FROM ISSUE NUMBER 8 ~ SUMMER 2011 GO TO TABLE OF CONTENTS

TEXT SIZE A A A

Spending in Disguise

DONALD B. MARRON

With the United States on track for a third year of trillion-dollar deficits, public debate is now focused on getting America's fiscal house in order. The challenge is straightforward: The federal government spends much more than it collects in tax revenue each year and will continue to do so even after the economy recovers.

The argument over how to close that gap is often dominated — sometimes debilitated — by sharp disagreements about how much should come from spending cuts and how much from tax increases. But that division can be misleading. A great deal of government spending is hidden in the federal tax code in the form of deductions, credits, and other preferences — preferences that seem like they let taxpayers keep their own money, but are actually spending in disguise. Those preferences complicate the code and often needlessly distort the decisions of businesses and families. The magnitude of these preferences raises the possibility of a dramatic reform of the tax code — making it simpler, fairer, and more pro-growth — that would amount to simultaneously cutting spending and increasing government revenue, without raising tax rates.

Such a reform would not eliminate the need for serious spending cuts, of course, nor would it take tax increases off the table. But it could dramatically improve the government's fiscal outlook and make the task of budget negotiators far easier. It will only be possible, however, if we clearly understand how spending is hidden in the tax code and what reformers might do about it — if we see that tax policy and spending policy are not always as distinct as we might think.

To illustrate, consider a dilemma that President Obama faced in constructing his 2012 budget. Because of inflation's ups and downs, Social Security beneficiaries did not receive a cost-of-living increase in 2011, just as they did not receive one in 2010. Nor did many benefit from the payroll-tax holiday enacted as a form of stimulus at the end of last year (since few retirees are still working and paying payroll taxes). For these reasons, President Obama wanted to make a special, one-time payment of $250 to each American receiving Social Security.

But that idea raised a question of basic fairness. Some retired government workers do not participate in Social Security. But they are just as retired — and arguably just as deserving of an additional benefit — as Social Security recipients. So the president wanted to give them $250, too. But how could he get the money to the retired government workers? The government doesn't have a master list of retirees outside of Social Security. Even if it did, it would make no sense to have Social Security send checks to retirees who had never participated in the program to begin with. Another agency had to deliver the benefit — one that was already in contact with most retirees, could enforce eligibility requirements, and, most important, could deliver the money. Only one agency fit the bill: the Internal Revenue Service. The president thus structured his special, one-time payment as a $250 refundable tax credit for any retiree who did not qualify for Social Security. In Beltway parlance, he offered these men and women a tax cut.

But was it really a tax cut? The president's $250 credit would have the same budgetary, economic, and distributional effects as his $250 boost in Social Security benefits. Both would deliver extra money to retirees, and both would finance those payments by adding to America's growing debt. One benefit would arrive as a Social Security check, the other as a reduced tax payment or a refund. These superficial differences aside, however, the proposed tax credit would be, in effect, a spending increase.

This is just a small example of a widespread practice that involves hundreds of billions of dollars each year. Because tax cuts often sound more appealing to policymakers and voters than spending increases — especially in today's political climate — the temptation to spend through the tax code is enormous. And the confusion surrounding such spending allows politicians to claim they are saving taxpayers' money when, in fact, they are really spending it.

With America deep in the red, this point is particularly important to keep in mind when considering proposals to reduce the federal deficit. It is neither feasible nor desirable to reduce deficits with tax increases alone. But revenues must be part of the conversation — even among lawmakers who loathe the very idea of "tax hikes." With our aging population and rising health-care costs, America will not be able to restrain spending enough to avoid increasing federal revenue above historical levels.

But these combined pressures do not mean that income tax rates have to go up. By properly taking account of spending disguised as tax breaks, policymakers can raise new revenue — and potentially even pay for some tax-rate reductions — by cutting back on the many spending-like provisions in our tax code. After all, that hidden spending should get the same scrutiny — and inspire the same enthusiasm for cuts — as the spending on entitlements, domestic programs, and defense that is targeted by today's fiscal hawks.

A SEA OF TAX PREFERENCES

Identifying all of the spending programs hidden in the tax code is no easy matter. The code is notoriously complex, and distinguishing between tax provisions and spending-like provisions can involve as much art and philosophy as it does science and accounting.

The best place to begin is the list of tax preferences that the Treasury Department compiles each year for the president's budget. This year, that list identifies more than 170 distinct preferences in the individual and corporate income taxes. These preferences fall into five categories.

First, credits reduce a taxpayer's liability dollar for dollar. If a taxpayer's total liability is low enough, and a credit is refundable, it can even result in a direct payment from the government to the taxpayer. The two largest credits are the Earned Income Tax Credit for low-income, working Americans (which provided its recipients with about $60 billion in 2010) and the child tax credit (which gave $48 billion to low- and moderate-income families). The most significant business credits include those for research and experimentation and for developing low-income housing, each of which amounted to almost $6 billion last year.

Second, deductions reduce the amount of income subject to tax. In the personal income tax, the most important deductions include those for mortgage interest ($79 billion in tax savings in 2010), charitable giving ($42 billion), and state and local income taxes ($27 billion). On the business side, the largest deductions are for accelerated depreciation (the ability to write off investment costs faster than capital actually depreciates), which amounted to a savings of $40 billion, and the deduction for manufacturing activities that take place in the United States ($13 billion).

Third, deferrals allow taxpayers to postpone the date at which income gets taxed. Individual taxpayers get deferrals through various tax-advantaged saving programs; 401(k)s, for example, allowed taxpayers to keep about $52 billion they otherwise would have paid to the IRS in 2010. Corporate taxpayers can defer taxes on most income earned by their foreign affiliates until the income is formally paid to the U.S. parent company, a delay that saved them about $38 billion last year.

Fourth, exclusions and exemptions allow certain types of income to avoid taxation entirely. The three largest are the exclusion for employer-provided health insurance ($260 billion in 2010), the exclusion for the imputed rental value of owner-occupied homes ($41 billion), and the exemption for interest earned by holders of state and local bonds ($30 billion).

Finally, preferential rates tax certain types of income at lower levels. The most important are the lower personal rates on long-term capital gains ($36 billion in 2010) and qualified dividends ($31 billion).

The estimated revenue losses from these five kinds of preferences total more than $1 trillion annually, almost as much as we collect from individual and corporate income taxes combined, and almost as much as we spend on discretionary programs. In 2010, for example, individual income-tax preferences totaled more than $900 billion in foregone revenue and corporate income-tax preferences more than $100 billion. When one factors in the money disbursed to individuals and corporations through refundable credits, these provisions boost spending by another $100 billion.

That's big money, but these figures come with two important caveats. First, eliminating all these tax preferences would not boost revenues as much as the numbers suggest. For instance, these calculations do not reflect how taxpayers might change their behavior if these preferences were eliminated. Treasury's estimates thus overstate how much money we could raise by eliminating preferences in the tax code. Still, even allowing for these adjustments, the sum of money at stake amounts to many hundreds of billions of dollars each year — and comes to trillions over the next decade.

The second caveat relates to how exactly the Treasury decides which provisions count as tax preferences. Identifying preferences inevitably invites controversy, because it requires a benchmark notion of an idealized tax system against which any deviations are deemed preferences. Perhaps not surprisingly, tax experts differ on what kind of system represents the ideal benchmark. The Treasury, for instance, uses a comprehensive, progressive income tax as its benchmark, with a few adjustments to reflect the practical realities of administering the tax system. Other analysts believe a broad-based consumption tax would be a better benchmark. In that case, several important preferences — including accelerated depreciation, lower rates on capital gains and dividends, and some retirement provisions — would be much smaller, or would not be identified as tax preferences at all. Meanwhile, other provisions, most notably companies' ability to deduct interest expenses, would be identified as preferences.

Although this disagreement reflects a fundamental debate about tax policy, it does not undermine the basic fact that tax preferences are enormous. Indeed, most provisions that are preferences relative to an income-tax-based system are also preferences relative to a system built around a consumption tax. My colleague Eric Toder and I have estimated, for example, that about two-thirds of the dollar value of tax preferences identified by Treasury for 2007 would also be foregone revenue under a consumption tax.

With budget pressures continuing to build, it is no surprise that this large pot of potential revenue has caught the eyes of policymakers. Some tax reformers, like senators Ron Wyden of Oregon and Dan Coats of Indiana, have recommended substantial cuts in tax preferences in order to finance broader reform of the tax code, including reductions in tax rates. President Obama's fiscal commission — as well as the debt-reduction task force chaired by former New Mexico senator Pete Domenici and former White House budget director Alice Rivlin (a committee on which I served) — proposed even deeper cuts, with the aim of financing significant deficit reduction and across-the-board reductions in individual and corporate tax rates.

This focus on tax preferences is a healthy development. Unfortunately, it has been accompanied by rhetorical flourishes that sometimes obfuscate America's real policy challenges. Tax reformers and deficit hawks often refer to tax preferences as loopholes or special-interest provisions. The president's fiscal commission even called them "tax earmarks." Those epithets make for good, quotable copy, and occasionally they even ring true. There is one tax provision, for example, that has as its sole purpose lowering taxes on NASCAR venues. That's certainly heading into earmark territory.

But the real money is not in earmarks, loopholes, or special-interest provisions. It's in the tax preferences that benefit large numbers of Americans throughout the income distribution and that play an important role in the lives of many voters. The exclusion for employer-provided health insurance and the deduction for mortgage interest, for instance, benefit tens of millions of households each year. The Earned Income Tax Credit and the various child-related credits are the two largest federal programs providing financial support to low-income families. Preferential rates on long-term capital gains and dividends reduce the tax burden on millions of individual investors. Tax preferences for research and experimentation reward firms for innovation.

It is therefore important for policymakers to recognize that tax preferences are not merely "loopholes" exploited by narrow interest groups or "earmarks" that favor some congressmen's pet constituencies. Tax preferences are social safety-net programs. They are middle- and upper-income entitlements. They are preferences for capital income. And they are incentives for activities — such as owning a home, saving for college, or investing in new research — that many believe enhance our society. Given these realities, we should not be lulled into believing that cutting tax preferences will be as painless as closing a few loopholes. Such cuts will be as politically painful as cutting popular spending programs.

ARE PREFERENCES REALLY SPENDING?

How did our tax code become so stubbornly riddled with preferences? Though tax preferences are widespread and have grown rapidly in recent years, they are not a new phenomenon. Indeed, Treasury officials began tracking them in the late 1960s. But rather than brand them as "tax preferences" or "tax breaks," they called them "tax expenditures" — a label intended to emphasize the similarity to spending programs. Congress then wrote that term into law in the landmark Congressional Budget Act of 1974 — legislation that established the rules still governing the federal budget process today.

The rationale for viewing the preferences as expenditures, rather than mere tax breaks, was (and is) that their budgetary, economic, and distributional effects are often indistinguishable from those of spending programs. Consider, for example, the tax-exempt bonds issued by state and local governments. Typically, these bonds pay much lower interest rates than taxable bonds of comparable risk. Because investors focus on after-tax returns, they are willing to accept the lower interest payments in return for the exemption from federal taxes. In the end, many investors come out ahead.

At first glance, this system might appear to offer a major tax break to investors. In reality, however, most of the benefit flows to state and local governments, which pay less in interest on their debts than they would otherwise have to in order to compete with higher-yield taxable bonds. Investors in municipal bonds pay an implicit tax by accepting lower returns, and the state and local governments receive an implicit subsidy. The fiscal effects would be the same if the tax exemption were eliminated, investors paid taxes on their interest, and the revenues were then disbursed to fund state and local projects. But because the money would pass through federal hands, we would call that taxing and spending, not a tax break.

The same is true of the exclusion for employer-provided health insurance, the deduction for charitable contributions, the corporate credit for U.S. manufacturing activities, and many other preferences. In each case, the government could accomplish the same goal — perhaps more cheaply — through an explicit cash subsidy, but has chosen instead to structure the spending as a tax break.

Princeton economist David Bradford once offered a simple thought experiment to illustrate how far such games could go. Suppose that policymakers wanted to slash defense procurement and reduce taxes, but did not want to undermine America's national security. They could square that circle by offering defense firms a refundable "weapons-supply tax credit" for producing desired weapons systems. The military would still get the weapons deemed essential to national security, defense contractors would get a tax cut, and politicians would get to boast about cutting both taxes and spending. But nothing would have changed meaningfully.

The government's ability to use such maneuvers has convinced many observers that tax preferences can be the equivalent of spending. But others continue to argue that the only tax preferences that should count as spending are the refundable credits that result in direct cash payments by the government. All other preferences are tax cuts, they insist.

The conflict between these two perspectives grabbed headlines this spring when Senator Tom Coburn, a Republican of Oklahoma, proposed ending the ethanol tax credit. The credit provides 45 cents to blenders of gasoline for each gallon of ethanol used in blending fuel. The Department of Energy could pay a direct 45-cent subsidy to blenders for each gallon of ethanol they use, but policymakers have instead chosen to run the incentive through the tax code. Senator Coburn introduced an amendment to eliminate the tax credit, arguing that doing so would cut corporate welfare and, in effect, cut spending. In his words, "[e]thanol subsidies are a spending program wrongly placed in the tax code that increases the burden of government [and] keeps tax rates artificially high."

Coburn's amendment was opposed by Americans for Tax Reform — an influential conservative group devoted to keeping taxes low. Since 1986, ATR has asked political candidates and officeholders to sign a pledge committing, among other things, to "oppose any net reduction or elimination of deductions and credits, unless matched dollar for dollar by further reducing tax rates." In the 112th Congress, 235 House members and 41 senators — mostly Republicans, but including a few Democrats — are pledge-signers; one is Coburn.

ATR contended that Coburn's amendment, which the Senate passed in June, would violate his pledge not to raise taxes. It also objected to Coburn's contention that the credit amounted to government spending. "Spending programs and tax relief are not the same thing," said ATR's tax-policy director, Ryan Ellis. He added: "If the government lets Tom Coburn keep a dollar of his own money, that is not the same thing as the government stealing a dollar from Ryan Ellis and giving it to Tom Coburn. The differences between tax relief and spending are unambiguous."

In April, similar views emerged from an unexpected source: the United States Supreme Court. In Arizona Christian School Tuition Organization v. Winn, several Arizona residents filed suit to overturn a state tax credit for charitable donations used to pay private-school tuition. Their concern was that Arizona's credit subsidized parochial schools, and thus violated the First Amendment's prohibition against any law respecting the establishment of religion. In a 5-4 decision, the Court ruled that the plaintiffs did not have standing to seek relief against the tax credit. The majority argued that, although taxpayers have long been understood to have standing to raise establishment-clause concerns about government spending, they do not have the same standing when it comes to tax incentives. Writing for the majority, Justice Anthony Kennedy concluded that the tax credit in question could not injure the plaintiff taxpayers because it did "not extract and spend [their] funds in service of an establishment." To argue otherwise, he went on, would be to assume "that income should be treated as if it were government's property even if it has not come into the tax collector's hands."

The key weakness in the arguments put forward by ATR and Justice Kennedy is that they emphasize the technicalities of budget accounting but overlook the practical effects of the tax preferences in question. As Justice Elena Kagan wrote in a dissent in the Arizona case, "Cash grants and targeted tax breaks are a means of accomplishing the same government objective — to provide financial support to select individuals or organizations"; they should therefore be judged by the same standards. Tax breaks can be viewed as a form of government spending, Kagan explained, "even assuming the diverted tax funds do not pass through the public treasury....Both deplete funds in the government's coffers by transferring money to select recipients."

Americans for Tax Reform and Justice Kennedy also misunderstand how money actually flows between taxpayers and the government. The ethanol credit, for example, may appear to allow blenders to "keep their own money," but that appearance is misleading. By levying an excise tax on the fuel that blenders produce, the government has already asserted a legal claim to a sum of money. In Ellis's language, the government has already asserted its authority to "steal" the money, but has not yet collected it. Before taking collection, the government offers to pay blenders to do something that the government wants them to do: use ethanol. If the blenders go along, they can subtract that payment when they send in their taxes. That payment does not let blenders keep their own money; rather, it is a reward for doing the government's bidding. If it were the blenders' own money, it would not come with strings.

Still, not all tax preferences are functionally equivalent to spending. One example is the lower tax rate on qualified dividends. Today, the top marginal tax rate on ordinary income is 35%, but the top rate on dividend income is only 15%. The Treasury identifies the lower dividend rate as a "tax expenditure," but it would be a mistake to confuse it with other tax expenditures that function as spending. The true purpose of the lower rate is to correct a design flaw in our current income tax, not to covertly distribute government spending.

The flaw in this instance is double taxation. If an investor buys stock in a corporation, he faces two layers of tax. The company (of which he is part owner) pays corporate income taxes on its profits, and then the investor pays personal income taxes on any of the remaining profits he receives as dividends. The two layers of tax can boost the effective tax rate on corporate income well above the rate for other sources of income. If both corporate profits and dividends were taxed at 35%, for example, the combined tax rate could reach 58%. Applying a lower personal rate to dividends is one way to soften that double taxation. With dividends taxed at 15%, the combined tax rate on corporate income paid out as dividends would be no more than 45%, much closer to the rate applied to other forms of income.

Accelerated depreciation is another example. The Treasury identifies accelerated depreciation as a tax expenditure because it allows businesses to write off their equipment investments faster than would be appropriate under a pure income tax. Under a consumption tax, however, businesses would be allowed to write off their investments immediately. Accelerated depreciation is thus a compromise between two visions of taxation, not spending hidden in the tax code.

Analysts and policymakers should thus take care to distinguish among the various preferences. Many tax expenditures are indeed spending in disguise, but not all.

BIGGER GOVERNMENT

Spending-like tax preferences not only complicate our understanding of fiscal policy, but also pose a challenge to the way we think about the size of government. This is because, in examining the scope of government, analysts usually focus on official budget measures. For example, we often hear that federal outlays averaged about 20.7% of gross domestic product over the past four decades, while revenues averaged about 18.1%. But those official budget measures do not fully account for tax breaks that effectively function as spending programs.

To get a sense of how large the federal government really is, Eric Toder and I recently added up all the spending-like tax preferences in effect in 2007, the last year before the Great Recession. Using the Treasury Department's estimates, we found that those preferences amounted to $600 billion (this is about two-thirds of the total tax expenditures that Treasury identified; the other third were preferences that are not spending in disguise). Spending-like tax preferences thus amounted to 4.1% of GDP. This means that government spending in 2007 was 23.7% of GDP when the spending-like tax preferences are included, but only 19.6% using the official budget measure. Our more inclusive measure of government spending was thus more than one-fifth larger than the usual measure.

A similar pattern holds with revenues. The official statistics indicate that federal revenues amounted to 18.5% of GDP in 2007, near their 40-year average. When we add back the 4.1% of GDP in potential revenues that were used to finance spending-like tax preferences, however, our broad measure of federal revenues increases to 22.6% of GDP.

The federal government is therefore bigger than we typically think it is. Conventional budget measures miss hundreds of billions of dollars that are implicitly collected and spent each year through spending-like tax preferences. That measurement error affects spending and revenues equally, so our measures of deficits and debts remain accurate. But the conventional measures do understate the extent to which fiscal policy redistributes income and influences economic activity.

For the same reason, conventional budget measures can misrepresent how changes in tax policy affect the real size of government. When we understand the size of government based on its spending alone, we assume that increases or reductions in revenue have no direct effect on the real size of government — all that matters is government outlays. And when we consider how much of our economy the government takes up based on the taxes it collects, we tend to assume that tax cuts make the government smaller and that tax increases make it larger.

But both approaches to assessing the size of government run into trouble when they encounter spending-like tax preferences. For example, using official budget measures, President Obama's proposed retiree tax credit (discussed earlier) would reduce tax revenues. So if we were to assess the size of government based on how much it spends, the president's proposal would have no effect on government's size; if we measured it based on how much it taxes, the proposal would actually shrink the government. But if we (correctly) recognized the retiree tax credit as a spending increase, we would conclude that the credit would actually increase government spending (making the government larger), while leaving the real scope of its tax collection unchanged.

The logic also works in reverse. Suppose policymakers decided that employer-provided health insurance should be subject to the same income and payroll taxes that apply to wages and salaries. Such a measure would increase government revenues by several trillion dollars over the next decade. But it would do so by eliminating the largest example of spending through the tax code. Under conventional measures, federal revenues would increase, but government's role in private insurance markets would actually narrow.

Advocates of smaller government are often skeptical of proposals that would increase federal revenues. But when it comes to paring back spending-like tax preferences, an increase in revenues would mean that government's role would get smaller. This, in essence, is the point Senator Coburn was trying to make with the ethanol tax credit. And it is a point that even the most ardent anti-tax, limited-government purists should keep in mind.

GIVING CREDIT

The fact that some tax preferences are actually hidden spending programs does not necessarily make them bad policy. Some tax preferences support important policy goals, just as many spending programs do. And sometimes the tax system is the most efficient way to administer specific policies. The personal income tax, for example, provides a natural mechanism for providing benefits that should vary with income, like the Earned Income Tax Credit. The corporate income tax provides a convenient administrative structure for incentives like the tax credit for research and experimentation.

The importance of labeling many tax preferences as spending is not to disparage them, but to account for them honestly. The goal is to highlight the resources that the government directs through these provisions and to encourage analysts, commentators, and policymakers to subject them to the same scrutiny they give traditional spending programs. Some tax preferences provide substantial benefits and can withstand that scrutiny even in times of fiscal tightening. Others should be left on the cutting-room floor.

In addition, many of the tax preferences that do stand up to scrutiny — or persist because of their political popularity — would benefit from serious restructuring. Today's preferences for low-income workers and families with children, for example, are painfully complex. Their byzantine rules impose unnecessary costs on beneficiaries and open the door to errors and fraud. Much better would be a system that consolidated these provisions into simple, streamlined preferences for holding a job and having children. The Domenici-Rivlin task force offered one such approach, in which all existing preferences for low-income families and children would be replaced with an earnings credit and a child credit available to all households, regardless of income.

Other major preferences could accomplish their intended goals at lower cost and with less economic distortion if they were redesigned as credits. The exclusion for employer-provided health insurance, for example, is an exceedingly inefficient way to encourage people to maintain health-care coverage. One flaw is that it offers bigger subsidies to high-income households. Because the exclusion matches the tax rate for each income bracket, the government picks up 35% of the insurance tab for an attorney earning $500,000, but only 15% of the cost for a truck driver earning $50,000. Not only is that "upside down" structure unfair, it also reduces the exclusion's efficiency in promoting health coverage — since high-income families are more likely to get health insurance without a subsidy than are lower-income families. If policymakers want to get as much health-insurance "bang" as possible for the many bucks devoted to health-insurance subsidies, it makes no sense to offer additional government assistance to people who are likely to carry insurance anyway.

Another flaw is that the tax exclusion raises the cost of health insurance for everyone. Because the exclusion is essentially open-ended (i.e., whatever amount an employer spends on a worker's health insurance is excluded from that worker's taxable income), workers have an incentive to choose expensive, high-end insurance plans that cover as much health care as possible. Such plans will usually minimize cost-sharing provisions like co-payments (which consumers usually pay for with after-tax dollars) while offsetting the costs through high premiums (which are paid for with pre-tax dollars). This system, in turn, drives up health-care costs overall — since the generous employer-provided coverage removes individual consumers' financial incentives to limit their use of health services. Absent this tax distortion, insurers would offer less expensive plans that relied more heavily on co-pays, co-insurance, and deductibles in order to both manage demand for health services and keep insurance costs low.

If policymakers wanted to retain a tax incentive for health insurance, they could correct both of these flaws by converting the current exclusion into a fixed, refundable credit. Individuals and families would qualify for the credit if they purchased health insurance that met some basic standard of coverage. Every taxpayer would have the same financial incentive, and would receive the same financial assistance, regardless of income. Every taxpayer would also be free to get more expensive insurance that exceeded the minimum standard, but the additional cost would not be offset by any extra tax subsidy. The credit would thus encourage the acquisition of basic health insurance across the income distribution, without undermining plans' ability to use co-insurance and other tools to influence the use of health services.

The same is true of the mortgage-interest deduction, another "upside down" preference that is more valuable for people in higher tax brackets. Researchers find that the mortgage deduction does little to encourage home ownership. Instead, it encourages middle- and upper-income taxpayers to buy bigger homes and take on more mortgage debt — neither of which is an important social goal. There is a good case, therefore, for simply eliminating the deduction. If policymakers want tax incentives for home ownership, they would be better off redesigning the deduction as a credit — one that would be both fairer and more effective. One option would be a fixed credit linked to home ownership, rather than to carrying a mortgage. For example, taxpayers might receive a fixed credit for each year they own a qualifying residence. Every taxpayer would face the same incentive and would receive the same assistance for purchasing a home. But no one would be encouraged to buy a larger house or to take on extra debt.

If policymakers want to use the tax code to encourage certain types of behavior, credits can often achieve the same results as exclusions and deductions, but more efficiently and at lower cost. Some observers may worry that greater reliance on credits would increase the amount of redistribution in the tax code, but changing the structure of tax rates could offset that effect in a broader tax reform. As both the president's fiscal commission and the Domenici-Rivlin task force demonstrated, eliminating spending-like preferences can allow for significant rate cuts even with significant deficit reduction.

AN ESSENTIAL REFORM

Washington's love affair with tax preferences has spawned a system that is needlessly complex, economically harmful, and often unfair. Tax breaks reach into many aspects of daily life and influence many personal choices — on matters including health care, education, charitable donations, investment, saving for retirement, owning a home, and even raising children. They represent a major exercise of government power, but face less oversight than many activities on the spending side of the budget. They conceal the true size of government, and they confer enormous power upon the tax-writing committees in Congress — which have the ability to simultaneously raise revenue and spend it inside the tax code.

The time has come for serious reform. America needs to fix its broken tax system and find additional revenue to help reduce our persistent budget deficits. The best way to achieve both aims is to take a hatchet to the thicket of spending-like tax preferences. Many preferences should simply be eliminated; those deemed to serve important policy goals should be restructured to be simpler, fairer, and more effective. Lawmakers can then use the resulting revenue to cut tax rates across the board and reduce the deficit.

Such reform is long overdue. It won't be easy, but the enormity of our budget problems may finally be enough to get liberals, moderates, and conservatives to join together to get it done.

Donald B. Marron is director of the Urban-Brookings Tax Policy Center. He has served as a member of the President's Council of Economic Advisers and as acting director of the Congressional Budget Office.