Tax Ti       tl  Statement of Joel Slemrod, Paul W. McCracken Collegiate Professor of Business Economics and Public Policy, Professor of Economics, and Director of the Office of Tax Policy Research, University of Michigan, Before the House Ways and Means Committee at the June 8, 2005, Hearing on Tax Reform

June 9, 2005

109th Congress

Testimony Submitted to the

Committee on Ways and Means

Hearing on Tax Reform

Washington , D.C. , June 8, 2005

Statement of Joel Slemrod

Paul W. McCracken Collegiate Professor of Business Economics and Public Policy, Professor of Economics, and Director of the Office of Tax Policy Research, University of Michigan

One-third of Americans say that complexity bothers them more than anything else about the tax system. This is more than twice as many who say their biggest complaint is "the large amount of taxes they paid." Half of Americans rate the system as "very complex," and someone holding that belief is 10% more likely to favor scrapping the income tax for another system. In this testimony I address the cost of collecting taxes under the current system, what causes this cost, and what kinds of tax reform promise to simplify and reduce the cost of collecting taxes.

Tax complexity has many dimensions. To suggest its magnitude, some count the number of pages or words in the tax code or forms. Others stress that even tax experts rarely agree on the true tax liability for a tax return of even moderate complexity, or note that it often takes several years to finally resolve the tax liability of a big corporation.

In my view the most informative measure of tax complexity is the resource cost of collecting taxes. This is equal to the IRS budget plus the value of the time and money spent by the taxpayers and third parties to the collection process (such as employers who withhold tax for their employees.) Measuring the IRS budget is straightforward, but measuring the other components of the collection costs --known as compliance costs --is not. What do we know?

A recent comprehensive study done by IBM Business Consulting Services under contract to the IRS concluded that, in tax year 2000, the 125.9 million individual taxpayers had a total compliance burden of 3.21 billion hours and $18.8 billion in monetary expenditures. This translates into an average burden of 25.5 hours and $149 per taxpayer. Although self-employed taxpayers represent only about 25% of all individual taxpayers, they experienced approximately 60% of the time and money burden. As a result, the average time and money burden of self-employed taxpayers (59.5 hours, $363) was substantially greater than that of those taxpayers with only wage and investment income (13.8 hours, $75). Not surprisingly, the average compliance burden was also consistently higher among taxpayers who have more complex tax returns; for example, non-self-employed taxpayers who itemize their returns spend an average of 21.3 hours and $114 on tax compliance, compared with 11.4 hours and $63 for non-itemizers. The current system is not, however, complicated for everyone, and is generally not that complicated or costly for the tens of millions of taxpayers who file Form 1040EZ or 1040A; survey evidence suggests that 30% of taxpayers spend fewer than five hours on all tax matters over an entire year. By applying a range of dollar values to each hour of time burden, IBM estimated that the annual resource cost of compliance for individual taxpayers is between $67 billion and $99 billion.

A series of analyses by the Office of Tax Policy Research at the Ross School of Business at the University of Michigan , under contract to the IRS, have examined the compliance cost of large and medium-sized businesses. In 1996 the total compliance cost for the 1,500 largest companies exceeded $2 billion, or over a million dollars per company; for Fortune 500 companies, the average cost was nearly $4 million. For businesses smaller than the biggest 1,500, but with assets in excess of $5 million, in 2002 the total compliance cost came to about $22 billion. There is clear evidence that business compliance costs are regressive --costs as a percentage of company size are higher for smaller companies than they are for larger companies. For instance, companies with between $100 million to $250 million of assets have only about seven times higher compliance costs than companies with between $5 million to $10 million of assets, even though they are between 10 and 50 times bigger.

In written testimony I submitted to this committee on June 15, 2004, I drew on these and other studies to estimate the total annual cost of collecting the federal income tax in 2004. My best estimate came to $135 billion per year. Of this total, $85 billion consists of the total compliance cost borne directly by individuals (including sole proprietorships), and another $40 billion relates to business. Adding an IRS budget of about $10 billion produces the overall collection cost estimate of $135 billion. This is 14.5% of individual and corporation income tax receipts in fiscal year 2004, and about 1.2% of 2004 GDP.

The $135 billion annual cost of complexity is the value of the resources used up in the process of collecting taxes, and the amount of resources that could --if taxes could be collected costlessly --be freed up for whatever else Americans would prefer to use their time and money for. This is an economic cost of collecting taxes that should be added to the cost incurred when the tax system discourages people from working as much as otherwise, businesses from investing as much as otherwise, and so on.

The negative consequences of tax complexity, though, go beyond what can be estimated in dollars. Complexity causes a capricious and inequitable distribution of tax burdens because it rewards those who have the means and inclination to find all the tax angles, and leaves the dutiful among us holding the bag. Moreover, the unfairness complexity causes --and the complexity itself --undermine trust in the fairness of the tax system, which may in turn undermine voluntary compliance. Complexity reduces the transparency --who bears how much burden and why --of the tax system, which I believe is inimical to a properly functioning democracy.

The rapid computerization of tax matters --in 2003 over two-thirds of self-prepared returns were done with software on a computer, compared to less than 20% in 1993 --is a double-edged sword. Although it has undoubtedly facilitated the tax collection process, I am concerned that computerization also may erode taxpayer understanding of the formula that turns the inputted items into what one owes the government. Tax return software may be helping to turn the tax system into a black box --a more efficient black box, to be sure, but a black box nevertheless.

Given a $135 billion annual resource cost, plus unquantifiable but significant negative effects on equity and even the functioning of democracy, tax simplification deserves to be taken seriously. We must, though, move beyond platitudinous support for simplification and address how, and how much, to simplify the tax system. As we begin what I hope will be a serious discussion of how we should tax ourselves, it is critically important to separately consider three important aspects of the tax system: 1) rate structure, including whether the rate structure has one rate for all or is graduated, 2) base integrity, that is whether the base is messy or clean, and 3) base choice, whether the base is income or consumption, or something in between.

Another fundamental point is that the simplest way to collect tax is not the best, and there are almost always tradeoffs between simplicity and the other criteria we use to evaluate our tax system. For example, consider the following simple tax system. Start with the annual revenue needs of about a trillion dollars and divide that by the 130 million or so individual taxpayers; this division yields about $8,000 per return. Here's a tax system: everybody owes $8,000 per year, period, including Bill Gates, a single mother earning $10,000 per year, and everyone else. Most of us, maybe all of us, would judge that very, very simple tax system to be very, very unfair.

As the example illustrates, in part tax complexity reflects a belief that simpler systems can cause an unfair distribution of the tax burden. Achieving a progressive distribution of the tax burden --one in which taxes as a fraction of income are higher for higher-income families --requires measuring, reporting and monitoring a measure of well being, such as income.

A simpler, or less costly, tax system may compromise the fairness of the tax burden in other ways. For example, loosening enforcement would save both administrative and compliance cost, but would also shift the tax burden toward those who view paying what they owe as a civic duty. Thus, the simplicity of a tax system can be assessed only with respect to a standard of enforcement and, ultimately, fairness. For this reason even the most thoroughgoing simplification imaginable that also meets our shared standards of fairness would cut the resource cost of collection by no more than in half or, at most, two-thirds.

Can a messy tax base be justified by an appeal to fairness? In some cases, yes. A deduction for involuntary medical expenses arguably improves how well taxable income measures a family's true well-being and thereby improves fairness. But most of the deductions, credits, and adjustments in our current system cannot be justified in this way. The fact is that the U.S. income tax system is an awkward mixture of a revenue-raising system plus scores of incentive and reward programs. A few of these programs may be justified, but most cannot, and a thoroughgoing pruning of these programs would, I believe, not only significantly simplify the tax system but would also be good for the economy.

Cleaning the tax base may be the hardest aspect of tax reform to achieve politically, because nearly every bell and whistle has a constituency behind it. Many observers and many taxpayers are cynical about the ability of Congress to keep its collective hands out of the cookie jar, leading some to question the wisdom of cleaning up the tax system only for it to get dirty again. The most cynical of all expect base-cleaning tax reform to happen precisely so the base exceptions can be introduced again for political rewards. I have no suggestions for how to stiffen Congressional resistance to special interest pleading, although I suspect that a thoroughgoing --rather than piecemeal --reform provides Congress with the opportunity to shrug its shoulders and say "we're doing it to everyone." In any event, if tax reform needs to be revisited every twenty years, so be it; that strikes me as a much better outcome than pursuing only policies that are likely to last forever.

Flaws in the political system also underlie an argument one hears in favor of a costly tax system --that it undermines big government. The idea is that the political system is flawed in such a way that we do not get the size of government we want --we get something bigger. Although the government cannot be downsized directly, the argument continues, it does contract when its source of funding gets more expensive. From this perspective, making the tax system more painful and more costly is a good thing. This argument (along with the separate and also controversial argument that it would make the tax burden more visible) lies behind the suggestion to abolish employer withholding and remittance of employees' income tax liability, which certainly makes the system much less costly to administer fairly. I don't buy this argument at all, because it relies on several arguable or dubious assumptions --that the political system systematically spends "too much," that spending responds more to raising the cost of funds than to direct political reform and that, if the muck-up-the-tax-system strategy is effective, the benefits of reducing spending exceed the higher costs of raising revenue. Whether one agrees with this argument or not, the important point is this: before we embark on serious tax reform we'd better get straight whether the objective is to make the tax system more, or less, costly. I recommend the latter, and so turn next to tax simplification.

One key to simplifying the tax system follows from the causes of tax complexity: minimize fine-tuning the tax liability of individuals and fine-tuning the economy by subsidizing and rewarding activities deemed to be especially valuable. Note, though, that this is illusory simplification --and probably not simplification at all --if the same set of subsidies and rewards is reconstructed outside of the tax system. If we're going to have these subsidies and rewards in any event, arguably they should be implemented through the tax system rather than by operating separate bureaucracies --by having just one financial account between the government and the people.

Simplifying the system enough along these dimensions would allow us to take advantage of large-scale final withholding by businesses through a return-free income tax system. Twenty years ago in its famous report that preceded the Tax Reform Act of 1986, the Treasury Department noted that, with significant base cleaning, as many of two-thirds of individual taxpayers need not file tax returns at all. Thus, base cleaning alone can, for the majority of taxpayers, achieve the ultimate simplification --no return at all --that is a natural feature of either a value-added tax (VAT) or a retail sales tax. This is, to be sure, a "populist" simplification, in that it would simplify the system for those (many) people whose tax matters are already relatively simple.

Structural changes to the way we now tax corporations, including multinational corporations, also have potential to simplify the tax system. Rationalizing the separate entity-level taxation of corporations (and the double taxation of dividends that accompanies it) would simplify tax matters by reducing or eliminating the need for complicated rules delineating when income passes from the corporation to the shareholder. This sort of reform, known as integration, has long been advocated by supporters of a comprehensive income tax, and is not inherently related to the choice of a consumption versus and income tax base, although it accomplished by all of the leading consumption tax proposals.

Although abandoning progressivity or thoroughly cleaning the tax base (or, for that matter, a rationalization of how we tax corporation-source income) would to my mind certainly qualify as fundamental reform, in recent years the idea of fundamental reform has become linked to basing taxation on consumption rather than income. Do we need fundamental reform in this sense to get a less complex system?

No. Replacing the income tax with a consumption base is neither necessary nor sufficient for significant tax simplification. European experience with the VAT has shown that a consumption tax is not sufficient for simplification: real-life VATs can be as costly to operate as real-life income taxes. Adopting a consumption tax is not necessary for simplification because a clean-base, largely return-free, structurally-improved income tax system can achieve a lot.

Recall that the three key aspects of tax systems --clean or messy base, progressive rate structure or not, and income or consumption base --are separate issues. A consumption tax can have a clean or a messy base, and can feature flat or graduated rates. So can an income tax. Most consumption tax proposals not only adopt a new base but also radically clean the base and sharply curtail progressivity.

By arguing that a consumption base is neither necessary nor sufficient for tax simplification, I do not mean to dismiss the potential inherent simplicity advantages of basing tax liability on consumption rather than income. All other things equal, and with some exceptions, tax systems are simpler when the base is cleaner, when the rate structure is flatter, and when the base is consumption. Although a reformed income tax can substantially simplify the system, it leaves some potential simplification opportunities on the table. Consumption taxes can dispense with the accrual accounting required for income taxation --depreciation, inventory accounting, and so on --and its inherent problems in an inflationary environment. Consumption tax systems need not measure real capital income; any system of taxing capital income is prone to inconsistencies that reward complicated transactions such as tax shelters and tax-oriented financial products. These structural problems will not go away with income tax reform.

We should not, though, fall victim to a "grass-is-greener" fallacy, for a few reasons. With more than a trillion dollars at stake, there will plenty of people looking for inconsistencies in a consumption tax system, and some will be found. Consumption taxes based on taxing retail sales are probably not administrable at our usual standards of equity and intrusiveness. Depending on how it is handled, the transition to any consumption tax system, even one that eventually would be much simpler, can be incredibly complex and riddled with loopholes that erode both revenues and fairness.

Abolishing the federal income tax would also eliminate the availability of a fairly reliable measure of the financial standing of families (as well as businesses) that is widely used throughout the economy. If individuals did not have to file income tax returns, they would still need to keep some records. But which ones? For example, what records would they have to provide mortgage lenders or college financial aid officers? Many federal transfer and other programs now rely on an annual measure of comprehensive income, for which labor income alone will not suffice. If many states continue to levy comprehensive income taxes, the compliance cost saving is reduced if taxpayers of those states still have to calculate income. To the extent that alternative ways of verifying income would arise, these new costs need to be netted out to obtain the true cost saving from moving away from an income base.

It costs us about $135 billion a year to collect the trillion dollars or so of income tax revenue. The complexity that contributes to this cost also undermines the fairness of the tax burden and erodes the transparency of the tax system. Some of the cost arises because we have high standards for the fairness of our tax system, and would remain under any system that meets those standards. Some of the cost occurs because we use the tax system for many things other than raising revenue, and could be eliminated if the tax base were cleaned of these features. Some of it occurs because of structural problems --some fixable, and some inherent --in the income tax system. Moving to a consumption tax base addresses some but not all of the sources of complexity.

The potential benefit from tax simplification is substantial, but choosing the right path to a simpler tax system requires facing up to several difficult tradeoffs. This should not be surprising, because dealing with tax complexity requires addressing the most fundamental questions about the relationship between government and the people it serves: how activist should the government be, how intrusive should it be, and when should it settle for rough justice?

Key References

Guyton, John L., John F. O'Hare, Michael P. Stavrianos and Eric J. Toder. 2003. "Estimating the Compliance Cost of the U.S. Individual Income Tax." National Tax Journal 56(3) (September): 673-88.

Slemrod, Joel. 1996. "Which is the Simplest Tax System of Them All?" in H. Aaron and W. Gale (eds.), The Economics of Fundamental Tax Reform, Washington , D.C. : The Brookings Institution, 355-91.

Slemrod, Joel. 2004. Testimony Submitted to the Committee on Ways and Means, Subcommittee on Oversight, Hearing on Tax Simplification, Washington , D.C. , June 15.

Slemrod, Joel and Jon Bakija. 2004. Taxing Ourselves: A Citizen's Guide to the Debate over Taxes. Third edition. Cambridge : MIT Press.

Joel Slemrod

Office of Tax Policy Research

Stephen M. Ross School of Business

The University of Michigan

701 Tappan Street , Rm. A2120

Ann Arbor , MI 48109-1234

 

734-936-3914

 

734-763-4032 (fax)

 

jslemrod@umich.edu