As part of the legislative process, the Congressional Budget Office supplies the Congress with cost estimates for legislation, economic and budget projections, and other economic assessments. Information from the research community is an important element of CBO’s analyses. This is the seventh in a series of blog posts discussing research that would enhance the quality of the information that CBO uses in its work. (Earlier posts in the series discussed the need for new research in the areas of energy and the environment, finance, health, labor, macroeconomics, and national security.) Please send comments to communications@cbo.gov.
CBO regularly provides the Congress with information about the ways that the government’s tax and transfer system affects the distribution of household income (for example, CBO 2022). That analysis is built on the models and data underlying the agency’s baseline projections of revenues and spending (CBO 2023). CBO is on the lookout for new research that would enhance its analysis of taxes and transfers, including research related to distributional analysis and projections of revenues from business income. The agency is currently working on those topics, and there are significant gaps in the relevant research literature.
In its distributional analysis, CBO indicates how different groups of people are affected by policies. Currently, the agency generally allocates certain taxes and transfer payments to households on the basis of the net dollar value of the resource costs incurred by the federal government, as measured in its budget (Habib and Heller 2022). In other cases, including when a firm interacts with the government, CBO relies on evidence in the economic literature (Gravelle 2010, Gravelle 2011) to determine how to allocate the resource costs of the government’s actions. For example, the agency allocates corporate income taxes to households in proportion to their capital income (75 percent) and labor income (25 percent). For public goods and other broad-based government spending, CBO has previously allocated the resources by using a combination of two methods: allocating in proportion to each household’s share of the population, and in proportion to each household’s share of total income.
CBO’s distributional analysis differs from an examination of economic incidence. The latter focuses on how prices and quantities of goods and services change when government revenues or spending change and assesses who is affected by those changes and their valuation of the change. Unlike that kind of analysis, the agency’s distributional allocation of resource costs is an extension of budgetary accounting that focuses on the government’s revenue collections, costs of providing a good or service, and who is affected.
Taxes on income derived from business activity are an important component of CBO’s distributional analysis and baseline projections of revenues. The tax rules and resulting budgetary effects differ in significant ways depending on the legal form of the business entity. For example, many large corporations pay the corporate income tax on their profits. Yet income from pass-through businesses—including sole proprietorships, partnerships, S corporations, and other businesses in which profits are “passed through” to owners—are reported and taxed under the individual income tax. In that case, the resulting tax liability depends on the total amount and composition of an owner’s income sources, as well as certain demographic characteristics and details related to the tax filing. CBO generally allocates profits of pass-through businesses to each individual owner, so the choice of a business’s legal structure also matters for the agency’s distributional analysis.
The method CBO uses for projecting each revenue source in its baseline varies and depends on available information. For example, individual income taxes are projected using a microsimulation model to simulate the effects of tax rules using a detailed sample of tax filers (CBO 2018), whereas revenues from the corporate income tax are projected using more aggregate methods.
CBO’s baseline projections reflect scheduled tax changes under current law, including the changes’ potential effects on the choices owners make about how they legally structure their businesses. For example, at the end of 2025, certain provisions of the 2017 tax act are scheduled to expire, which will increase the tax rate on income earned within pass-through entities—whereas the corporate income tax rate of 21 percent will remain unchanged.
How Do Changes in Federal Policy Affect Different Households?
Research that provides more information about the impact of federal policies on households’ net resources could enhance CBO’s analysis of the policies’ economic, distributional, and budgetary effects. For example, additional research on how resource costs for Medicaid are allocated could enhance CBO’s analysis. Some researchers have estimated the value of Medicaid spending using data from a randomized expansion of Medicaid coverage in the state of Oregon (Finkelstein, Hendren, and Luttmer 2019). They found that only about 40 percent of the resource cost of the program accrued to people who received coverage because of the expansion, with the remainder accruing to providers and other parties. It is unclear whether the results from that study apply broadly to the entire Medicaid program; but given the program’s size, even modest differences in the allocation could have significant effects on the pattern and trends of household income, especially among lower-income groups. Additional research that provides estimates of the distributional effects of Medicaid and also of other health-related spending, especially for different time periods and in different institutional settings, could help CBO better allocate the spending on those programs in its analysis.
Moreover, CBO’s allocations generally do not account for differences over time. For the payroll tax, for example, the agency has estimated that employers make adjustments in the allocation of resource costs to employees between the short term and the long term, but the average amount of time they take to make that adjustment is unclear (Carloni 2021). CBO generally focuses on the long-term effects of policy changes when assessing how to make allocations—not only for the sake of simplicity and tractability, but also because distributional effects can be compared consistently with the baseline distribution of household income and other policy changes in the long term. The Congress, however, may be particularly interested in differences in the short- and long-term effects of policies or in the effects of policies that differ depending on the overall state of the economy. Additional research could enhance CBO’s estimates of such effects.
How Do Taxes Affect the Way Businesses Are Legally Structured?
CBO’s baseline projections of the federal budget could be enhanced by research that provides new evidence about the likely impacts, after 2025, of taxes on the way businesses are legally structured. Much of the existing literature on that front focuses on periods before the increase in pass-through business activity that occurred in the past few decades, (Goolsbee 1998, Mackie-Mason and Gordon 1997) or on areas outside the United States (Tazhitdinova 2020).
The role of taxes in determining choices about how businesses are structured has implications for CBO’s analysis beyond the federal budget. For example, tax-motivated changes in organizational structure are estimated to account for about one-third of the decline in the measured share of income going to labor (Smith et al. 2022). That effect on measurement arises in part because of differences in the taxation of wages and business income between corporations and other pass-through entities (CBO 2012).
John McClelland is CBO’s Director of Tax Analysis. This blog post includes contributions from the following CBO staff: Dorian Carloni, Scott Craver, Justin Falk, Bilal Habib, Rebecca Heller, Jeffrey Kling, Sarah Masi, Xiaotong Niu, James Pearce, Joseph Rosenberg, Molly Saunders-Scott, Emily Stern, and Julie Topoleski.