A Trillion Little Subsidies: The Economic Impact of Tax Expenditures in the Federal Income Tax Code

October 25, 2012


The loopholes known as “tax expenditures” reduce individual and corporate tax obligations by more than $1 trillion each year. But while these tax deductions are hugely popular and fiercely protected, are they really a good deal for most Americans?

A new Mercatus Center study looks at the ten largest tax expenditures for individuals and corporations and weighs the economic impact of each. The study also reviews the intended-vs-actual beneficiaries and outcomes of particular tax expenditures and considers the economic and political implications of eliminating all expenditures in a single swipe.

To read “A Trillion Little Subsidies: The Economic Impact of Tax Expenditures in the Federal Income Tax Code” and to learn more about its authors, please click here.



  • According to the Office of Management and Budget, FY2011 federal tax expenditures were more than $1 trillion—with 80 percent, or $891 billion, going to individuals, and 20 percent, or $181 billion, to corporations.
  • To put this in perspective: FY2011 tax expenditures were nearly equal to all federal income tax collected in that year, or to the entire FY2011 discretionary budget; they were also greater than the annual federal spending on Medicare or Social Security.
  • Although applied through the tax code, their effect is similar to spending provisions, hence the name tax expenditures: They encourage certain kinds of government-supported behavior by subsidizing it. Because they are part of the tax code, however—reducing revenue from what it otherwise would be, rather than overtly increasing spending—they mask the true size and scope of government.




  • Tax expenditures hinder economic growth by distorting individuals’ and corporations’ behavior toward qualifying for tax loopholes rather than making the best economic decisions.
  • By distorting behavior, tax expenditures distort the entire economic system by altering spending on goods and services; distorting capital allocation; changing the distribution of income; and encouraging lobbying and rent-seeking to maintain and expand these provisions.
  • For example, corporations must divert significant resources away from pro-growth activities to the non-productive—but critical to compete with other U.S. corporations— activity of fighting for more and bigger tax breaks.
  • Of the largest tax expenditures studied, the authors found the stated legislative intent was seldom realized.
  • Intended economic benefits seldom materialized.
  • Intended beneficiaries were seldom the greatest beneficiaries; most tax-expenditure benefits accrue disproportionately to higher-income earners, and encourage “gaming” the system by those in a position to take advantage, often resulting in cronyism and the capture of the tax code for private gain.
  • For example, while the encouragement of home ownership has become the common justification of the home mortgage interest deduction, it does not effectively achieve this goal. 
  • Of the 33 percent of taxpayers who itemize deductions, only 20 percent claim the home mortgage deduction. Of those, two thirds make more than $100,000 a year.
  • Consequently, individuals and families on the margin who could be motivated to become homeowners by incentives—that is, lower-income individuals and families—are unlikely to use this deduction.
  • With respect to tax expenditures, the authors conclude that economically optimal tax reform must include two key pieces: eliminate all tax expenditures and lower marginal rates across the board.
  • Eliminating expenditures without simultaneously lowering tax rates amounts to a tax increase on the economy as a whole. This would result in slower economic growth and, thus, lower future tax collections.
  • While a one-shot elimination of tax expenditures seems less unlikely, the authors find it more probable than a piecemeal elimination. This is because the former would give all taxpayers an immediately recognizable benefit (such as an increase in individuals’ paychecks or a significant reduction in lobbying and accounting expenses for businesses) to offset later loophole losses.


Read the paper as a PDF

Who Pays the Tax? Theoretical and Empirical Considerations of Tax Incidence

September 24, 2012

The question of who pays a tax has a discernible answer that need not be shrouded in technical mystery. This paper offers those who have little or no formal exposure to economics a tutorial in the economics of tax incidence. The analysis begins by considering a simple market for butter in a provincial town in France. A tax is placed on butter and its impact on both buyers and sellers is examined. A number of general principles of tax incidence are derived. The paper then examines three taxes used in the United States. Evidence as to the incidence of each tax is presented. The empirical analysis is broadly consistent with the theoretical predictions. The upshot is that a bit of basic economic analysis coupled with some common sense will take the untrained analyst a long way in making reliable predictions as to tax incidence.

Of the 44 states that made budget predictions for fiscal year 2012,  all anticipate budget shortfalls ranging from 2 percent in Indiana to over 45 percent in Nevada.[1]  Moreover, the federal government is currently running a budget deficit estimated to be 8.7 percent of U.S. GDP.[2]

As a result of this fiscal crisis, legislators will likely consider tax increases in future sessions. Such proposals will inevitably be debated, and an important component of these debates will be who bears the burden of the tax increase. Unfortunately, which party ultimately bears the burden of a tax can be a source of much confusion and misstatement. The purpose of this paper is to outline the basic principles of what economists call tax incidence.

A simple way of thinking about tax incidence is to consider the following conceptual exercise. Suppose a $1 tax is imposed on the sale of butter at the retail level and it is the retail outlet’s responsibility to remit the tax revenues to the government. If the price consumers pay for butter rises by a full dollar, then the burden—the incidence of the tax—is fully borne by consumers. In this case, although the retailer incurs the statutory incidence of the tax, it is able to shift forward the full economic incidence of the tax to consumers.

If, on the other hand, the price rises by less than $1, then someone other than the consumer must pay part of the tax. This tax revenue could come from retailers or from someone in the butter supply chain in the form of a reduction in their compensation. The retailer may shift backward part or all of the tax burden to the butter wholesaler, the butter manufacturer, the dairy farmer, or someone else in the production process.

The entity with the obligation to remit the tax revenues to the taxing authority may not be the entity who pays the tax. In other words, the statutory incidence is not necessarily the same as the economic incidence. This observation is important because the general public and policymakers often assume that the statutory incidence and the economic incidence are the same. What determines who pays the tax?

The side of the market with participants least able to adjust to a price change bears most of the tax. This point is the central theme of this essay.

A recent political controversy in Minnesota illustrates the problem. In an attempt to balance the state budget, legislators proposed a package of tax increases that included a surtax on lender income from credit card balances.[3] The proposed legislation would have imposed a tax “at the rate of 30 percent on any income attributable to interest collected from the portion of an annual percentage rate that exceeds 15 percent on [credit card balances and] transactions.”[4] The Minnesota Department of Revenue estimated the tax would yield $118.9 million in the first year.[5]

The bill imposes the statutory burden of the tax on those credit card lenders who charge an annual rate of interest in excess of 15 percent. As most credit card issuers are banks, it seems the bill’s intent is to balance the state budget by delivering a blow to rich banks that gouge poor consumers. Given that the credit card tax is buried in a larger tax package and is accompanied by public pronouncements that it is a tax on banks, the average voter might believe the incidence would indeed fall on banks. But even a cursory appeal to economics reveals a different story.

Suppose the hypothetical XYZ bank charges Minnesotan John Doe a rate of 21 percent on his unpaid credit card balance. Doe has an average annual balance of $10,000, yielding the bank $2,100 in income. Under the bill, the bank pays the state $180 in taxes.6 However, what prevents the bank from simply raising the rate it charges Doe to make up for the tax or from canceling Doe’s credit card? XYZ has many alternatives to lending to Doe. The high interest rate Doe pays indicates he has few alternatives to borrowing from XYZ. Thus, consumers would bear the actual burden of the tax although banks would bear the statutory incidence.

And just who are these credit card consumers? St. Thomas University economist John Spry’s examination of the proposed tax concludes that “Minnesota’s proposed thirty percent surtax on consumer interest in excess of fifteen percent would create a highly regressive tax.”[7] Specifically, “twenty percent of the new tax would be paid.

by Minnesota families with the lowest 10 percent of income. Thirty-seven percent of the tax would be paid by families with the lowest 20 percent of income.”[8]

Minnesota did not adopt the tax. The plan was derailed after a sponsor of the bill stated, “I don’t know what the consequences will be.”[9] Although the poorly designed tax did not inflict damage on Minnesota, the experience offers a cautionary tale: The party who is assessed a tax and the party who actually pays a tax are not necessarily the same. But can it be known before a tax is enacted who will actually bear its incidence? Will the retailer always be able to shift a tax burden forward or backwards? Are there general and readily discernible principles of how tax incidence is apportioned, or are policymakers forced to rely on the esoteric modeling of economic experts?

The answer is that the economic incidence of any given tax is generally knowable. Although the detailed split of the tax may require statistical expertise beyond the scope of most policymakers, some very simple principles and a bit of common sense will take one far down the road in understanding how a tax is likely to play out.

To promote an understanding of these principles, this paper offers a simple example that explains the possible ways tax incidence can be apportioned between market participants and what forces determine the apportionment. We then turn our attention to three taxes that are or have been actually imposed in the United States with an eye to what statistical analysis suggests are the actual patterns of tax incidence. In general, the evidence confirms the principles of tax incidence as valid.

Continue Reading

The Political Economy of Medicaid Reform: Evidence from Five Reforming States

April 26, 2012

Fiscal policy at both the federal and state levels is on an unsustainable path. Entitlement reform in America—particularly Medicaid reform—is shifting from a question of whether cuts should be made, to how much must be cut? To better understand best practices in Medicaid reform, we explore five recent state-level Medicaid reforms and their ability to simultaneously reduce costs, maintain or increase access, and survive the politics of reform.


Medicaid is the United States’ major health care financing system for the poor, some elderly, and the disabled. In 2010, the most recent available data, annual Medicaid spending totaled around $400 billion and accounted for more than 15 percent of U.S. health expenditures[1]. While final data for 2011 is not yet available, average monthly enrollment is estimated to exceed $55 million, with over 70 million Americans covered for one or more months[2].

Although every state has ostensibly implemented Medicaid reforms over the last 10 years, combined federal and state expenditures grew from 2.0 percent of gross domestic product (GDP) in 2000 to 2.7 percent in 2007. Some cost-saving reforms are politically dead ideas, while other—more popular—reforms only drive up costs. Successful reforms that provide a combination of cost reduction and maintained or increased access are hard to find. Thus, the easier path has been to increase expenditures.


Florida Medicaid reform of 2005

The main feature of Florida’s 2005 pilot program was the shifting of enrollees in Broward and Duval counties to managed care networks—three other counties were later added. In theory, these networks would control costs by matching enrollees with service providers. They would act as a buffer against inappropriate use (e.g., emergency care for basic treatments). The 2005 reforms also introduced benefit flexibility, incentives for healthy decisions, and premium assistance.

The results of Florida’s five-county pilot project were described as “a decided success.”[3] But a highly publicized Georgetown University study from April 2011 claimed the program did not produce significant results, limited access to prenatal care, and saved costs through low provider reimbursement rates.[4]

Florida is currently awaiting federal approval for expansions to the program. Many parties, such as researchers at Georgetown University’s Health Policy Institute,[5] are opposed to the expansion, but the case has weak evidence of cost and coverage reductions, and significant—though not insurmountable—political barriers to reform.

Idaho Medicaid reform of 2006

Idaho’s Medicaid reform aimed to tailor coverage to enrollee needs. The greater flexibility promised to reduce costs and provide care more consistent with needs. The reforms also called for healthy choice initiatives, premium sharing, consolidated purchasing of prescription drugs and medical supplies at lower prices, and other cost reductions.[6]

Idaho’s spending increased following the 2006 Idaho Medicaid Simplification Act due to its segmented approach, which reduced the risk pool causing significant adverse effects. By segmenting people into separate risk categories, the high-risk pool became underfunded, and an excessive number of risky enrollees were wrongly placed in safer pools.

The reforms were popular at the time, and then-Governor Dirk Kempthorne encountered minimal opposition.[7] This was partly because of his willingness to hold town hall meetings and public forums to discuss the reform proposal. Kempthorne combined the topics of cost savings and state control versus federal control: Idaho’s reforms promised to not leave anyone without coverage, to reduce costs, and to return more power and control to the state.

Rhode Island Global Consumer Choice Compact of 2008

Rhode Island’s Consumer Choice Compact of 2008 set spending for five years and gave state leaders flexibility to introduce market principles to the Medicaid program. The plan encouraged cost control through an incentive system: When the state spent less than the capped amount, it could keep a fraction of the federal money. In its first two years, the program cut Medicaid spending by $1.1 billion. The global waiver alone saved more than $100 million in its first year.[8]

Then-Governor Donald Carcieri ignited Rhode Island’s Medicaid reform and won people over with a compelling case: His proposals did not deny people care, but targeted care at different groups’ needs. For example, when asked if people would be denied coverage, Gary Alexander, secretary of Rhode Island’s Office of Health and Human Services at that time, replied, “If there is an elderly population that needs podiatry, we want to be able to offer it to them, rather than to the entire population.”[9]

Liberal Democrats in Rhode Island were skeptical but went along with the plan because of its popularity. With political momentum on their side, Carcieri and Alexander were able to get a large coalition on board. Thus, the major challenge for state legislators was not Republicans versus Democrats, but Rhode Island versus federal agencies handling the state’s waiver.

TennCare of 1993

TennCare dates back to June 1993, when then-Governor Ned McWherter and the Tennessee Department of Health applied for a Medicaid waiver. The main objective was to shift Tennessee’s Medicaid program from public provision to man- aged care organizations. McWherter sought rapid approval due to concern over pushback from the Tennessee Medical Association.[10] The savings from the shift would be used to expand coverage to the uninsurable and non-poor uninsured groups.

Tennessee’s reforms promised cost savings and were viewed as a radical new approach to Medicaid. However, the rapid implementation led to great confusion about care and weakened McWherter’s overall base. People were suddenly told their traditional means of health care was shifting and policies were not in place to get their questions answered. The Tennessee Medical Association took a strong position against the reforms due to the low capitation rates.

By 2000, the plan was viewed as a failure. By shifting cost savings to uninsured groups, the system faced constant demand-side pressure and never reduced taxpayers’ costs. The low capitation rates, the lowest in the country for some time, reflected the state’s failure to appropriately price risk. In essence, an adverse selection problem confronted TennCare: eligibility was expanded without concomitant increases in capitation rates. As the Tennessee Medical Association warned, poor pricing led to major increases in expenditures.[11]

Washington State’s SB 5596 of 2011

Washington State’s 2011 Medicaid reform authorized greater flexibility and a block-grant-like approach to funding. While time will tell, the general tenor of the reforms and the block-grant style, which cap expenditures and encourage innovation, are reasons for optimism.

Unlike Tennessee’s program, which was six months in the making, Washington spent six years building consensus and calculating the best reforms. Governor Chris Gregoire developed a commission tasked with recommending reforms “for Washingtonians by Washingtonians.”[13] After years of careful study, the commission submitted recommendations for reform. By taking enough time and working to overcome the concerns of interest groups and the opposition, Gregoire succeeded in getting a radical-looking reform bill passed.While time will tell, the general tenor of the reforms and the block-grant style, which cap expenditures and encourage innovation, are reasons for optimism.


If cost-saving reforms are necessary, and do not compromise well-being, what blocks their passage? Politics. Politics can kill the best of ideas. Reforms in Tennessee and Florida were radical in scale and scope. But they were not as successful as reforms in Rhode Island and Washington because they were rushed and stakeholders were less engaged. Reformers must work to bring key interest groups—even opponents—into discussions. Skeptics can provide input while gaining an understanding of the seriousness of the fiscal problems.

Rhode Island and Washington have, in effect, implemented block-grant reforms. Yet, political leaders there do not make a big deal about the radical nature of their reforms, nor do they want their reforms to be called “block grants” due to negative connotations. Reformers who consider the impact of their words and actions on their opponents are more likely to succeed.

Tennessee’s reforms are an example of bad messaging: Tennessee’s attempted reforms lacked the buy-in needed for long-term success. In contrast, Rhode Island and Washington’s leaders didn’t fixate on ideological purity. Their actions spoke louder than words.


Many states are initiating their own experiments with radical Medicaid reform. New York and Utah are both pursuing cost-saving reforms to shift Medicaid participants to privately run managed care plans. By incorporating lessons from other states, policy makers should be able to save time, build greater consensus, and deliver more effective Medicaid services to participants at a lower cost to taxpayers.


1. U.S. Department of Health & Human Services, “Fiscal Year 2010 Bud- get in Brief: Medicaid,” HHS.gov, 2011, http://dhhs.gov/asfr/ob/ docbudget/2010budgetinbriefm.html; The National Association of State Budget Offices, “State Expenditure Report,” NASBO.org, 2011, http:// www.nasbo.org/sites/default/files/2010%20State%20Expenditure%20 Report.pdf.

2. Kaiser Commission on Medicaid and the Uninsured, “Medicaid Enroll- ment: December 2010 Snapshot,” December 20, 2011, http://www.kff. org/medicaid/enrollmentreports.cfm; Kathleen Gifford et al., A Profile of Medicaid Managed Care Programs in 2010: Findings from a 50-State Survey (Washington, DC: Kaiser Commission on Medicaid and the Unin- sured, 2011), http://www.kff.org/medicaid/upload/8220.pdf. CMS data came from CMS, National Health Expenditure Projections 2010– 2020 (Washington, DC: CMS, 2010), https://www.cms.gov/National- HealthExpendData/downloads/proj2010.pdf.

3. Tarren Bragdon, Florida’s Medicaid Reform Shows the Way to Improve Health, Increase Satisfaction, and Control Costs (Washington, DC: Heritage Foundation, 2011), http://www.floridafga.org/wp-content/ uploads/Combined-Medicaid-Reform-Pilot-Nov-2011.pdf. Bragdon estimates $28.6 billion in savings for the United States as a whole if all 50 states embraced the Florida model.

4. See Joan Alker and Jack Hoadley, Understanding Florida Medicaid Today and the Impact of Federal Health Care Reform (Washington, DC: Health Policy Institute at Georgetown University, 2011), http://hpi.georgetown. edu/floridamedicaid/pdfs/Health_Reform_FL_2011.pdf. See also Alker and Hoadley, As Legislators Wrestle to Define Next Generation of Florida Medicaid, Benefits of Reform Effort Are Far from Clear (Washington, DC: Health Policy Institute at Georgetown University, 2011), http://hpi. georgetown.edu/floridamedicaid/pdfs/Medicaid_Reform_FL_2011. pdf.

5. Joan Alker, Jack Hoadley, and Jennifer Thompson, Florida’s Experience with Medicaid Reform: What has Been Learned in the First two Years? (Washington, DC: Health Policy Institute at Georgetown University, 2008), http://hpi.georgetown.edu/floridamedicaid/pdfs/briefing7.pdf.

6. Center on Disabilities and Human Development (CDHD), Issue Brief: Federal Approval of Idaho Medicaid Reform (Moscow, ID: CDHD, 2006), http://www.idahocdhd.org/Portals/44/docs/federal-approval- of-idaho-medicaid-reform.pdf.

7. National Association of Social Workers, “Lawmakers so far like Medicaid plan...” The Gatekeeper, November 2005 (reprint from Idaho Statesman, October 30, 2005), http://www.naswidaho.org/newsletter/The%20 Gatekeeper%2011-05.pdf.

8. John Graham, “In the Nick of Time: Rhode Island’s Medicaid Waiver Shows How States Can Save Their Budgets from Obamacare’s Assault,” Health Policy Prescriptions 9, no. 1 (San Francisco: Pacific Research Insti- tute, 2011), http://www.pacificresearch.org/publications/in-the-nick- of-time-rhode-islands-medicaid-waiver-shows-how-states-can-save- their-budgets-from-obamacares-assault. See also Benjamin Domenech, “The Rhode Island Medicaid Experience,” Reform + Medicaid, March 19, 2011, http://reformmedicaid.org/2011/03/the-rhode-island-medicaid- experience/.

9. John Buntin, “Ending Medicaid As We Know It,” Governing the States and Localities, June 2011, http://www.governing.com/topics/finance/ ending-medicaid-as-we-know-it.html.

10. Mark Ross Daniels, Medicaid Reform and the American States (West- port, CT: Greenwood Publishing Group, 1998).

11. Adam Zaretsky, “Revamping Medicaid: A Five-Year Check-up on Ten- nessee’s Experiment,” The Regional Economist, Federal Reserve Bank of Saint Louis, 1999, http://www.stlouisfed.org/publications/re/ articles/?id=1751.

12. AP and WATE, “Bredesen scraps TennCare,” WATE News (Knox- ville, TN), November 10, 2004, http://www.wate.com/Global/story. asp?s=2547662.

13. Washington Governor’s Office, “Governor Gregoire Enacts Health Care Reforms,” news release, May 11, 2011, http://www.governor.wa.gov/ news/news-view.asp?pressRelease=1706&newsType=1.