. First, at 4 and 6 percent, state sales tax rates are an order of magnitude lower than the required rate in a NRST. The higher rates in an NRST would encourage more taxpayers to engage in time-consuming taxpayer avoidance and evasion activities than under the existing state sales taxes, and this, in turn, would increase the required tax rate and compliance and administrative costs. Second, state sales tax bases are very different from the proposed federal base. States sales taxes typically include a significant amount of business purchases (Ring 1999). This reduces compliance costs, since distinguishing business and retail sales is costly. To avoid taxing business in a NRST may require all businesses to file returns and receive rebates, which would raise costs. State sales taxes often exclude hard-to-tax sectors. All states exempt financial services from their retail sales taxes, but the NRST would not. Third, states do not provide demogrants.

On the other hand, states often exempt from taxation goods such as food, housing, rent, and health care, for political or social reasons. This increases compliance costs relative to taxing a broader base because defining the boundaries of the exemption (for example, distinguishing “food” and “candy”) can be difficult, and record keeping requirements can be extensive. However, if a NRST required high rates, there would be massive political pressure to exempt goods like food, health care, and rent.

C. Other sales taxes

To address administrative problems and other concerns with the retail sales tax, many countries have employed value-added taxes (VATs). VATs are paid by businesses and impose taxes on all sales, including business-to-business transactions. Each business owes taxes on its sales and receives deductions or credits to account for the taxes it paid on its purchases. Controlling for administrative factors, the net economic effect of a VAT should be the same as an NRST. The key administrative advantage of a VAT over an NRST is that the existence of a paper trail can improve compliance rates. The chief disadvantage is the added compliance costs created. See Cnossen (1999) for further discussion.

Mieszkowski and Palumbo (1998) describe a “hybrid NRST” which would add the following features to a retail sales tax: (a) taxes would be due on all sales of multi-purpose goods and services used as final consumption goods or business inputs, (b) businesses would file for rebates for the taxes collected on business inputs, and (c) sales taxes would be withheld at pre-retail stages of production and distribution, such that taxes collected at one stage of production and distribution are credited at the next stage. This system would improve compliance relative to the NRST by developing a more extensive paper trail to identify suspicious returns and facilitate tax audits. However, the tax would also be more complex. A system of cross-checks and cross-reporting would be needed to limit fraud. The number of firms required to file would rise much closer to VAT levels than NRST levels. And businesses would file more frequently, perhaps on a bi-weekly or even weekly basis, in order to claim refunds. Mieszkowski and Palumbo concur with those who claim, as we do above, that the compliance experience of state sales tax is not very relevant for formulating cost estimates for a high-rate national sales tax. They note that the compliance costs of a hybrid NRST would likely be “several multiples” of the $20 billion compliance cost estimates they cite for an NRST. Note that if several equals “four,” the costs of complying with and administering this system would be as high as Slemrod’s estimated costs for the income tax.

VI. Simplification and the flat tax

Originally developed by Robert Hall and Alvin Rabushka (1983, 1995), the flat tax has been proposed in legislative form by Representative Richard Armey (R-TX) and Senator Richard Shelby (R-AL). Under the flat tax, businesses would pay taxes on the difference between gross sales (including business-to-business transactions) and the sum of wages, pension contributions, and purchases from other businesses, including the cost of materials, services, and capital goods. Individuals would pay taxes on their wages and pension disbursements, less exemptions of $21,400 for a married couple ($10,700 for single filers) and $5,000 for each dependent. Both individuals and businesses would pay the same flat tax rate, estimated by Treasury (1996) to be 20.8 percent (on a tax-inclusive basis). As with the sales tax, realistic versions of the flat tax will require higher rates. Unlike the sales tax, however, the required rate estimate for the flat tax already incorporates evasion and avoidance and does not assume that government tries to raise revenue by taxing itself. The only significant adjustments are for transition relief and the possible retention of some major deductions and credits due to political pressures. Adjusting for those factors, the required rates range between 21 percent and 32 percent (Aaron and Gale 1996).

A. Sources of complexity

As with the sales tax, the proposed flat tax would change the tax base to consumption, flatten tax rates, eliminate all deductions and credits in the tax code, and vastly simplify tax compliance. For taxpayers who were not self-employed, the individual filing requirement could probably be eliminated. For those that did have to file, the tax form could be a relatively short postcard with simple calculations. The tax would eliminate individual-level taxation of capital gains, interest and dividends and the individual AMT.

Any well-functioning business already retains records of wages, material costs and investments, so tax filing would impose little additional cost. The flat tax would eliminate the differential treatment of debt versus equity, the uniform capitalization rules, the corporate alternative minimum tax, depreciation schedules, rules regarding defining a capital good versus a current input, depletion allowances, corporate subsidies and credits, the potential to arbitrage across different accounting systems, and a host of other issues. The tax distortions currently caused by inflation would vanish.

Nevertheless, the flat tax would retain some existing sources of complexity and exacerbate others. It would also create entirely new areas of complexity, and the types of complexity it would abolish could easily creep back into the code. Some areas of the existing tax code are also common to the flat tax and would prove just as difficult as ever. These include rules regarding independent contractors versus employees, qualified dependents, tax withholding for domestic help, home office deductions, taxation of the self-employed, and non-conformity between state and federal taxes. The treatment of travel and food expenses might also cause problems. To the extent they are a cost of doing business, the expenses should be deducted in the flat tax. To the extent they are a fringe benefit, they should not. Making this determination may prove difficult. Graetz (1997) emphasizes the numerous problems in the existing system that would be retained in the flat tax.

A potentially more troubling issue is that, since the flat tax makes different distinctions than the existing system does, the flat tax will create new “pressure points,” and so could create a host of new compliance and sheltering issues. For example, under the existing income tax, a firm must pay taxes on interest income as well as income from sales of goods. In the business portion of the flat tax, receipts from sales of goods and services are taxable, but interest income is not. This creates an important incentive in transactions between businesses subject to the flat tax and entities not subject to the business tax (households, governments, non-profits, and foreigners): the business would like to label as much cash inflow as possible as “interest income.” The other party (not subject to the business tax component of the flat tax) is indifferent to such labeling. The same possibility occurs for cash outflows from businesses. Outflows that are labeled purchases of goods and services or capital investments are deductible, while outflows that are labeled interest payments are not deductible. This creates obvious incentives for businesses to label as “purchases” as much of their cash outflow as possible, and possibly seriously erode the tax base and tax revenues. Thus, while it equates the tax treatment of debt and equity flows, the flat tax creates a new wedge between inflows labeled “sales” and those labeled “interest,” and a new wedge between outflows labeled “purchases” and those labeled “interest expense.” Concerns that these wedges would be easily manipulated led McLure and Zodrow (1996) to conclude that the business tax “contains unacceptable opportunities for abuse.”

Another new area of complexity concerns wages, fringe benefits, and current operating expenses. Under the current system, all are deductible to firms. Under the flat tax, however, fringe benefits are not deductible. Gruber and Poterba (1996) speculate that this wedge could bring back the “company doctor.” In the flat tax, a firm’s contribution for health insurance would not be deductible, but its payment for in-house doctors, nurses, and medical equipment would be deductible.

Some flat tax rules will exacerbate existing tax complexities. The sheltering of personal consumption, especially durable goods, through business would become more important due to the more generous deduction for expensing. Conversion of business property to individual use ought to generate taxable income for the business, but would be hard to monitor.

The tax treatment of mixed business and personal use raises a number of issues. A family who rents rooms in its home or has a vacation home must currently follow fairly complex rules for allocating expenses and depreciation between personal and rental use. The flat tax is based on cash flow, however, so it is not clear how such items would be handled. Suppose a taxpayer bought a home in year 1 and in year 5 decided to begin renting a room in the house. What deduction for the cost of the capital good would the homeowner be able to take? The answer should not be “none” since depreciation is a legitimate business expense. Nor should the answer be “expensing” since the flat tax is based on cash flow and the house did not become a business property until year 5, during which there was no house purchase. But any other answer will lead to a potentially complicated new set of rules (or the same rules that currently exist). Also, if a deduction were allowed, then the decision to stop renting the room or the vacation home after a period of time would implicitly convert a business asset to personal use and so should be taxed at the business level under the flat tax (Feld 1995).

Current law imposes limits on how taxes or losses may be allocated among different taxpayers. These provisions regarding consolidated returns, S-corporations, and partnerships stop firms from merging solely for tax purposes. They appear to have no counterpart in the proposed flat tax. However, as Feld (1995, p. 610) notes: “the logical conclusion of unregulated allocation of deductions would allow free transferability of losses. Historically, however, the outcry against the opportunity by wealthy

Taxpayers may also create pressure to find ways to transfer income between wages and business income. Under the flat tax, business and wage incomes are recorded on separate forms. Thus, a business loss may be carried forward, but-unlike in the current system-it can not be used to offset current wage income.

The flat tax would create several incentives regarding cross-border flows. Firms would have incentives to engage in transactions that shifted interest expense offshore and interest income into the United States. Transfer pricing would probably work to encourage firms to locate more profits in the United States, since the tax rate would be lower here than in most other industrialized countries. Both of these issues would be easy to exploit and would drain revenues from foreign countries. Some sort of retaliation, adjustment or treaty negotiation might be expected, which would then require changes in the tax treatment of international flows under the flat tax (Hines 1996).

Feld (1995) highlights a variety of additional concerns with the business tax, including the role of in-kind transfers to a corporation, the definition of a business input (and the possible need to exempt passive assets from the definition), and possibly complex rules for hedging transactions to distinguish those that are part of the business from those that are investments by the individual.

Despite its apparent simplicity, the individual tax also creates some potential areas of complexity (Feld, 1995). First, the flat tax would effectively renegotiate every alimony agreement in the country. Under the flat tax and other reform proposals, alimony payments would no longer be deductible and alimony receipts would no longer be taxable. Second, suppose that Victim earns money and then Robber takes it away. Under the flat tax, Victim is still liable for taxes, and Robber is not. Under the income tax, it is the other way around. Third, prize money won by contestants would be deductible by the sponsoring organization as an expense, but does not appear to be taxable as wages. Addressing any of these problems would make the flat tax more complex.

Lastly, a new system will inevitably create unintended loopholes that will need to be addressed via corrective tax measures. It would be a mistake to underestimate the creative ingenuity of America’s accountants, attorneys, and tax planners.

To be clear, all of the concerns noted above could be resolved by writing carefully detailed rules covering each contingency. But of course that is what the current system already does. There is little reason to believe that the ultimate resolution of most of these issues will be simpler under the flat tax than in the current system. Feld (1995) concludes that to avoid losing revenues, the flat tax will either generate complicated business transactions (to skirt the simple rules) or complicated tax laws (to reduce the gaming possibilities), or both. This conclusion seems quite reasonable to us.

All of the discussion above focuses on the pure flat tax. However, if the flat tax were implemented, “[w]e should expect near unanimity that it will be necessary to provide transition relief” (Pearlman, 1996). Zodrow (2000) concurs that some transitional relief is “virtually inevitable.” Pearlman and Zodrow discuss the various types of transition relief that could be provided, including compensating firms for lost depreciation deductions and carry forwards of AMT credits, net operating losses and foreign tax credits. The treatment of interest deductions will also require attention. More generally, because taxes are embedded in the fabric of existing legal contracts, transitioning to a new tax system could potentially affect numerous aspects of agreements in other areas. The effect on alimony, noted above, is one such example. Pearlman concludes (p. 419) that “inevitably, any approach [to transition relief] will make the new law more complex for a long time.”

Another potential source of complexity is the reintroduction of social policy into the tax code. The pure flat tax would be devoid of all social policy initiatives. Thus, the flat tax would not only change tax policy, but also reduce the generosity of subsidies toward housing, the charitable sector, family and children, education, health insurance, state and local governments, etc. For each existing deduction and credit, however, a political case would be made that the subsidy should be retained. To the extent that social policy creeps back into the flat tax, there will be added complexity. Notably, because the flat tax has an individual component-whereas the sales tax, for example, does not-social policy in the flat tax can be tailored to individual circumstances. However, credits for children, child care, and education all raise issues of eligibility, compliance and phase-outs. Retention of popular deductions would require additional record-keeping, reporting and monitoring. Retention of the mortgage interest deduction, in a system that does not tax interest income, could create arbitrage opportunities and added record-keeping costs. Corporate subsidies for research, environmental clean-up, and other goals could easily wend their way back into the business tax. And to the extent that the demand for any of these programs remained and the tax system was able to remain clean, there is a possibility that the programs would return as spending or regulatory initiatives.A third source of complexity in a modified flat tax concerns the real and perceived distributional effects. Families in the very highest income or consumption strata would see tax burdens fall dramatically (Gale, Houser, and Scholz 1996). The flat tax would make poor families worse off, because it would eliminate the earned income tax credit, but the increased burdens on the poor would not be as large as the reduced burdens on high-income families. The difference would be made up by increased taxes on middle-class families (Dunbar and Pogue 1998; Gale, Houser and Scholz 1996; Gentry and Hubbard 1997; Mieszkowski and Palumbo, 2000).

It seems unlikely that these distributional effects will pass political muster. Retaining the earned income tax credit would reduce much of the distributional loss of the pure flat tax (Gale, Houser, and Scholz 1996), but would raise compliance costs. Moving to a Bradford-style X-tax (which would use the flat tax base, but has graduated tax rates on wages and sets the business tax rate at the highest wage tax rate) would provide more progressivity, but would also create administrative and compliance problems. It would significantly increase the revenue loss from transition relief. This would require higher tax rates on the remaining tax base. It would re-introduce taxpayer incentives and attempts to redistribute income across people or over time to exploit tax rate differentials. By raising tax rates at the high end of the income distribution, it would increase political pressure to restore popular itemized deductions.

A number of issues regarding what economists might describe as perceptions of fairness also arise. For example, there will be an inexorable tendency to compare the flat tax to an income tax because both are collected from individuals and businesses. Despite the fact that taxes on capital income will be collected at the business level, the non-taxation of capital income at the individual level may upset citizens who are used to seeing people remit taxes directly to the government on the capital income they receive.

Several perception issues arise in the business tax. Unlike the current corporate or individual business taxes, the business tax in the flat tax does not attempt to tax profits. Changing the entire logic and structure of business taxation will create several situations that will be perceived as problems by taxpayers and firms, even if they make perfect sense within the overall logic of the flat tax.

First, some businesses would face massive increases in their tax liabilities. For example, Hall and Rabushka (1995) note that General Motors’ tax liability would have risen from $110 million in 1993 under the current system to $2.7 billion under a 19 percent flat tax. Despite economists’ view that individuals-not businesses—bear the burden of taxes, there will likely be massive resistance at the business level to such changes. Businesses who oppose such change will demand reductions in the tax base or other types of relief.

Second, some businesses with large profits will pay no taxes. Profit (before federal taxes) is equal to revenue from sales and other sources less deductions for depreciation, interest payments, materials, wages, fringe benefits, payroll taxes, and state and local income and property taxes. The tax base in the business tax, however, is equal to revenue from sales minus materials, wages, pension contributions, and new investment. Thus, if a firm had large amounts of revenue from financial assets (i.e., not from sales of goods and services), it could owe no taxes or even negative taxes under the flat tax even though it reported huge profits to shareholders. This situation is consistent within the context of the flat tax. But in the past, precisely this situation led to the strengthening of the corporate and individual alternative minimum taxes, which are universally regarded as one of the most complex areas of the tax code. It is hard to see why those same pressures would not arise in the flat tax.

The third issue is the flip side of the second: some firms with low or negative profits may be forced to make very large tax payments. For example, a firm with substantial amounts of interest expense, fringe benefits, payroll taxes, and state and local income and property taxes could report negative profits, but since these items are not deductible in the flat tax, the firm could still face stiff tax liabilities. Again, this makes sense within the context of the flat tax, but will not be viewed as fair by firm owners who wonder why they have to pay taxes in years when they lose money and who will push for reforms. Misunderstanding of this point could be very important. For example, the Wall Street Journal editorial board (February 5, 1997), a strong supporter of the flat tax, nevertheless complains about a German tax that can force companies to pay taxes “even when they are losing money.” The flat tax, however, would do exactly the same thing for some firms. This will lead to efforts by businesses to retain currently existing deductions for health insurance, payroll taxes and state and local income and property taxes. Taken together, these deductions would cut the business tax base by more than half.

B. Compliance cost estimates

Slemrod (1996) and Hall (1996) have attempted to quantify the compliance costs of the pure flat tax. Both authors’ estimates ignore transition issues and the potential reemergence of social policy. Using the ADL model for taxpayer hours described above and valuing taxpayer time at $39.60 per hour, Hall estimates that the costs of record-keeping, learning about the tax law, form preparation, and packaging/sending would equal $8.4 billion. The projected 93 million individual returns are estimated to take an average of one hour and eight minutes. The projected 24.4 million business returns are estimated to take an average of three hours and 24 minutes to complete

Hall’s estimates seem both significantly too large in some respects and significantly too small in others. For example, valuing individuals’ time at $15 per hour and business time at $25 per hour, as Slemrod does, would reduce the estimate by about half. On the other hand, some of the time estimates seem implausibly low, and possibly off by orders of magnitude. Individual taxpayers are estimated to spend an average of 2.4 minutes per year doing record keeping for tax purposes. Businesses are estimated to spend only 2.3 hours per year on record keeping for tax purposes. Remarkably, especially in light of the discussion above on possible areas of complexity, businesses are estimated to spend an average of only 18 minutes learning about the tax law, and 24 minutes gathering all the relevant documents and preparing the return. In addition, Hall’s estimate leaves out many components of compliance costs, such as tax planning and auditing.

Slemrod concludes (1996, p. 375) that “it is impossible to confidently forecast the collection cost of the business part of the flat tax on the basis of observable systems, because none exists.” Instead, he offers an educated guess that the flat tax would cut business compliance costs (which were $17 billion in the individual income tax and $20 billion in the corporate tax) by one-third, and cut individual filing costs by 70 percent (from $33 billion to $10 billion), for total compliance costs of about $35 billion. This is $35 billion less than his compliance cost estimate for the income tax, or about 0.5 percent of GDP in 1995.

VII. Conclusions

As a purely technical matter, tax complexity and tax evasion can be reduced, and tax administration can be made more just and efficient. As a political and policy matter, however, making these improvements has proven quite difficult. Efforts to simplify the tax system typically run up against conflict with other tax policy goals, political factors, taxpayers’ efforts to avoid and evade taxes, and revenue requirements. Each of these factors tends to shape the base, credits, deductions, rate structure and administrative aspects of the tax system in ways that raise complexity. Efforts to reduce evasion sometimes run into similar problems.

To the extent that simplicity is a goal of tax reform, many improvements could be made within the existing system. Pure versions of both the national retail sales tax and the flat tax could be vastly simpler than even an improved income tax. But realistic versions of the flat tax and especially the sales tax would require tax rates much higher than advertised by their proponents. These higher rates complicate tax compliance and enforcement. The sales tax would face potentially serious problems with enforceability and political pressure for exemptions. The flat tax would face the same political pressures, and while enforceability is not a major issue, the tax would likely become significantly more complex than currently proposed.

Thus, simplification is an important goal of tax reform, but lasting and significant simplification may prove difficult to establish. Policy makers and voters should, therefore, weigh the costs and benefits of simplification against the other goals of tax policy.

# # # # #

This paper is a slightly revised version of testimony presented in the Congress on July 17, 2001. My analysis of tax simplification has been influenced by discussions and collaborative research with Len Burman, Janet Holtzblatt and Joel Slemrod. I thank Samara Potter for assistance. The views expressed are the author’s and should not be ascribed to other researchers, or to the trustees, officers, or staff of the Brookings Institution.

Table 1

The Economic Growth and Tax Relief Reconciliation Act:

Effective Dates and Revenue Costs of Selected Provisions



Highest Annual Tax Cut*

2001-2011 Tax Cut*

Phase-in Begins

Phase-in Complete

Phased Out By


Reduce marginal income tax rates







Abolish estate tax







Create 10 percent bracket







Double child credit







Marriage penalty







Repeal restrictions on itemized deductions and personal exemptions







Pension and IRA provisions




Nonrefundable credit






Roth 401(k)s







AMT Relief







Deduction for education expenses







All provisions






*Billions of dollars.

Source: Joint Economic Committee on Taxation JCX-50-01. “Summary of Provisions Contained in the Conference Agreement for H.R. 1836, the Economic Growth and Tax Relief Reconciliation Act of 2001”. May 26, 2001.

Table 2

Surveys of Individual Taxpayer Time


Arthur D. Little

Slemrod and Sorum

Blumenthal and Slemrod

Data Source

National random survey of 6,200 individuals1

Random survey of 2,000 Minnesota residents

Random survey of 2,000 Minnesota households

Response Rate




Sample Size2




Types of Returns

1983 federal income

1982 federal and state income

1989 federal and state income

Hours Per Activity (in billions)









Time with Preparer




Preparing Return








Rearranging Financial Affairs




Total Hours




Value of Time


$10.65/hour in

$13.69/hour in


$10.09 in

Out-of-Pocket Costs


$44/return in

$57/return in
1989$ 3


$66/return in

Total Costs for Individuals


$26.7 billion in

$34.1 billion in


$37.6 billion in 1989$

$34.8 billion w/ same activities as 1982

to Survey
(if any)

Survey results were used to obtain models for estimating taxpayer burden. Estimates above from models.

Weighted nationally.

Accounting for biases in estimates, authors suggest estimates could be as low as $17 billion.

Weighted nationally.


Arthur D. Little also surveyed 4,000 corporations and partnerships, with a response rate of 36.8 percent. Businesses found to spend 1.6 billion hours on recordkeeping, 0.1 billion hours on learning, 0.1 billion hours on obtaining materials, 0.1 billion hours on finding and using a preparer, 0.7 billion hours on preparing the return, and 0.1 billion hours on sending. Total business time: 2.7 billion hours.


The sample size was reduced by incomplete or inconsistent responses, as well as nonrespondents.


Blumenthal and Slemrod report that the average out-of-pocket expenditure for 1982 taxpayers (in 1989 dollars) was $45. This appears inconsistent with the estimate shown in the Slemrod and Sorum study, which shows that the average out-of-pocket expenditure for 1982 taxpayers was $44 in 1982 dollars—which would be consistent with $56.5 in 1989 dollars.

Table 3

Estimates of Costs of Operating Income Tax System: 1995


Slemrod 1995


Payne (1985 in 1995$)


Hours Data

Blumenthal and Slemrod

OMB estimates of average hours (ADL models for 1995)

ADL models for 1985

Total Hours

2.8 billion

1.2 billion

1.8 billion


$15/hour (after-tax hourly wage)

$39.6/hour (average labor cost of IRS and Price-Waterhouse)

$40/hour (average labor cost of IRS and Arthur Andersen)

Value of Time

$42 billion

$46 billion

$73 billion


$8 billion


$8 billion

Total Costs

$50 billion

$46 billion

$81 billion


Hours Data

ADL survey in 1983

OMB estimates of average hours (ADL models for 1995)

ADL models for 1985

Total Hours

800 million

2.4 billion

3.6 billion



$39.6/hour (average labor cost of IRS and Price-Waterhouse)

$40/hour (average labor cost of IRS and Arthur Andersen)

Total Costs

$20 billion

$95 billion

$145 billion

Other Taxpayer Costs



$27 billion2 (avoidance, evasion)

$18 billion (enforcement burden)

Total Compliance Costs

$70 billion

$141 billion

$271 billion

Total Administrative


$5 billion


$6 billion

Total Operating Costs

$75 billion


$277 billion


Hall includes individual income tax returns with Schedules C and F in the business category. The other estimates include these returns in the individual category.


Payne would include $236 billion that he estimates represents the distortionary effects of the income tax. Such costs are not typically included in the operating costs of the tax system and are not included in the table.


Aaron, Henry J. and William G. Gale. 1997. Fundamental Tax Reform: Miracle or Mirage. In Setting National Priorities, edited by Robert D. Reischauer. Washington, DC: Brookings Institution Press, 235-62.

Arthur D. Little. 1988. Development of Methodology for Estimating the Taxpayer Paperwork Burden. Final Report to the Department of the Treasury, Internal Revenue Service. Washington, DC, June.

Astrachan, Joseph H. and Craig E. Aronoff. 1995. A Report on the Impact of the Federal Estate Tax: A Study of Two Industry Groups. Unpublished paper. Kennesaw State College, Family Enterprise Center.

Astrachan, Joseph H., and Roger Tutterow. 1996. The Effect of Estate Taxes on Family Business: Survey Results. Family Business Review 9(3): 303-14.

Blumenthal, Marsha and Joel Slemrod. 1992. The Compliance Cost of the U.S. Individual Income Tax System. National Tax Journal 45(2): 185-202.

Blumenthal, Marsha and Joel Slemrod. 1996. The Income Tax Compliance Cost of Big Business. Public Finance Quarterly 24(4): 411-38.

Burman, Leonard and William G. Gale. 2001. A Golden Opportunity to Simplify the Tax System. Brookings Institution Policy Brief No. 77. April.

Calegari, Michael. 1998. Flat Taxes and Effective Tax Planning. National Tax Journal 51(4): 689-713.

Cnossen, Sijbren. 1999. On the Feasibility of RST, VAT, BTT, FT, CIT. Mimeo.

Crenshaw, Albert B. 2001. Tax Planning May Be Futile, New Law Will Shift From Year to Year, Then Expire in 2011. Washington Post June 3, H2.

Davenport, Charles and Jay A. Soled. 1999. Enlivening the Death-Tax Death-Talk. Tax Notes 84(4) special report: 591-630.

Department of the Treasury. 1996. Internal Revenue Service. Federal Tax Compliance Research: Individual Income Tax Gap Estimates for 1985, 1988, and 1992. Internal Revenue Service Publication 1415, Revision 4-96.

Departure Tax. 1997. Wall Street Journal (February 5): A18.

Dunbar, Amy, and Thomas Pogue. 1998. Estimating Flat Tax Incidence and Yield: A Sensitivity Analysis. National Tax Journal 51(2): 303-324.

Feld, Alan. 1995. Living with the Flat Tax. National Tax Journal 48(4): 603-17.

Fullerton, Don and Diane Lim Rogers. 1996. Lifetime Effects of Fundamental Tax Reform. In Economic Effects of Fundamental Tax Reform, edited by Henry J. Aaron and William G. Gale. Washington, DC: Brookings Institution Press, 321-352.

Gale, William G. 1997. Tax Reform is Dead, Long Live Tax Reform. Brookings Policy Brief No.12.

Gale, William G. 1998. Let’s Revamp the Tax Code-But How? Wall Street Journal (April 15), 1.

Gale, William G. 1999. The Required Tax Rate in a National Retail Sales Tax. National Tax Journal 52(3): 443-57.

Gale, William G. and Janet Holtzblatt. 2000. The Role of Administrative Factors in Tax Reform: Simplicity, Compliance and Administration. Mimeo.

Gale, William G. and Janet Holtzblatt. 1997. On the Possibility of a No-Return Tax System. National Tax Journal 50(3): 475-485.

Gale, William G., Scott Houser, and John Karl Scholz. 1996. Distributional Effects of Fundamental Tax Reform. In Economic Effects of Fundamental Tax Reform, edited by Henry J. Aaron and William G. Gale. Washington, DC: Brookings Institution Press, 427-431.

Gale, William G. and Joel Slemrod. Overview. In Rethinking Estate and Gift Taxation, edited by William G. Gale, James R. Hines, Jr., and Joel Slemrod. Washington, DC: Brookings Institution Press, 1-64.

General Accounting Office. 1996. Tax Administration: Alternative Filing Systems. GAO/GGD-97-6. October.

General Accounting Office. 1998. Tax Administration: Potential Impact of Alternative Taxes on Taxpayers and Administrators. GGD-98-37.

Gentry, William M., and R. Glenn Hubbard. 1997. Distributional Implications of Introducing a Broad-Based Consumption Tax. Tax Policy and the Economy 11: 1-48.

Gillis, Malcom, Peter Mieszkowski, and George R. Zodrow. 1996. Indirect Consumption Taxes: Common Issues and Differences Among the Alternative Approaches. Tax Law Review 51: 725-774.

Graetz, Michael J. 1997. The Decline (and Fall?) of the Income Tax. New York: W.W. Norton and Company.

Gruber, Jonathan, and James Poterba. 1996. Fundamental Tax Reform and Employer-Provided Health Insurance. In Economic Effects of Fundamental Tax Reform, edited by Henry J. Aaron and William G. Gale. Washington, DC: Brookings Institution Press, 125-170.

H.R. 2001. “The National Retail Sales Tax Act of 1997.”

H.R. 1040 and S. 1040. “Freedom and Fairness Restoration Act of 1997.”

Hall, Arthur. 1996. Compliance Costs of Alternative Tax Systems II. Testimony before House Ways and Means Committee. March 20.

Hall, Robert, and Alvin Rabushka. 1983. Low Tax, Simple Tax, Flat Tax. New York: McGraw-Hill.

Hall, Robert, and Alvin Rabushka. 1995. The Flat Tax, Second Edition. Stanford, CA: Hoover Institution Press.

Hines, James R. Jr. 1996. Fundamental Tax Reform in an International Setting. In Economic Effects of Fundamental Tax Reform, edited by Henry J. Aaron and William G. Gale. Washington, DC: Brookings Institution Press, 465-502.

Joint Committee on Taxation. 2001a. Estimated Revenue Effects of the President’s Fiscal Year 2002 Budget Proposal. May 4. JCX-31-01.

Joint Committee on Taxation. 2001b. Study of the Overall State and Federal Tax System and Recommendations for Simplification, Pursuant to Section 8022(3)(B) of the Internal Revenue Code of 1986. April. JCS-3-01.

Joint Committee on Taxation. 2000. Budget Neutral Tax Rate for H.R. 2525. Memorandum. April 7.

Joint Economic Committee. 1998. The Economics of the Estate Tax. December.

McLure, Charles E. Jr., and George R. Zodrow. 1996. A Hybrid Approach to the Direct Taxation of Consumption. In Frontiers of Tax Reform, edited by Michael J. Boskin, Stanford, CA: Hoover Institution Press, 70-90.

Metcalf, Gilbert E. 1997. The National Sales Tax: Who Bears the Burden? Cato Policy Analysis No. 289.

Mieszkowski, Peter, and Michael Palumbo. 1998. Is a National Retail Sales Tax Administrable?: A Proposal for a Hybrid NRST. Working Paper.

Mieszkowski, Peter and Michael Palumbo. 2000. Distributive Analysis of Fundamental Tax Reform. Mimeo.

Munnell, Alicia H. 1988. Wealth Transfer Taxation: The Relative Role for Estate and Income Taxes. New England Economic Review Nov./Dec.: 3-28.

Murray, Matthew N. 1997. Would Tax Evasion and Tax Avoidance Undermine a National Retail Sales Tax? National Tax Journal 50 (1): 167-82.

Payne, James L. 1993. Costly Returns: The Burdens of the U.S. Income Tax System. San Francisco, CA: ICS Press.

Pearlman, Ronald. 1996. Transition Issues in Moving to a Consumption Tax: A Tax Lawyer’s Perspective. In Economic Effects of Fundamental Tax Reform, edited by Henry J. Aaron and William G. Gale. Washington, DC: Brookings Institution Press, 393-434.

Pechman, Joseph A. 1990. The Future of the Income Tax. American Economic Review 80(1): 1-20.

Repetti, James R. 2000. The Case for the Estate and Gift Tax. Tax Notes March 13: 1493-1510.

Ring, Raymond J., Jr. 1999. Consumers’ Share and Producers’ Share of the General Sales Tax. National Tax Journal 52(1): 81-90.

Seltzer, David R. 1997. Federal Income Tax Compliance Costs: A Case Study of Hewlett-Packard Company. National Tax Journal 50(3): 487-93.

Slemrod Joel. 1984. Optimal Tax Simplification: Toward a Framework for Analysis. In Proceedings of the Seventy-Sixth Annual Conference on Taxation. Columbus, Ohio: National Tax Association-Tax Institute of America.

Slemrod, Joel. 1989a. Complexity, Compliance Costs, and Tax Evasion. In Taxpayer Compliance, Volume 2: Social Science Perspectives, edited by Jeffrey A. Roth and John T. Scholz. Philadelphia: University of Pennsylvania Press.

Slemrod, Joel. 1989b. The Return to Tax Simplification: An Econometric Analysis. Public Finance Quarterly 17(1): 3-27.

Slemrod, Joel. 1996. Which is the Simplest Tax System of Them All? In Economic Effects of Fundamental Tax Reform, edited by Henry J. Aaron and William G. Gale. Washington, DC: Brookings Institution Press, 335-391.

Slemrod, Joel and Jon Bakija. 1996. Taxing Ourselves: A Citizens Guide to the Great Debate Over Tax Reform. Cambridge, MA: MIT Press.

Slemrod, Joel and Marsha Blumenthal. 1996. The Income Tax Compliance Cost of Big Business. Public Finance Quarterly 24(4): 411-438.

Slemrod, Joel and Nikki Sorum. 1984. The Compliance Cost of the U.S. Individual Income Tax System. National Tax Journal 37(4): 461-74.

Slemrod, Joel and Shlomo Yitzhaki. 2000. Tax Avoidance, Evasion, and Administration. NBER Working Paper No. 7473. Cambridge, MA: National Bureau of Economic Research.

Steuerle, C. Eugene. 1997. Paying Taxpayers to File Electronic Returns. Tax Notes October 27: 477-478.

Office of Management and Budget. 1998. Draft Report to Congress on the Costs and Benefits of Federal Regulations. Federal Register 63(158): 44034-44099.

Vaillancourt, Francois. 1986. The Administrative and Compliance Costs of the Personal Income Tax and Payroll Tax System in Canada. Toronto: Canadian Tax Foundation.

Weisbach, David A. 1999. Implementing the Flat Tax. Mimeo. University of Chicago Law School, July 20.

Zodrow, George R. 1999. The Sales Tax, the VAT, and Taxes in Between-Or, Is the Only Good NRST the ‘VAT in Drag?’. National Tax Journal 52(3): 429-42.

Zodrow, George R. 2000. Transitional Issues in the Implementation of a Flat Tax or a National Retail Sales Tax. Mimeo.