Report

Tax Expenditure Database

quick summary

Pew's Tax Expenditure Database helped policy makers, journalists, researchers and the public better understand the role that tax expenditures play in the nation's budget and economy. Note: Pew's Tax Expenditure Database is no longer active. Downloadable files containing the underlying datasets can be found here.

Exploring the linkages between the federal tax code and state tax systems is an important part of the Fiscal Federalism Initiative's work to illuminate the complex fiscal relationship between the two levels of government and provide relevant data to federal and state policy makers.

More than half the states have tax codes linked to the federal code, as many states conform, or “piggyback,” their income tax codes to the federal code in some way. Further, certain federal tax expenditures, such as the deduction for state and local taxes and the tax exclusion for state and local bond interest, benefit state governments. This means federal tax policy changes can impact state tax codes, revenues, and spending.

Pew's Tax Expenditure Database provided a baseline for understanding existing federal tax expenditures and served as a starting point for tracking the effect of federal changes on state tax systems and revenues.

Report Contents

About Tax Expenditures

Tax expenditures are a measure of the government revenue losses resulting from provisions in the tax code that allow people or businesses to reduce their tax burden by taking certain deductions, exemptions, exclusions, preferential rates, deferrals or credits. By reducing the revenue that would otherwise have been collected by the government, tax expenditures are similar to government spending. As Figure 1 illustrates, the sum of tax expenditure estimates rivals discretionary spending in some years.1

However, these estimates are not the same as revenue estimates. Tax expenditure estimates do not model any behavioral responses. For example, if one tax expenditure is repealed or changed, a taxpayer could choose to use a different one, which would then change the estimate.2

Figure 1: Discretionary Outlays vs. Tax Expenditure Estimates in 2011 Constant Dollars ($ billions) 

Fig. 1 NEW

Blue: Total Discretionary Outlays

Orange: Sum of Tax Expenditure Estimates

Source: GAO analysis of OMB budget reports on tax expenditures, FY1985-FY2011.

Note: The tax expenditure data from FY2000-FY2011 can also be found in Pew's Tax Expenditure Database. All data presented in the database are nominal.

Pew's Tax Expenditure Database

Pew's Tax Expenditure Database helped policy makers, journalists, researchers and the public better understand the role that tax expenditures play in the nation's budget and economy.

It was a first-of-its-kind database of federal income tax expenditure estimates from the Department of the Treasury and the Joint Committee on Taxation (JCT). The database allowed users to easily view specific estimates from the Treasury and JCT and make side-by-side comparisons (see Figure 2). Users also could select and aggregate tax expenditures across 17 budget functions or economic sectors (such as energy or health) or drill down to find information about a particular tax expenditure. Specifically, users could:

  • Select a particular tax expenditure to find both the Treasury and JCT estimates reported each year from 2000 to FY13.
  • View aggregate tax expenditure totals for all individuals or all corporations, or individuals and corporations combined. 3
  • Access aggregate totals for all tax expenditures contained in each sector.
  • Examine how single or aggregate estimates change from year to year, based on historical estimates of each federal tax expenditure.
  • Download data to perform individual analysis.

Figure 2: Total Tax Expenditure Estimates by the Treasury and JCT FY2000-2011 ($ Billions)

Fig. 2

Orange: Treasury; Blue: JCT.

Source: Pew analysis of data from OMB. “Analytical Perspectives.” Budget of the U.S. Government, Fiscal Years 2002-2013 and JCT. "Estimates of Federal Tax Expenditures, Fiscal Years 2002-2013."

Notes: Numbers provided are from the most recent estimate available. Data are for individuals and corporations combined. Summing tax expenditures often provides a reasonably good estimate for the total cost of groups of tax expenditures, though it does not capture the potential interactions among tax expenditures or behavioral responses if any single one is changed or repealed

Select Findings From the Database

  • Certain budget functions account for much greater spending through the tax code than others. In FY2011, the total estimated revenue loss in the commerce and housing sector was more than four times the estimated amount lost in the 10 smallest sectors combined (in both the Treasury and JCT). Looking more closely at the Treasury's estimates, as evident in Figure 3, the four largest sectors (commerce and housing; health; income security; and education, training, employment and social services) account for nearly 80 percent of spending through the tax code.

Figure 3: Treasury Tax Expenditures by Sector (History and Projections, $ Billions) 

Fig. 3

Source: Pew analysis of data from OMB. “Analytical Perspectives.” Budget of the U.S. Government, Fiscal Years 2002–2013.

Notes: Numbers provided are from the most recent estimate available. Data are for individuals and corporations combined. Summing tax expenditures often provides a reasonably good estimate for the total cost of groups of tax expenditures, though it does not capture the potential interactions among tax expenditures or behavioral responses if any single one is changed or repealed.

  • Estimates of revenue lost from a specific tax expenditure may change significantly from year to year. For instance, the Treasury's estimate in FY2008 for "Expensing of exploration and development costs, fuels" more than tripled between its FY2009 report ($0.5 billion) and FY2010 report ($1.7 billion). Such adjustments in the estimates might result from new information about tax returns, as well as specific assumptions about tax law and economic conditions used by the government in each document; however, these data and assumptions are not made available.
  • The Treasury and JCT have different estimates for some tax expenditures. In many cases, this is due to different methodologies and assumptions; in other cases it is because of presentational differences. Consider the list of the five largest tax expenditures in FY2011, as presented in Table 1. Both report the "Exclusion of employer contributions for medical insurance premiums, and medical care" as the largest tax expenditures, and the "Deductibility of mortgage interest on owner-occupied homes" as the third largest. Yet, the order of the other three tax expenditures on each list is different. Some of these differences are largely due to presentation, or how each entity chooses to present the estimates. For example, the Treasury displays the dividends and capital gains estimates separately and JCT presents them together. For refundable credits, such as the "Earned income credit," the Treasury does not include the refundable portion, which is considered an outlay effect, in the total estimate. Rather, it lists these amounts in footnotes, which are not captured in the table of estimates. JCT does include the refundable portion in the estimate, which contributes to differences when comparing estimates in the table. Table 1 illustrates how presentation choices can affect the resulting estimates.

Table 1: Top Five Largest Treasury and JCT Tax Expenditures, FY2011 ($ Billions)

Tax Expenditure Fig 5

Source: Pew analysis of data from OMB. "Analytical Perspectives." Budget of the U.S. Government. Fiscal Year 2013 (Treasury) and JCT. "JCS-1-12: Estimates of Federal Tax Expenditures for Fiscal Years 2011-2015."

Note: Data are for individuals and corporations and are taken from FY2011 estimates.

*Part of the difference between the Treasury and JCT estimates is related to how they are displayed. The Treasury displays the capital gains and dividends tax expenditure estimates separately; JCT presents them together. For refundable credits, such as the "Earned income credit" the Treasury does not include the refundable portion, which is considered an outlay effect, in the total estimate; rather, it lists these amounts in footnotes that are not captured in the table of estimates. JCT does include the refundable portion in the estimate, which contributes to differences when comparing estimates in the table.

References

1. Discretionary spending refers to spending that is subject to annual appropriation. In contrast, mandatory spending (or entitlements) continues from year to year unless Congress chooses to change it, and the amount spent is not capped.

2. For more on why tax expenditure estimates are not exact estimates of the amount of federal revenue that would be brought in if they were eliminated, see the Methodology.

3. Summing tax expenditures often provides a reasonably good estimate for the total cost of groups of tax expenditures, though it does not capture potential interactions among tax expenditures or behavioral responses if any single one is changed or repealed. For more on why tax expenditure estimates are not exact estimates of the amount of federal revenue that would be raised if they were eliminated, see the Methodology. For more on summing tax expenditures and their interaction effects, see Burman, Leonard, Eric Toder and Christopher Geissler. "How Big Are Total Individual Income Tax Expenditures, and Who Benefits from Them?" The Urban Institute. Washington, DC. December 2008.