来源:Admin5 作者:Susannah Camic Tahk 时间:2016-03-09 11:38


This Article explores the relationship between how taxes distribute their costs and benefits and how likely taxes are to succeed as revenue sources over time. Public choice theory contends that how a law allocates its costs and benefits affects how well that law accomplishes its goals. This Article employs and evaluates this contention in the context of federal and state tax law.


The federal income tax base is regularly shrinking, an outcome that public choice theory would not find surprising. At present, income tax proceeds go into a pool of general revenues, whose funds benefit the large and yet diffuse and amorphous U.S. public as a whole. As a result of this arrangement, no particular beneficiary group has an incentive to prevent the tax base from shrinking or to rally behind any of the wide-ranging reforms that lawmakers have been trying in vain to implement. In contrast, the interest groups that do play a role in tax lawmaking have incentives to contract the tax base by protecting special exceptions and carve-outs.


While the cost-benefit structure of the federal income tax leads to tax base erosion and stymies reform, other tax laws, particularly state tax laws, offer an alternative. In particular, many states earmark1 specific taxes for specific programs that benefit concentrated groups. Public choice theory suggests that these earmarked taxes are more likely to remain long-term sources of revenue. The interest groups that benefit from earmarked taxes will work to preserve and to expand these taxes, guarding existing tax bases and even supporting reforms if necessary. In particular, earmarked taxes that benefit concentrated groups and impose costs on diffuse groups should be the most successful at raising revenue over time. In contrast, earmarked taxes that benefit diffuse groups and impose costs on concentrated groups should be the least successful at raising revenue over time.


This Article tests these public choice theory suggestions about the success of taxes with different cost-benefit configurations. To do this, I describe and analyze a comprehensive dataset I compiled on the recent history of approximately 1500 state-level earmarked taxes, along with a case study of the politics of a recently enacted earmarked tax. Using this recent historical evidence, this Article examines the extent to which the costs and benefits of a tax connect to its capacity to generate revenue over time. This analysis also investigates which specific types of earmarked taxes are the most productive sources of revenue. While these findings may suggest normative conclusions that I do discuss, the focus of the Article is positive.


This analysis indicates that how a tax law allocates its costs and its benefits relates to how revenue productive that law is. In particular, in this Article’s model, the earmarked taxes that have concentrated benefits and diffuse costs are associated with a substantial increase in revenue over time relative to taxes with other cost-benefit configurations. Conversely, the earmarked taxes that have diffuse benefits and concentrated costs are associated with smaller increases in revenue over time relative to other taxes. For these reasons, whether a tax has a concentrated beneficiary group to protect it may determine how well that tax performs as a revenue source.


These findings are the first to provide empirical support for a claim that has long been central to public choice theory: How laws distribute their costs and benefits influences how well these laws achieve their goals. In addition, this evidence furnishes information that state tax lawmakers may find useful in designing earmarked taxes.


This Article proceeds as follows: Part II describes tax lawmaking dynamics generally in light of public choice theory, looking at the federal tax legislative process and discussing the ways in which the mechanism of tax earmarking, often employed at the state level, offers an alternative. Part III presents and analyzes recent historical data and a case study on state-level earmarking practices and considers the implications that the findings of this analysis have for tax law design.



The federal income tax base has been eroding for decades.2 Efforts to stem this trend and reform the federal income tax have met resistance, and even successful endeavors are difficult to sustain. Based on public choice theory, I suggest that this is in large part a result of the way the federal income tax distributes its costs and benefits. Taxes that are earmarked for specific purposes, however, offer different cost-benefit configurations that I predict, again based on public choice theory, give these taxes different political fates than that of the federal income tax. The following two Sections of this Article briefly describe the legislative problems that have emerged at the federal level and explain the earmarking alternative.


A. Tax Lawmaking Dynamics

1. Federal Tax Lawmaking in Practice


Looking at the political dynamics of the federal income tax reveals a tax whose base erodes easily and a tax that is difficult to reform.3 As former Assistant Secretary of the Treasury for Tax Policy and Harvard law professor Stanley Surrey once said: “A sort of political paralysis appears to be forming . . . under which no clear way has appeared to move away from the combination of high rates tempered by many avoidance possibilities.”4 As Surrey noted, while lawmakers remain able to raise taxes across the board from time to time, they are unable to prevent the tax base from eroding. Similarly, they cannot achieve tax reform goals that would halt this trend.


Tax base erosion is not a new phenomenon. Responding to the proliferation of targeted tax preferences, presidents from both parties have advocated tax reform.5 In the 2012 presidential campaign, both  candidates ran on tax reform platforms.6 Writing during that campaign, prominent economic commentator Ezra Klein stated: “As polarized as Washington is over tax and budget issues, a base-broadening, rate-lowering tax-code overhaul has become the one policy every wonk in town can agree on.”7 Klein went on to point out that, in the two years prior to his post, two national bipartisan commissions, the Democratic chair of the Senate Finance Committee, the Republican chair of the House Ways and Means Committee, the Republican chair of the House Budget committee, as well as the Democratic president had all endorsed tax reform plans.8


As Klein’s observations suggest, the kind of tax reform around which bipartisan consensus has emerged entails broadening the income tax base by subjecting more income to taxation by pruning the Code of special preferences.9 Such pruning allows tax lawmakers to raise more revenue or to cut rates across the board. Base-broadening tax reform has widespread appeal in part because it provides benefits to millions of taxpayers at the expense of only a few. A larger tax base taxed at a lower rate can raise the same amount of revenue as a smaller tax base taxed at a higher rate. As a result, broadening the tax base allows tax rates to fall across the board.10 In contrast, when federal lawmakers add to special preference provisions to the Code, including exclusions, deductions, credits, and other targeted carve-outs, tax rates must rise across the board to keep revenue levels constant.11


Although many lawmakers have at times advocated base-broadening tax reform, in almost every instance, they have failed to achieve it. To take one of the earlier examples, President Kennedy’s administration witnessed a major tax reform collapse. The interest groups whose preferences the Kennedy administration targeted vehemently opposed the bill.12 Senator Harry Byrd, then chair of the Senate Finance Committee, informed Kennedy and his advisers that interest-group opposition would kill the bill in committee.13 Like many other Kennedy-era legislative reforms, this one was unable to overcome legislative resistance,14 and the tax base continued to crumble.


Then, in the late 1960’s, under the leadership of tax reform proponent Treasury Secretary Joseph Barr, Congress revisited the issue of tax reform. Barr instigated a public campaign to rid the Code of the loopholes that were driving up rates for the majority of American taxpayers.15 In the wake of that endeavor, Congress passed the Tax Reform Act of 196916--only to watch as “subsequent legislation reopened most of the closed loopholes.”17 Following this, special preference items continued to be enacted at an unprecedented rate.18 Even the Tax Reform Act of 197619 contained more exclusions, deductions, credits, and other targeted favors than it included base-broadening measures.20 Hoping to protect the rapidly eroding tax base, President Carter ran on a platform of tax reform, promising to “broaden . . . the base” and eliminate “tax reduction provisions.”21 As President, Carter did manage to bring a tax reform package to Congress.22 Once there, however, just as in 1976, special interest groups inserted so many preference items into the so-called “reform” bill that the legislation that eventually passed represented “a complete renunciation of the Carter tax proposals and any notion of tax reform.”23


Lawmakers have managed to break the tax reform curse only once. In 1986, President Reagan, along with several of his top economic advisers and much of the congressional leadership, achieved what observers have called a legislative “miracle,”24 the Tax Reform Act of 1986.25 In this memorable instance, key policy entrepreneurs overcame massive interest-group resistance to pass a bill that cut taxes for 80% of Americans.26 Political scientists and journalists have studied this political triumph in an effort to figure out how the political entrepreneurs involved attained the “impossible.”27 No one has offered a general explanation for how this bill became law, although many cite the rare political gifts of the involved parties and their particular commitment to tax reform.28


Regardless, most observers can agree that the 1986 Act was a remarkable tax policy feat. The bill eliminated hundreds of targeted tax preference in exchange for lowering the top marginal income rate from 50% to 28% and substantially simplifying the pre-existing rate structure.29 Observers from a variety of political perspectives continue to herald the triumph that this represented. In 2011, Tax Notes ran a retrospective twenty-fifth anniversary series on the Tax Reform Act of 1986. Commentators from Robert McIntyre of the left-leaning Citizens for Tax Justice30 to President Reagan’s economic adviser Martin Feldstein wrote paeans to the law, celebrating “the most monumental tax reform bill in American history.”31


For all this hoopla, however, in the years since 1986, the bill’s accomplishments have almost entirely evaporated. The tax base has eroded even more rapidly, and revenue crises have become more common. In his recent book Reforms at Risk, political scientist Erik Patashnik uses the 1986 Act as a case study of a legislative reform package that was unable to survive in the long term.32 Patashnik’s analysis demonstrates that the 1986 Act, while successful at excising exclusions, deductions, and credits, did nothing to transform permanently the underlying political dynamics of tax lawmaking.33 For this reason, as soon as the 1986 Act had passed, the tax base soon began to erode again, perhaps even more substantially than before. Interest groups began to petition members of Congress for particular tax favors.34 By 2005, when President Bush convened a bipartisan tax reform commission to study potential tax code overhauls, the Code had changed 15,000 times, amounting to an astonishing average of two changes each day.35 The tax reform panel observed: “Each one of these changes had a sponsor, and each one had a rationale to defend it. Each one was passed by Congress and signed into law.”36


2. Using Public Choice Theory to Understand Tax Lawmaking Dynamics


Public choice theory provides a framework for understanding the twin phenomena of the shrinking federal income tax base and the paralyzed tax reform apparatus. The relevant strands of public choice theory emerge in James Q. Wilson’s classic work on the politics of regulation.37 For Wilson, laws have, on the one hand, costs: costs that are either more concentrated or more diffuse. On the other hand, laws provide benefits: benefits that are also either more concentrated or more diffuse.38 Laws with concentrated costs or concentrated benefits allocate their costs or benefits to members of a relatively narrow subgroup. Most subgroup members pay a substantial cost or receive an important benefit.39 In contrast, laws with diffuse costs or benefits spread those costs or benefits across a broader group so that most members of the large group bear a relatively modest cost or receive a small benefit.40 Figure 1 lays out the four possible types of laws that Wilson’s cost-benefit framework differentiates.



Cost-Benefit Framework


Diffuse Costs


Concentrated Costs


Diffuse Benefits


(1) Diffuse costs, diffuse benefits (“majoritarian politics”)


(3) Concentrated costs, diffuse benefits (“entrepreneurial politics”)


Concentrated Benefits


(2) Diffuse costs, concentrated benefits (“client politics”)


(4) Concentrated costs, concentrated benefits (“interest group politics”)




Most relevant in this table for understanding federal tax politics are, as I elaborate below, Quadrants 2 and 3, the client-politics quadrant and the entrepreneurial-politics quadrant. As Table 1 indicates, laws in what Wilson calls the client-politics quadrant have diffuse costs and concentrated benefits.41 As such, not only are laws in Quadrant 2 relatively easy to enact, but, once enacted, they are also relatively easy to maintain and to extend. The costs of these laws are spread thinly across the members of a large social group. As a result, no individual group member has a compelling incentive to organize against the law. However, these laws concentrate their benefits on narrow groups. The groups then work hard to advocate for the passage, maintenance, and expansion of these laws.42 Wilson’s account of client politics explains, some small, easily organized group will benefit and thus will have a powerful incentive to organize and lobby; the costs of the benefit are distributed at a low per capita rate over a large number of people, and hence they have little incentive to organize in opposition--if, indeed, they even hear of the policy.43


Wilson’s paradigmatic examples of laws from the client-politics quadrant are “less conspicuous regulatory programs, such as state laws that license (and protect) occupations” or instances “where the government is supplying a cash subsidy to an industry or occupation.”44


In contrast, to client-politics laws, what Wilson calls entrepreneurial-politics laws, or Quadrant 3 laws, are frequent nonstarters.45 In these cases, the concentrated subgroups that might bear the law’s financial costs work hard to block the law’s passage. Even when an entrepreneurial-politics law does manage to pass, the concentrated oppositional groups work to erode the law’s provisions over time. In contrast, no subgroup has a sufficient incentive to fight in favor of the law. Wilson terms this political dynamic entrepreneurial politics because laws from Quadrant 3 cannot pass without extraordinary effort on the part of “political entrepreneurs.”46 He writes: “Since the incentive to organize is strong for opponents of the policy but weak for the beneficiaries, and since the political system provides many points at which opposition can be registered, it may seem astonishing that regulatory legislation of this sort is ever passed.”47


Lawmaking in the remaining two quadrants of Wilson’s framework is neither as easy as the client politics of Quadrant 2 nor as difficult as the entrepreneurial politics of Quadrant 3. Quadrant 1 is home to the majoritarian politics that accompanies laws whose costs and benefits are both widely distributed.48 Here, in Wilson’s words, “all or most of society expects to gain; all or most of society expects to pay,” and “interest groups have little incentive to form around such issues because no small, definable segment of society (an industry, an occupation, a locality) can expect to capture a disproportionate share of the benefits or avoid a disproportionate share of the burdens.”49 With no dedicated interest groups to support or oppose them, majoritarian-politics laws struggle to find places on the political agenda. When they do, however, majoritarian-politics laws may pass and survive. Whether they do depends largely on their capacity to gather ideological support from the political parties in majoritarian control of Congress.50 Wilson’s examples of majoritarian-politics laws include the Sherman Antitrust Act and the Federal Trade Commission Act.51


Quadrant 4 laws present what Wilson calls interest-group politics.52 Wilson explains that, with a law demonstrating interest-group politics, “[a] subsidy or regulation will often benefit a relatively small group at the expense of another comparable small group.”53 As a result, “each side has a strong incentive to organize and exercise political influence.”54 Whether the law then passes hinges on a struggle among contending interest groups. In “most examples of interest group politics,” neither of the opposed interest groups emerges as the total victor and there is, instead, “something in the final legislation to please each affected party.”55


Wilson’s public-choice framework offers an illuminating lens through which to view tax lawmaking, starting with the problems that the federal government has had with stymied reform efforts and continued tax base erosion. Efforts to pass Quadrant 3, or entrepreneurial-politics laws, that is, laws that would have concentrated costs and diffuse benefits, paralyze lawmakers. Reforming the Code by removing tax preference items presents a classic entrepreneurial-politics scenario.  The benefits of excising particular exclusions, deductions, or credits are diffuse. Successfully excising preference items enables lawmakers either (1) to lower tax rates across the board, (2) to grow the pool of general revenues that support broad-based programs, or (3) to lower the federal deficit. This means that the beneficiaries of eliminating special provisions are (1) the amorphous millions of taxpayers whose rates might fall, (2) the millions of citizens who receive benefits from government programs funded out of general revenues, and (3) the unborn millions of future taxpayers who otherwise will have to bear the deficit burden. At the same time, the subgroups of taxpayers that would bear the costs of eliminating any particular tax preferences are highly concentrated. The members of these subgroups then have strong incentives to protect the preferences in question and may lobby vigorously in favor of keeping the benefit in place.


To take an example of how these dynamics play out, the recent American Taxpayer Relief Act of 2012, which many observers hoped would constitute a tax reform bill, renewed, among many other tax preferences, an existing tax credit for expenses to maintain railroads.56 Removing this credit would be an entrepreneurial-politics Quadrant 3 law. It has concentrated costs for the railroad industry and diffuse benefits for everyone else. The Joint Committee on Taxation estimated the revenue loss for this credit at $232 million in 2013.57 Thus, the railroad industry had an incentive, worth $232 million in 2013 alone, to organize against reducing or eliminating the credit. In contrast, had Congress managed to remove this credit as part of a tax reform law, taxes might have fallen by $232 million, government spending out of general revenues might have risen by $232 million, or the deficit might have diminished by $232 million. For instance, say that, as in the 1986 Act, Congress had chosen to allocate the $232 million saving by lowering rates across the board. In 2009, approximately 149 million U.S. households filed tax returns.58 This means that if Congress had used the money it would have netted (from passing a law to eliminate the railroad credit) in order to give every household a tax cut, that tax cut would be worth less than two dollars per household.


In contrast, bills that will further erode the tax base are relatively ease to pass or renew since these are client-politics Quadrant 2 measures, with concentrated benefits and diffuse costs. From this perspective, laws that enact or expand targeted tax preferences are client-politics laws. Return to the railroad credit example. The law renewing the railroad credit has benefits that are highly concentrated in the railroad industry. As described above, the railroad industry is the primary beneficiary of the approximately $232 million that the railroad credit costs the federal government each year. As a result, the railroad industry has a major incentive to push through Congress legislation preserving and then expanding this credit. In contrast, the costs of this credit are highly diffuse. The separate members of the American public who may be facing--but only very slightly-- higher tax rates, reduced government spending, and growing deficits have no real incentive to spend time attempting to block pro-credit legislation. The same holds for the passage of all targeted tax preferences that erode the tax base.


Viewing the passage or renewal of tax exclusions, deductions, credits, and carve-outs as client-politics laws helps to illuminate why there are so many of these measures. Using Wilson’s theory, it is not surprising that members of Congress passed nearly 15,000 special tax measures between 1986 and 2005.59 Understanding these provisions as client-politics laws also highlights why the current dynamics of tax lawmaking paralysis are so alarming. Laws that chip away at the tax base produce revenue crises, either forcing tax rates up or government revenues down. Tax reform is the way to stop this shrinkage of the tax base. However, while tax base erosions are easy-to-pass client-politics laws, tax reform measures are hard-to-pass entrepreneurial-politics laws. As a result, the tax base continues to diminish, while the laws that would reverse the trend are stymied.


B. Earmarking as an Alternative

Wilson’s theory also suggests that the dynamics of federal tax lawmaking are not the only possible tax lawmaking dynamics. In particular, at the state level, all fifty states earmark tax revenues for specific purposes. When a governmental unit earmarks a tax, it sets aside the revenue for a specific purpose or recipient. This arrangement stands in contrast to nonearmarked taxes, which go into the pool of general revenues--a large fund that the government later disburses to many of its spending programs. While states vary in their earmarking practices, states derive on average 25% of their revenues from earmarked taxes.60


In this regard, the states differ from the federal government. Aside from Social Security and Medicare, the federal government has eschewed the use of earmarked taxes. Federal income tax revenue goes into the pool of general revenues that Congress then allocates for a variety of purposes. Although federal income taxes have no predesignated destination, this arrangement could change. Earmarking federal tax revenues for particular purposes has the potential to move tax reform legislation into one of the quadrants--Quadrants 2 or 4--that does not suffer from paralysis problems. Further, earmarking tax revenues could make targeted tax preference items harder to pass by transforming them from client-politics Quadrant 2 laws into interest-group politics Quadrant 4 laws. The following subsections discuss how this is the case.


3. Earmarking in Practice


Currently, states earmark tax revenues for a broad variety of purposes. Moreover, different earmarked taxes distribute costs and benefits differently, and they thus have different political dynamics. This subsection considers some common instances of earmarked taxes and their configuration of costs and benefits.


State-level earmarked taxes have funded many types of governmental activities. For instance, in 2008, when the National Conference of State Legislatures last surveyed states about their earmarked taxes, some of the most frequently used earmarked taxes fund transportation. In particular, highways were receiving motor fuel taxes in forty-five states.61 States also funded highways through motor vehicle registrations (eight states) and general sales taxes (seven states).62 State governments used earmarked taxes, often on motor fuels or highway use, to benefit local governments in forty-six states.63 Primary and secondary education was also a common beneficiary of earmarked taxes, receiving designated funds in thirty-five states.64 States also typically earmarked revenues for health and social services, (thirty-four states).65 The social-services taxes are particularly notable because they benefit subpopulations like the mentally ill and children, neither of which has traditionally carried weight with most legislators. Often, tobacco taxes (twenty-three states) and alcoholic beverage taxes (thirteen states) funded these health programs.66


Another frequent destination for earmarked tax revenue was environmental programs, which received precommitted funds in thirty states.67 In addition to these popular destinations for earmarked tax proceeds, states used these taxes to support a plethora of other activities. To take a few examples, Alabama earmarked part of its gas tax for airport development,68 Indiana earmarked part of its gambling tax for the Purdue School of Veterinary Medicine,69 Nebraska earmarked part of its documentary stamp tax for homeless shelters,70 and Washington state earmarked a tax on soft drinks for violence reduction and drug enforcement programs.71 Despite the prevalence of state-level earmarked taxes, however, no previous research has studied which of these targeted activities constitute revenue productive destinations for earmarked tax revenues.


States also earmark multiple kinds of taxes. As mentioned above, the motor fuel tax is the most popular, and as of 2008 it was earmarked in all but one state.72 Other common earmarked taxes were general sales (thirty-five states), tobacco (twenty-six states), alcoholic beverages (twenty-three states), and insurance and severance (twenty-six states each).73 Twenty states earmarked some of their income tax.74 Narrower sales taxes were also common. For example, Maine taxed the sale of mahogany quahogs to fund a toxin prevention program;75 Indiana taxed riverboat admissions to fund the Northwest Indiana Law Enforcement Training Center, counties and cities, county convention and visitors’ bureaus, the State Division of Mental Health and Addiction, and the State Fair Commission ;76 Oregon taxed “amusement devices” to pay for a youth group conservation corps;77 and Virginia taxed the sale of eggs to pay for research, education, and promotion of the sale and use of eggs.78


All the earmarked taxes share one key feature: They designate the purpose of tax revenues before the state government actually collects the tax revenue.79 At the time taxpayers pay the tax, they know where the revenue will go. This precommitment feature amounts to a significant spending promise from lawmakers.


This promise becomes particularly consequential given that courts have ruled that the designated beneficiaries of earmarked tax revenues have legal recourse if state governments attempt to divert the earmarked revenues in a previously unspecified direction.80 Even if some earmarked taxes do not include the legislative language necessary to create a funding promise enforceable in court, beneficiaries of earmarked taxes presumably believe that that they have a greater claim on an earmarked revenue stream than on a pool of general revenues that could go toward any destination.


2. Earmarking as a Cost-Benefit Reconfiguration


This earmarking mechanism that states so frequently employ offers a different cost-benefit structure than the one that currently characterizes the federal income tax. Unlike the federal income tax, with its diffuse roster of beneficiaries, earmarked taxes tend to have concentrated benefits. For this reason, many of the state earmarked taxes fall into the two concentrated-benefits quadrants of Wilson’s framework--Quadrants 2 and 4, client politics and interest-group politics. There are some earmarked taxes in every quadrant of the cost-benefit matrix. Even so, earmarked state taxes fall preponderantly into the quadrants with concentrated benefits. For example, state sales taxes that are earmarked for education benefit students represented by educational advocacy groups and teachers’ unions. State taxes on natural resources earmarked for environmental programs benefit environmental organizations and their constituencies. Sales taxes for health programs benefit health care providers and health policy groups. As a result of cost-benefit arrangements such as these, earmarked taxes and the programs that they fund have identifiable political constituencies to support them.


In particular, state earmarked taxes typically exhibit the attributes of laws that are easiest to pass, preserve, and grow: They fall into the diffuse cost/concentrated benefit quadrant-- client-politics Quadrant 2. This is so because many of these state taxes tax broad groups--people who buy things, people who earn income, people who use the roads--while their benefits accrue to relatively narrower subgroups. Among the many examples are income taxes like Illinois’, which is earmarked for local governments and for public education,81 or sales taxes like Florida’s, which is earmarked for local and county governments, sports facilities, an ecosystem restoration fund, and a public employee labor relations board.82 The laws that levy these client-politics taxes should be especially easy to pass, to preserve, and to expand. Thus, one would expect earmarked taxes of this type to provide particularly productive revenue sources.


States also have a number of interest-group politics Quadrant 4 tax laws, which have concentrated costs and concentrated benefits. Interest-group politics Quadrant 4 laws include natural resource extraction taxes like Louisiana’s timber severance tax earmarked for forest protection programs,83 as well as targeted excise taxes like Kansas’ excise tax on new tires earmarked for tire waste management at the city and county levels.84


Figure 2 presents the cost-benefit framework of Figure 1 with some examples of the state earmarked taxes that fall into each quadrant.



Cost-Benefit Framework with Earmarked Tax Examples


Diffuse Costs


Concentrated Costs



Diffuse Benefits


(1) Diffuse costs, diffuse benefits


(3) Concentrated costs, diffuse benefits



Maryland’s real property tax earmarked for debt service on state bonds85


Nebraska’s cigarette tax earmarked for state buildings87



Florida’s auto title tax earmarked for transportation86


Nevada’s minerals tax earmarked for local governments and debt service88


Concentrated benefits


(2) Diffuse costs, concentrated benefits


(4) Concentratedcosts, concentrated benefits



New Hampshire’s business profits tax earmarked for schools89


Arkansas’ tax on oil and brine severance earmarked for the Arkansas Museum of Natural Resources Fund91



Ohio’s income tax earmarked for libraries90


South Carolina’s tax on hospitals earmarked for indigent health care92





The earmarked taxes in the two concentrated-benefits quadrants, Quadrants 2 and 4, rest on fundamentally different dynamics than the federal income tax. Rather than impose a cost on a politically powerful subgroup to provide a benefit spread across a broad population, these taxes do the reverse. They spread a cost across a wide population to benefit a concentrated subgroup, which makes these Quadrant 4 state earmarked taxes relatively easy to pass, maintain, and grow.


For this reason, public choice theory would predict that the bases of the many earmarked taxes from Quadrant 2 and Quadrant 4 should be less likely to shrink than the federal income tax base, and the earmarked taxes should be less difficult than the federal income tax to reform. This is so because a targeted preference carved out of one of those earmarked taxes constitutes an interest-group politics Quadrant 4 law. A law of this type helps the concentrated beneficiaries of the new tax exclusion, but adversely impacts only the concentrated recipients of the tax. Sometimes, a carve-out of an earmarked tax will pass, but sometimes it will not. Often, the two concentrated but competing subgroups involved reach a compromise. Then, even if, after years of jockeying between these competing interest groups, the earmarked tax base does erode somewhat, lawmakers may be able to reform the earmarked tax. A reform bill that removed the exclusion would not be a paralyzed entrepreneurial-politics law, but an interest-group politics law, pitting two interest groups against each other. One or the other might win; more likely, the two would again compromise. In short, an earmarked tax from Quadrants 2 and 4, as so many are, should be relatively easy to pass and to sustain.


C. Potential Doubts about Earmarking

Earmarking tax revenues offers a potentially promising way to reconfigure the costs and benefits of tax lawmaking. Despite this promise, earmarking also raises certain challenges and concerns. Consequently, before proceeding to test some of the above claims empirically, I address a few of the downsides of the earmarking mechanism.


First, the specific manner in which earmarking configures statutory costs and benefits could threaten the progressivity of the tax system. Although laws with concentrated benefits and diffuse costs are easiest to pass, some of these taxes are not very progressive. The current federal income tax is (relatively) progressive, while, state sales and excise taxes are regressive. Many earmarked taxes are regressive.


As discussed above, state governments earmark a wide variety of taxes and these taxes actually vary in their degree of progressivity (as they do in many other respects). Some of these taxes, such as the several state income taxes, are in fact quite progressive. Others are progressive in a broader and more substantive sense of that word. As noted earlier, many earmarked taxes support subgroups such as poor children or the homeless--subgroups whose members may draw popular sympathy, but who are not politically powerful. The earmarked taxes that fund programs for these groups tax other groups that are relatively well-off (high-income individuals, financial institutions, people who purchase certain consumer goods) to help subgroups that are not well-off. Federal earmarked taxes could be rendered progressive in this same way. For instance, the federal government could impose a surtax on the investment income of millionaires, and earmark this revenue for services for homeless veterans. The overall structure of the tax would be progressive in effect. The program would tax a sizeable but well-off group of taxpayers to help a group that is smaller, more concentrated, but less well-off. The challenge to the federal government would be to find ways to earmark revenue while maintaining or even increasing the progressivity of the tax system as a whole.


A second and related challenge of earmarking is to ensure that its costs do not fall too heavily on disempowered groups. As noted above, state-level evidence shows that disempowered groups, instead, have fared reasonably well in obtaining revenues from earmarked taxes. However, certain risks remain. Earmarked taxes give the organized beneficiaries of a revenue source a voice in the legislative decisions that pertain to that source. This is an advantage over the current federal system in which tax beneficiary groups have no say in decisions regarding their revenue streams. However, groups without influence in the political process will remain unable to have a say over revenue sources that pertain to their members. For this reason, organized interest groups may be able to reap the benefits of earmarked taxes while politically unorganized constituencies fail to do so.


This problem is not particular to earmarked taxes, however. Politically marginalized groups also have trouble getting benefits out of the general revenue pool. Nothing suggests that the political dynamics of earmarking hurts those groups any more than do the political dynamics of the existing tax lawmaking process. As in the current tax lawmaking process, lawmakers who seek to use earmarking need to rise to the challenge of finding ways to empower groups that historically have not been able to participate meaningfully in the political process.


A third challenge that earmarking presents involves preserving legislative flexibility. If lawmakers earmark revenues for a particular purpose, those revenues are not then available for future use in the general revenues pool. The current generation of lawmakers would not then have the flexibility to tap those earmarked funds as new revenue needs arise. Future generations of legislators might be in the same situation. Political theorist Stephen Holmes’ classic work on this subject observes that democracy suffers when lawmakers “remove certain decisions from the democratic process, that is . . . tie the community’s hands.”93


Earmarking’s threat to legislative flexibility, however, may not be as great as it initially appears. This is true for two reasons. First, lawmakers who implement an earmarked tax may limit its lifespan. For instance, state governments often earmark revenues to service particular debts. As soon as the state pays off the debt, the earmarked tax expires. Federal lawmakers considering earmarking might address the flexibility challenge by adopting this same approach. Second, future legislatures can always revise statutes. If lawmakers come to question the value of an earmarked tax, they can eliminate it just as easily as any other program. In this case, lawmakers could earmark the proceeds of the tax for a new purpose or eliminate the tax altogether.





a1 Assistant Professor of Law, University of Wisconsin Law School. Thanks to Charles Camic, Lindsay Clayton, Andrew Coan, Lee Epstein, Keith Findley, David Gamage, Cori Harvey, Alexandra Huneeus, Brian Jenn, Cecelia Klingele, Shu-Yi Oei, Jason Oh, Katherine Pratt, Florence Roisman, Theodore Seto, Alexander Tahk, Bree Grossi Wilde, Lua Yuille, and Eric Zolt for extremely helpful comments. Thanks also to participants in the 2013 Loyola-Los Angeles Tax Policy Colloquium, the 2012 Midwest Junior Tax Roundtable, the 2013 Junior Tax Workshop, and audience members at the 2013 Law & Society Association conference for useful discussions. In addition, thanks to the Graduate School at the University of Wisconsin and the Arthur Andersen Summer Research Fellowship for funding; to Kyle Lawrence, Patrick Meyer, Emma Shakeshaft, and Benjamin Wright for data collection; and to Monica Mark and Steven Lloyd Wilson for truly outstanding research assistance. All errors are my own.

1   By “earmark” a tax’s revenues, I mean designate a tax’s revenues for a particular purpose. The term earmark also refers to an unpopular form of federal special interest spending. See Rebecca M. Kysar, Listening to Congress: Earmark Rules and Statutory Interpretation, 94 Cornell L. Rev. 519 (2009).

2   See Jeffrey H. Birnbaum & Alan Murray, Showdown at Gucci Gulch: Lawmakers, Lobbyists, and the Unlikely Triumph of Tax Reform 13 (1987) (referring to 1913-1986); Eric M. Patashnik, Reforms at Risk: What Happens After Major Policy Changes Are Enacted 53-54 (2008) (referring to period since 1986).

3   See Susannah Camic Tahk, Making Impossible Tax Reform Possible, 81 Fordham L. Rev. 2683, 2687-706 (2013).

4   Joseph J. Thorndike, Stanley Surrey Knew a Thing or Two About Loopholes, 138 Tax Notes 663 (Feb. 11, 2013).

5   See George W. Bush, Acceptance Speech to the Republican National Convention (Sept. 2, 2004), available at; Jimmy Carter, Address Accepting the Presidential Nomination at the Democratic National Convention in New York City (July 15, 1976), available at; Gerald Ford, “Whip Inflation Now” Speech (Oct. 8, 1974), available at; John F. Kennedy, Special Message to the Congress on Taxation (Apr. 20, 1961), available at; Barack Obama, Address Before a Joint Session of the Congress on the State of the Union (Jan. 20, 2015), available at; Ronald Reagan, Radio Address to the Nation on Tax Reform (June 1, 1985), available at

6   See President Barack Obama & Former Governor Mitt Romney, Candidates Debate (Oct. 16, 2012) (transcript available at

7   Ezra Klein, Tax Reform Is Going To Be Really, Really Hard, WonkBlog (Aug. 10, 2012),

8   Id.

9   See C. Eugene Steuerle, Urban-Brookings Tax Pol’y Ctr., Tax Reform, Federal, (last visited June 5, 2015).

10 See Tahk, note 3, at 2687-90.

11 See id. at 2691.

12 See Birnbaum & Murray, note 2, at 13-14.

13 See id.

14 Id. For more on the difficulties the Kennedy administration had getting economic bills out of committee, see Robert A. Caro, The Years of Lyndon Johnson: The Passage of Power 345-50 (2012).

15 See John F. Witte, The Politics and Development of the Federal Income Tax 166 (1985).

16 Pub. L. No. 91-172, 83 Stat. 487 (1969).

17 Birnbaum & Murray, note 12, at 14.

18 See Patashnik, note 2, at 37; Witte, note 15, at 196-97.

19 Pub. L. No. 94-455, 90 Stat. 1520 (1976).

20 See Witte, note 15, at 196.

21 Id. at 204-05.

22 Id. at 212-13.

23 Id. at 213.

24 Birnbaum & Murray, note 2, at 285; see also Michael Graetz, The Truth About Tax Reform, 40 U. Fla. L. Rev. 617, 619 (1988).

25 Pub. L. No. 99-514, 100 Stat. 2085 (1986).

26 Patashnik, note 2, at 40.

27 President Ronald Reagan, Remarks on Signing the Tax Reform Act (Oct. 22, 1986), available at

For analysis of how the 1986 Act passed, see Birnbaum & Murray, note 2; Timothy J. Conlan, Margaret T. Wrightson & David R. Beam, Taxing Choices: The Politics of Tax Reform (1990) [hereinafter Taxing Choices]; David R. Beam, Timothy J. Conlan & Margaret T. Wrightson, Solving the Riddle of Tax Reform: Party Competition and the Politics of Ideas, 105 Pol. Sci. Q. 193, 194 (1990); Dennis Coyle & Aaron Wildavksy, Requisites of Radical Reform: Income Maintenance Versus Tax Preferences, 7 J. Pol. Analysis & Mgmt 1, 1 (1987); James M. Verdier, The President, Congress, and Tax Reform: Patterns over Three Decades, 499 Annals Am. Acad. Pol. & Soc. Sci. 114, 144 (1988).

28 For the most complete versions of this argument, see Birnbaum & Murray, note 2; Conlan et al., Taxing Choices, note 27.

29 Tax Reform Act of 1986, Pub. L. No. 99-514, § 131-34, 301-31, 411-13, 100 Stat. 2085, 2113-21, 2216-21, 2225-29 (codified as amended in scattered sections of 26 U.S.C.).

30 Robert S. McIntyre, Remembering the 1986 Tax Reform Act, 133 Tax Notes 351 (Oct. 17, 2011)

31 Id. at 356.

32 Patashnik, note 2.

33 See id. at 35.

34 See id. at 43 (“After the dust settled breaks began creeping back into the tax code, shredding the principles of base broadening and horizontal equity beyond recognition.”).

35 Letter from The President’s Advisory Panel on Federal Tax Reform on a Proposal to Fix America’s Tax System, to the Secretary of the Treasury (Nov. 1, 2005) (available at

36 Id.

37 James Q. Wilson, The Politics of Regulation, in The Politics of Regulation 357, 369 (James Q. Wilson ed., 1980). For applications of this work, see, e.g., Michael D. Reagan, The Politics of Regulatory Reform, 36 W. Pol. Q. 149, 158-59 (1983); Elaine B. Sharp, The Dynamics of Issue Expansion: Cases from Disability Rights and Fetal Research Controversy, 56 J. Pol. 919 (1994); Charles R. Shipan, Regulatory Regimes, Agency Actions, and the Conditional Nature of Congressional Influence, 98 Am. Pol. Sci. Rev. 467 (2004); Bruce A. Williams & Albert R. Matheny, Testing Theories of Social Regulation: Hazardous Waste Regulation in the American States, 46 J. Pol. 428 (1984); B. Dan Wood & Richard A. Waterman, The Dynamics of Political Control of the Bureaucracy, 85 Am. Pol. Sci. Rev. 801 (1991); B. Dan Wood, Does Politics Make a Difference at the EEOC?, 34 Am. J. Pol. Sci. 503 (1990).

38 See Wilson, note 37, at 367-70.

39 See id.

40 See id.

41 Id. at 369.

42 See id.

43 Id.

44 Id.

45 Id. at 370.

46 Id.

47 Id.

48 Id. at 367.

49 Id.

50 See id.

51 Id. at 367-68.

52 Id. at 368.

53 Id.

54 Id.

55 Id.

56 American Taxpayer Relief Act of 2012, Pub. L. No. 112-240, § 306, 126 Stat. 2313, 2329 (2013).

57 Staff of the Joint Comm. on Tax’n, Estimated Revenue Effects of the Revenue Provisions Contained in an Amendment in the Nature of a Substitute to H.R. 8, the “American Taxpayer Relief Act of 2012,” as Passed by the Senate on January 1, 2013, (2013), available at

58 IRS, SOI Tax Stats, Table 1, All Returns: Sources of Income, Adjustments, Deductions and Exemptions, by Size of Adjusted Gross Income, Tax Year 2009, (last visited June 5, 2015).

59 See Letter from The President’s Advisory Panel, note 36.

60 Arturo Pérez, Nat’l Conf. of St. Legislatures, Earmarking State Taxes (2008),

61 Id at 3.

62 Id.

63 Id.

64 Id.

65 Id.

66 Id.

67 Id.

68 Ala. Code § 40-17-31(d) (repealed 2011).

69 Ind. Code 4-31-9-3 (2015).

70 Neb. Rev. Stat. § 76-903 (2014).

71 Wash. Rev. Code § 82.64.20 (amended 2009).

72 Pérez, note 60, at 6.

73 Id.

74 Id.

75 Me. Rev. Stat. tit. 36, § 4718 (2015).

76 Ind. Code § 4-33-12-6 (amended 2015).

77 Ore. Rev. Stat. § 320.100 (2015).

78 Va. Code Ann. §§ 3.2-1604-1610 (2015).

79 For more on how earmarked taxes precommit revenues and for a deeper look at the topics discussed in the next few paragraphs, see my initial exploration of earmarking in Susannah Camic, Earmarking: The Potential Benefits, 4 Pitt. Tax Rev. 55 (2006).

80 See Tuttle v. New Hampshire Med. Malpractice Joint Underwriting Ass’n, 159 N.H. 627, 992 A.2d 624 (2010); Wisconsin Med. Soc’y v. Morgan, 328 Wis. 2d 469, 787 N.W.2d 22 (2010).

81 35 Ill. Comp. Stat. 5/201, 5/901 (2013).

82 Fla. Stat. § 212.0305 (2015).

83 La. Rev. Stat. Ann. § 3:4321 (2013).

84 Kan. Stat. Ann. § 65-3424(g), (f) (West 2013).

85 Md. Code Ann., State Fin. & Proc. §8-134 (2014).

86 Fla. Stat. §319.32.

87 Neb. Rev. Stat. § 77-2602 (2014).

88 Nev. Rev. Stat. § 362.115 (2015).

89 N.H. Rev. Stat. Ann. § 77-A.20-9 (2015).

90 Ohio Rev. Code Ann. § 5747.47 (2013).

91 Ark. Code Ann. § 26-58-301 (2014).

92 S.C Code Ann. § 12.23.840 (2014).

93 Stephen Holmes, Precommitment and the Paradox of Democracy, in Constitutionalism and Democracy 195, 196 (Jon Elster & Rune Slagstad eds., 1988).

中国税法评论 | 中国财税法治论坛 | 中国政法大学财税法研究中心 | 中国税务律师论坛 | 联系我们 | 投稿