New JCT Model Distributes Taxes for Business Income Between Owners of Capital and Labor

This document outlines the approach that the Joint Committee staff employs, as of October 16, 2013, for distributing the taxes on the income of corporations and passthrough entities.¹ The first section provides background on economic research about distributing business taxes and describes the approaches of other agencies to distribute these taxes. The second section provides a detailed discussion of the new approach employed by the Joint Committee staff for estimating the distribution of business taxes. The third section considers an example of distributing a reduction in business taxes impacting both corporations and passthrough entities to illustrate the effect of this change. The fourth section briefly summarizes this report and indicates that the new methods adopted by the Joint Committee staff for distributing taxes on corporations and on businesses conducted in passthrough entities will continue to be refined to reflect the current understanding in the economics profession.

The Joint Committee staff welcomes comments regarding the modeling of the distribution of taxes on business income.

Harberger's conclusion that current corporate capital owners bear the entire corporate tax burden in the short run is reaffirmed by subsequent research.³ However, recent research has noted limitations of the model in other respects. The most significant of these limitations is that Harberger's original results were built on the assumption of a closed economy, in which capital could not flow internationally among countries. In an open economy, where capital is mobile and goods can be traded internationally, capital owners can escape a portion of the long-run burden of business taxes and shift that burden onto domestic labor through lowering wages or reducing employment.⁴ Some economists, however, emphasize that since the capital stock of the United States is large relative to the rest of the world, one may expect this outflow to reduce the worldwide rate of return on capital and limit the ability of capital owners to escape the incidence of the tax.⁵ Even with this limitation, some researchers have estimated that labor actually bears the majority of the burden of the corporate tax. Under the assumption that capital is perfectly mobile while the stock of labor is fixed, one study estimates that domestic labor bears approximately 70 percent of the long-run burden of the corporate income tax while owners of domestic capital bear just 30 percent.⁶ Similarly, another study estimates that, for a large country such as the United States, if capital is mobile and labor is immobile, then labor bears almost the entire corporate tax burden.⁷

Nevertheless, while some research suggests that, when accounting for capital outflows, labor may bear most, or even all, of the corporate tax incidence, there are also reasons to believe that such results may understate the share borne by owners of domestic capital. Models concluding that most corporate income taxes are borne by labor generally assume that the U.S. after-tax rate of return to capital has no effect on worldwide capital supply,⁸ fail to capture the current tax-deductibility of corporate debt,⁹ ignore foreign reactions to U.S. tax changes,¹⁰ or assume that all capital receives just a normal rate of return.¹¹ Also, they typically assume that labor is completely immobile, while in reality one may expect some labor mobility. If these assumptions do not hold, the burden of the corporate tax shifts partially back towards owners of domestic capital.¹² Furthermore, most researchers adopt an open-economy general equilibrium model, as discussed above, to estimate the incidence of corporate income taxes. However, under an alternate differential allocation model, approximately 94 percent of the corporate tax falls on owners of domestic capital in the long run.¹³

The debate over the incidence of corporate income taxes is ongoing, with a range of estimates on the precise breakdown depending on the assumptions of underlying models. Nevertheless, the existing research has arrived on two clear points of consensus. One is that the burden of the corporate income tax falls largely on domestic individuals, and therefore the corporate income tax does impact the well-being of these individuals. The second is that the burden of corporate income taxes is not borne entirely by capital owners, and is instead shared between capital owners and labor with the share borne by each being the subject of ongoing debate.

In response to the evolving economic consensus on corporate income taxes, other government agencies interested in tax distribution have updated their distributional methods. For example, from 1996 through 2010 the Congressional Budget Office ("CBO") allocated the entire corporate income tax to owners of domestic capital -- which was in line with the initial Harberger assumptions.¹⁴ More recently, however, CBO updated its distribution methods to reflect that a portion of the corporate income tax is borne by labor. The CBO now allocates 75 percent of corporate income taxes to capital owners and 25 percent to labor in its long-run estimates.¹⁵

Similarly, the Department of the Treasury (the "Treasury") assumed that the incidence of corporate income taxes rested entirely on owners of domestic capital until recent years.¹⁶ Starting in 2012, the Treasury adopted a more complex approach, noting that capital owners receive both a normal (risk-free) rate of return and a supernormal return.¹⁷ The supernormal return represents the additional return on their risky investment, returns to monopolies, or any other returns that result in profits above those which could be obtained in a risk-free environment. This approach allows the Treasury to estimate differential treatment of tax changes based on the extent to which the change impacts normal returns to capital and labor or the supernormal returns received by some capital owners. Nevertheless, despite this additional complexity, the overall distribution of corporate income taxes to capital (both normal and supernormal returns) and labor is relatively similar to that estimated by the CBO. Overall, the Treasury allocates approximately 82 percent to capital and 18 percent to labor.

The Joint Committee staff has traditionally not distributed corporate income taxes to individuals.¹⁸ However, given the current economic research on the distribution of corporate income taxes, it is appropriate to distribute these taxes to individual taxpayers rather than to ignore these taxes in a distributional analysis, effectively assuming that such taxes do not affect the overall distribution of tax liabilities. Thus, the approach described below distributes corporate income taxes in such a way that is consistent with the current economic research. However, some uncertainty remains regarding the division of the incidence between owners of domestic capital and labor so the Joint Committee staff will continue to evaluate this approach to be consistent with new research findings as they arise.

Thus far, the discussion of methodologies for distributing business taxes has focused on corporate income taxes rather than taxes on the income of businesses organized as passthrough entities. The Joint Committee staff has traditionally distributed some passthrough tax changes to the passthrough owner, and left some passthrough tax changes undistributed. This distinction was based on whether the legislative change occurred on the individual tax return or occurred at the level of the passthrough entity, which would have required determining how much income, deductions, and tax credits to pass through to each individual taxpayer in the data. For example, a change in the depreciation schedule of passthrough entities would not be distributed since it occurred at the level of the passthrough entity, but a change in the tax rate on passthrough income of individuals would be distributed because this change occurs on the individual tax form. However, given the reevaluation of the distribution of corporate income taxes to capital owners and to labor based on the recent business tax incidence literature, it is also appropriate to reevaluate the treatment of taxes on the income of passthrough entity businesses to reflect that both capital owners and labor bear a portion of the incidence of taxes on passthrough businesses.

While there is little direct research on the incidence of taxes on passthrough entity businesses, the same principles of the general business tax literature should apply. Like corporations subject to tax under subchapter C ("C corporations"); sole proprietorships and passthrough entities such as corporations subject to the rules of subchapter S ("S corporations") and partnerships (including limited liability companies, which are generally treated as partnerships for Federal tax purposes) use capital and labor as factors of production to produce income. These entities often compete in the same domestic market for labor with C corporations. They also compete with C corporations for capital. Accordingly, the taxes on the income of passthrough entity businesses should be borne partially by labor and partially by capital owners. However, passthrough capital owners may have less ability to shift their capital abroad due to the size and nature of their businesses. As will be discussed further below, passthrough entities receive substantially less foreign income than C corporations, reflecting the more domestic nature of their operations. As a result of their limited ability to shift capital abroad, the incidence is expected to fall more heavily on owners of domestic capital than is the case for taxes on C corporations.

The new method for distributing business taxes matches the longstanding practice of the Joint Committee staff of distributing individual taxes on an annual basis for each year in the budget window. In doing so, the current incidence of present-law taxes on corporations and on owners of passthrough entities is assumed to match the long-run incidence of corporate taxes and taxes on passthrough owners.¹⁹ The market effects of tax changes, however, take time to reach their long-run equilibrium.

In the very short run, the incidence of business tax changes should fall entirely on the holders of the existing capital stock and bond holders.²⁰ This is because the owners of the existing capital have almost no capacity to immediately shift their assets to other sectors of the economy or to foreign holdings, or to substitute labor for capital in the production process. Similarly, bond holders are impacted by the after-tax rate of return on capital since it determines the returns they can demand on their investment, but current bond holders have no ability to immediately adjust bond holdings prior to the change in the bond values that results from tax changes. Further, it takes time for wages and employment to adjust to reflect labor's share of these business taxes. As a result, in estimating the distributional effects of business tax changes, the initial burden of business tax increases (or benefits from business tax reductions) is assumed to fall entirely on capital owners, which includes both the owners of the business and the bond holders. In the short run, there is no incidence on labor. This is true both for corporate income taxes where the entire tax falls on current owners of corporate equity and debt, and for passthrough entities where the entire incidence initially falls on the passthrough owners who are unable to shift any of the burden to labor in the very short run.

As previously noted, economic analysis concludes that in the long run owners of domestic capital are more easily able to escape some of the burden of the tax so business taxes are at least partially passed on to labor. In estimating the long-run burden of corporate income taxes on capital and labor, the Joint Committee staff follows the middle range of the current economic literature by assuming that 25 percent of corporate income taxes are borne by domestic labor and 75 percent are borne by owners of domestic capital. Some of these owners of domestic capital, however, are foreign individuals, who are excluded from Joint Committee staff distribution tables. Following the standard view expressed in the economic literature, the Joint Committee staff's distributional methodology assumes that none of the burden of corporate income taxes flows through to consumers. These long-run incidence assumptions match those currently made by the CBO.

Some passthrough entities are multinational businesses that are largely indistinguishable from C corporations in terms of their capital mobility. However, others are smaller closely-held businesses that are less able to move capital overseas while still operating the business. Furthermore, to the extent that smaller passthrough entities have a domestic (rather than foreign) market for their goods and services, they may require operations that are geographically close to their customer base in the United States, further restricting the ability to shift operations abroad.

Recognizing these constraints, however, businesses conducted in both corporations and passthrough entities may be able to escape some of the tax burden through international flows. A Joint Committee staff analysis of sources of income of corporations and passthrough entities observed that approximately 20.6 percent of corporate gross income is from foreign income, while only about 7.6 percent of passthrough income is from foreign sources. This implies that while some passthrough entities do have international operations to which they may shift some capital, the ability of passthrough entities to shift capital abroad generally is more limited than that of C corporations. In those situations where capital in passthrough entities is less able to avoid the burden of the tax through international capital flows, capital owners of passthrough entities bear a larger share of the tax burden. This relatively larger incidence of tax on domestic capital owners of passthrough entities gives rise to a relatively smaller incidence on labor, the other factor of production. Given the assumption of a 25-percent incidence on labor for C corporation taxes, the Joint Committee staff determined that the relatively smaller ability of passthrough owners to shift capital abroad implies between a five- and 10-percent incidence on labor of taxes on passthrough entity business income. Given the paucity of empirical and theoretical literature on incidence effects in passthroughs, the Joint Committee staff believes that it is appropriate to choose an incidence on labor at the low end of this range.²¹ Thus, the Joint Committee staff assumes that five percent of taxes on business income of passthroughs is borne by domestic labor and 95 percent is borne by domestic capital owners in the long run.

Although economists generally agree that owners of domestic capital bear the entire burden of business tax changes in the very short run, and that the burden is shared between owners of domestic capital and labor in the long run, a crucial challenge in developing a practical distribution analysis is that there is no clear definition of when the long-run equilibrium is reached. Further, the time-frame for reaching such equilibrium may differ by firm or by tax change.²² For some tax changes, the long-run equilibrium may be reached quite quickly, but in other cases it may be a slow transition. In the absence of a clear theoretical determination of the long-run, the Joint Committee staff generally assumes that a provision enacted in the first year of the ten-year budget window reaches a long-run equilibrium in the final year of the 10-year budget window. However, for proposals with delayed effective dates, the phase-in may be adjusted to reflect the fact that individuals may not fully adjust to the tax change by the end of the budget window.

Joint Committee staff distribution analyses are produced using the Joint Committee staff individual tax model, which is built around a sample of individual income tax returns and associated information returns provided by the Statistics of Income Division of the Internal Revenue Service.²³ While the data in this model is well suited to the analysis of the distributional effects of most provisions affecting individual income taxation, additional modeling and assumptions are necessary to add information about business tax incidence.

In allocating labor's share of the corporate tax, the Joint Committee staff begins with the assumption that labor is mobile across corporate and non-corporate employment. Since labor pays its share of corporate income taxes through lower compensation (based on the after-tax value of marginal product), if an increase in corporate income taxes results in lower demand for labor and lower compensation at corporate firms, labor moves towards non-corporate employment. This, in turn, drives up compensation for corporate firms as their supply of labor contracts and drives down compensation for non-corporate firms as their supply of labor expands until balanced pay across sectors is restored. Thus, once labor's share of the corporate income tax is determined, this share should be distributed across all labor regardless of the type of employer. Additionally, the impact of corporate income taxes on labor compensation should be based on each worker's total compensation package and not just based on cash wages. Each worker's share of the corporate income tax is therefore computed as labor's total share of the corporate tax change times the individual's compensation relative to the total compensation of all workers. Compensation is defined for this purpose as the sum of the worker's wages as reported on the tax return, payroll taxes paid by the employer on behalf of the worker, employer contributions to health insurance plans, and untaxed voluntary contributions to retirement plans made by the worker.²⁴

The distribution of total employee compensation by the prior expanded income definition is shown in Table 1. This distribution reflects the distribution that the Joint Committee staff assumes for distributing labor's share of business taxes. As will be discussed further in the subsequent sections of this document, the distribution of labor compensation is less concentrated among the top income classes than is observed for either corporate capital income or income received by owners of passthrough entity businesses.

        Table 1.-- Estimated Distribution of Labor Compensation

In contrast to the approach for distributing labor's share, where it is assumed that wages equalize across the corporate and non-corporate sectors, the Joint Committee staff does not assume equalization of rates of return for corporate capital and capital of passthrough entities. As such, corporate tax changes are not assumed to be borne by owners of passthrough entities and tax changes with respect to passthrough entities are not assumed to be borne by corporate capital owners. This is partially because there are restrictions on entity choice for individuals forming a business, such as the restriction that S corporations may have no more than 100 shareholders, all of whom must be either individuals who are U.S. citizens or residents, or certain estates, trusts, or tax-exempt organizations. Such restrictions prevent after-tax rates of return from equalizing as between corporate and non-corporate capital. This determination is also consistent with the earlier observation that C corporations are more likely to have a greater international presence than passthrough entities have, reflecting the fact that entity forms are not completely interchangeable. For distributing capital's share, note that virtually all domestic corporate capital is held by individuals in taxable or tax-deferred accounts, by other corporations, by tax-exempt entities, or by foreign entities and individuals. The distributional effect of each of these types of holding is discussed below in turn.

The goal of the distributional analysis is to identify the incidence of taxes paid by individual taxpayers. In doing so, capital held by other corporations has no direct impact on the distribution of corporate income taxes. This is because, when Alpha Corporation holds stock in Beta Corporation, it is functionally economically equivalent to the shareholders of Alpha holding that portion of Beta shares directly. The return to Beta's capital simply flows through to Alpha's capital owners.²⁵ Because the method does not attempt to link capital owners to individual firms and instead focuses on the total level of capital individuals have invested across all firms, this secondary level of ownership does not impact the results and can therefore be ignored.

Capital held by tax-exempt entities is less straightforward, but similarly is assumed to have no effect on the distribution of capital's share. Although tax-exempt organizations have stakeholders rather than shareholders, it is assumed that the stakeholders reap the benefits of the returns to capital of the tax-exempt organization. These stakeholders are generally domestic individuals. Unfortunately, however, there is no clear way to identify the stakeholders in tax-exempt organizations. Therefore, for simplicity it is assumed that stakeholders in tax-exempt organizations are distributed through the population in a way that matches the distribution of domestic individual capital owners and no adjustment is made for capital ownership by tax-exempt organizations.²⁶ However, there are limitations of this assumption and the approach may be updated in the future based on continued research on the income distribution of tax-exempt organizations' stakeholders as well as additional data on the amount of tax exempt organizations' asset holdings.

In contrast to capital held by corporations or tax-exempt organizations, domestic capital held by foreign entities and individuals does have a distributional effect. This is because the foreign capital owner is the end-individual bearing the burden of the tax. Thus, to the extent that corporate capital is owned by foreign entities and individuals, the tax burden is borne by foreign citizens. This, in turn, reduces the share of the tax burden borne by U.S. citizens and residents.²⁷

            Table 2. -- Foreign Ownership of U.S. Corporate Equity,

The remaining 89.2 percent of capital's share of corporate income taxes is distributed to U.S. individuals based on their capital holdings. Capital holdings of individuals are held across a range of account types, each of which are reported differently in economic data including the tax data used for Joint Committee staff tax models. Specifically, capital holdings can be held as (1) taxable equities, (2) taxable and tax-exempt bonds, (3) assets in tax-deferred individual retirement accounts ("IRAs") and Roth IRAs, and (4) assets in defined contribution retirement plans. Assets in defined benefit retirement plans are discussed separately below. The Joint Committee staff estimates each individual's share of all assets held in each of these four holding types. The overall distribution of capital holdings is then determined based on each individual's share of assets in each holding type along with information on how assets are divided across the four holding types.

Table 3 illustrates, from the Federal Reserve Flow of Funds data, the division of assets held by individuals across these four types of asset holdings. Over the period from 2005 through 2012, 60.2 percent of individuals' capital holdings are corporate equity in taxable accounts, 9.5 percent are taxable corporate bonds, 16.8 percent are assets in tax-deferred and Roth IRA accounts, and 13.5 percent are assets in defined contribution retirement plans. These allocations illustrate how assets are divided across holding types, but, without information on the distribution of assets within each holding type, they yield no direct information on capital ownership by income level as is necessary to distribute capital's share of corporate taxes. To determine the relationship between capital holdings and income, the Joint Committee staff estimates the distribution of capital assets to all individuals for each of these four classes of asset holdings. The overall population-wide distribution of domestic capital's share of the corporate income tax burden is calculated as a weighted-sum of these four distributions. In other words, each individual's share of the corporate income taxes that fall on domestic owners of corporate capital equals 0.602 times their share of taxable corporate equity holdings in the population plus 0.095 times their share of taxable corporate bond holdings plus 0.168 times their share of IRA holdings plus 0.135 times their share of defined contribution plan holdings.

     Table 3. -- Fraction of Capital Holdings of Individuals by Asset Type

Given the theoretical reasons for treating defined benefit retirement plan holdings as owned by the plan sponsor, it is appropriate to return to the original four individual holding classes listed previously. For taxable corporate equity holdings, individuals report on their tax returns the level of dividends and capital gains received in each year. However, the distribution depends not on the flow of income generated by assets but instead on the stock of taxable corporate equity held by each individual.

To determine the stock of taxable corporate equity held by each individual, the analysis focuses on the dividend receipts reported in tax return data. Unlike reported realized capital gains that represent the sale of an asset (which therefore is no longer held by the individual), dividends are paid based on assets that an individual currently owns. Additionally, outside of closely-held corporations, dividend recipients are unable to time the receipt of payments in an effort to delay tax liabilities in the way that capital gains realizations may be timed. Furthermore, while the flow of realized capital gains is only loosely tied to the stock of an asset sold, varying based on the initial purchase price of a holding and which lots the capital owner chooses to sell, dividends are based completely on the current level of capital held. As a result, two individuals with equal amounts of equity in the same company on a given day receive the same level of dividends even though they might report dramatically different levels of capital gains (or capital losses) should each choose to sell the asset.

Of course, dividends are not a perfect measure of corporate equity capital holdings. Not all corporations pay dividends, and even among those that do the dividend yields are not uniform. While some investors certainly seek out assets that pay dividends and others seek out assets that do not, it is assumed that these differences are randomly distributed throughout the income distribution. Thus, while this approach may overestimate or underestimate the corporate equity capital holdings of specific individuals, it is assumed that this is not the case for the aggregate classes of individuals presented in the distribution tables.²⁹

To test this assumption and the sensitivity of the analysis to the use of dividends data, rather than capital gains or asset sale data, the Joint Committee staff compares the distribution of sales of taxable assets reported on Schedule D of the tax return from the 2007 Sale of Capital Assets ("SOCA") data to the distribution of dividend receipts reported on 1040 tax returns. This data reports the sale price and capital gains realizations from all assets sold by individuals and reported on tax returns. This data is from a special study and is not available in every year, making it unsuitable to directly incorporate into the distributional analysis. However, it is valuable for testing the approach of using dividends to estimate the distribution of taxable equity holdings. As is the case with dividends data, the asset sales data is an imprecise measure of asset holdings, although the imprecision is different than that seen when considering using dividends data. The imprecision in the asset sales data arises because it includes only the value of the subset of assets that are sold rather than the value of all assets that are held. Because the sources of the imprecision in the two measures should be uncorrelated, to the extent that they produce similar results it supports the assumption that both dividend distributions and the asset sales distribution represent the true underlying distribution of asset holdings.

This comparison is provided in Table 4 which shows the 2007 distribution of dividends and asset sales by the original expanded income definition. Outside of the income classes with expanded incomes under $10,000 or over $1 million, dividend distributions are relatively well aligned with the distribution of asset sales. However, asset sales show that a lower fraction of assets is held by taxpayers with expanded income greater than $1 million than that observed based on dividend payments. This partially reflects the fact that dividends are included in expanded income, creating a natural correlation of dividends appearing higher in the income distribution than asset sales. However, it may also reflect the fact that the SOCA data only includes assets that are actually sold. To the extent that individuals tend not to sell assets to take advantage of the step-up in basis at death or other tax-minimization and tax-deferral techniques, such behavioral responses to tax rules could explain the drop-off in asset sales for high-income individuals. As a result, one may expect the share of asset sales at the top of the distribution to be relatively low compared to the share of asset holdings. Given the general consistency of these two measures for the vast majority of the AGI distribution and the expected divergence for the top, the Joint Committee staff uses dividends data to estimate the distribution of taxable capital assets.³⁰

     Table 4. -- Comparison of the Distribution of Dividends by AGI to the

Recognizing that not all capital holdings are held in taxable accounts, the analysis also incorporates the major tax-preferred avenues for holding capital: personally held IRAs and employer-sponsored defined contribution retirement accounts such as 401(k) retirement plan accounts. In contrast to all other holdings of capital, IRA custodians are required to report the fair market value of each IRA account at the end of each year, providing direct information about IRA assets. The 16.7 percent of capital held in IRA accounts is distributed to individuals based on the end-of-year fair-market value of IRA holdings.³¹

Finally, for 401(k) and other employer-sponsored defined contribution retirement accounts, the distribution is imputed based on the age and estimated current 401(k) contribution level for each individual. Thus, as one may expect, as individuals age or increase their incomes (and thus their 401(k) contribution amounts), the expected balances in their defined contribution retirement accounts grow. For the population as a whole, 13.4 percent of capital assets are held in defined contribution retirement plans such as 401(k) accounts.

The distribution of corporate capital by holding type, calculated as described above, is shown in Table 5, which presents the 2007 capital holdings by income source by the prior expanded income classes, along with the overall distribution of capital across all holding types. Each class of corporate capital holdings is predominantly concentrated among the top expanded income classes. The taxable equity holdings and the taxable and tax-exempt bond holdings in particular are held predominantly by the top income groups, with approximately 30 percent of each being held by tax-units with expanded incomes over $1 million. In contrast, tax units with expanded incomes over $1 million hold just 3.3 and 2.0 percent of the tax-deferred IRA and 401(k) assets respectively. This is expected since the income and contribution limits for contributing to IRA accounts and the contribution limits for contributing to 401(k) accounts both limit the ability for these tax-preferred assets to be highly concentrated among a small number of individuals.

Because the primary objective of distributing corporate income taxes is to understand the distributional effects of tax legislation, the Joint Committee staff imputes the growth of capital and wages over the course of the 10-year budget window in a manner consistent with the Joint Committee staff's individual income tax model. This imputation is calibrated annually to match the Joint Committee staff's and CBO's targets for gross domestic product, tax receipts, number of tax units, and a wide range of other economic and demographic variables. Using this imputation, the Joint Committee staff can estimate the distributional effects of corporate tax changes for each of the 10 years in the budget window.

     Table 5. -- Estimated Distribution of Capital Holdings by Holding Type

The Joint Committee staff has also reexamined the incidence assumption used for distributional analysis of tax changes with respect to business income of passthrough entities and sole proprietorships. The revised analysis assumes that 95 percent of a tax change is borne by the business owner and five percent by labor. As described above in the context of attributing labor's share of corporate tax incidence to individuals, wages are expected to equalize across business sectors. The same analysis applies in the attribution to individuals of labor's share of taxes on business income of passthrough entities and sole proprietorships. In other words, labor's share of the taxes affecting income earned through passthrough entities is distributed to all workers in proportion to their total compensation, including wages as reported on the tax return, payroll taxes paid by the employer on behalf of the worker, employer contributions to health insurance plans, and untaxed voluntary contributions to retirement plans made by the worker.

For capital's share of the taxes affecting income earned through passthrough entities and sole proprietorships, it is assumed that this tax is borne by the passthrough owners or the sole proprietor. This tax, therefore, falls on individuals who report income from the passthrough entity or sole proprietorship on their individual tax returns. This is calculated as the sum of reported business income and Schedule E income on individual tax returns. Like the procedure for distributing corporate tax changes, the growth pattern of passthrough income and sole proprietorship income is imputed to the Joint Committee staff's individual tax model over the 10-year budget window for estimating purposes.

Expanded income is further adjusted to reflect the revised treatment of taxes with respect to business income of passthrough entities and sole proprietorships. Since passthrough owners report the passthrough income on their individual tax returns and then pay the tax with their personal income taxes, this treatment accurately reflects a pre-tax and transfer income concept if the tax burden were borne entirely by the passthrough owner. However, since the Joint Committee staff estimates that five percent of the passthrough tax is actually borne by workers rather than owners, the income reported by passthrough owners on their individual tax returns actually exceeds that in a pre-tax environment. Thus, taxes attributable to business activities of passthrough entities are accounted for in expanded income by reducing the expanded income of passthrough owners by five percent of the amount of taxes attributable to their passthrough entities and increasing the expanded income of workers by that same amount. The same is true for income received by owners of sole proprietorships. Combining these modifications with the prior expanded income definition described in JCX-15-12 results in expanded income being defined as follows:

The impacts of this change on the 2013 distribution of expanded income and the taxes paid by expanded income level are presented in Table 6 through Table 9. Because of the additions to the pre-tax definition of expanded income, these changes shift the overall expanded income distribution upwards, decreasing the number of returns in the expanded income classes below $75,000 and increasing the number in the classes above $75,000 (Table 6). This movement of tax returns up the income distribution similarly increased the total level of expanded income reported by each of the income classes above $75,000 (Table 7). Furthermore, when considering the tax liabilities by economic income group, including the corporate income taxes also increases the total taxes (Table 8) and average tax rate (Table 9) of individuals throughout the distribution, but particularly among the higher income groups.

     Table 6. -- Impact of New Distribution Methods on the Baseline

Over the course of the budget window, the fraction of the tax change allocated to owners of capital and bond holders gradually declines to 75 percent, with the fraction allocated to domestic labor increasing from zero to 25 percent. Because the fraction of corporate taxes distributed to all owners of domestic capital declines, the fraction allocated to foreign owners of capital also declines. Therefore, by 2023, the last year of the budget window, only around $0.8 billion of the $10 billion increase in corporate income taxes is distributed to foreign persons and excluded from the domestic income tax distribution table. Considering the distributions to domestic individuals, since workers tend to be lower in the income distribution than capital owners, the shift of a portion of the burden to labor slightly reduces the extent to which the corporate income tax increase is borne by individuals in the top income classes. In 2023, 77 percent of the domestically distributed corporate income tax increase is allocated to individuals with expanded incomes over $100,000 and slightly less than 23 percent is allocated to individuals with expanded incomes over $1 million.

              Table 10. -- Distribution of a $10 Billion Per Year

As was the case for a corporate income tax increase, the burden of a tax increase on income of passthrough entities is concentrated on higher income individuals. However, this concentration is even greater when considering taxes with respect to businesses in passthrough entities than it was for the corporate income tax increase in Table 10. In the short-run, the tax increase is completely borne by the owners of passthrough entities. However, by 2023, the last year of the budget window, five-percent of the burden is assumed to fall on labor.

The increased share of the tax increase borne by labor in the long-run relative to the short-run also slightly increases the share of the tax increase distributed to individuals lower in the income distribution in the latter years of the budget window. However, this shift in the burden towards labor is overwhelmed by the Joint Committee staff's estimation that passthrough ownership will be increasing among individuals in the highest income classes. As a result, the share of the tax increase with respect to income earned in passthroughs that is borne by individuals with expanded incomes over $1 million is the highest in 2023, when approximately 47 percent of the tax increase is borne by individuals in this income class. This is an increase from approximately 40 percent being distributed to individuals with expanded incomes over $1 million in 2013.

In all years of analysis the share of the tax increase distributed to the highest income classes is greater for the tax increase with respect to income of passthrough entities than was seen for the corporate tax increase. For example, in 2023 approximately 84 percent of the tax increase with respect to income of passthrough entities falls on individuals with expanded incomes over $100,000. In contrast, only 77 percent of the corporate tax increase in that year fell on these individuals. The larger concentration of the tax burden with respect to income of passthroughs borne by individuals in the top income classes results from two principal factors. One factor is that labor's share of the the tax burden with respect to income of passthroughs is smaller than it is for the corporate income tax burden. The other factor is that passthrough owners are more heavily represented among higher income groups than are corporate capital owners.

          Table 11. -- Distribution of a $10 Billion Per Year Increase

² Arnold C. Harberger, "The Incidence of the Corporate Income Tax," Journal of Political Economy, 70:3, June 1962, pp. 215-240.

³ Alan J. Auerbach, "Who Bears the Corporate Tax?" in James Poterba, ed., Tax Policy and the Economy Vol. 20, ed., Cambridge, MA: MIT Press, 2006, pp. 1-40, and Arnold C. Harberger, "Corporate Tax Incidence: Reflections on what is Known, Unknown, and Unknowable" in John W. Diamond and George R. Zodrow, eds., Fundamental Tax Reform: Issues, Choices, and Implications, Cambridge, MA: MIT Press, 2008.

⁴ Arnold C. Harberger, "The ABCs of Corporate Tax Incidence: Insights into the Open-Economy Case," Prepared for Symposium of the American Council for Capital Formation Center for Policy Research, Tax Policy and Economic Growth, June 8, 1994.

⁵ Harry Grubert and John Mutti, "The Taxation of Capital Income in an Open Economy: The Importance of Resident-Nonresident Tax Treatment," Journal of Public Economics 27, August 1985, pp. 291-309 and Julie Anne Cronin et al. "U.S. Treasury Distributional Analysis Methodology," Office of Tax Analysis Working Paper, 2012.

⁶ William C. Randolph, "International Burdens of the Corporate Income Tax," CBO Working Paper 200609, 2006.

⁷ Arnold C. Harberger, "Corporate Tax Incidence: Reflections on what is Known, Unknown, and Unknowable" in John W. Diamond and George R. Zodrow, eds., Fundamental Tax Reform: Issues, Choices, and Implications, Cambridge, MA: MIT Press, 2008.

⁸ Mihir A. Desai, C. Fritz Foley, and James R. Hines, "Labor and Capital Shares of the Corporate Tax Burden: International Evidence," Prepared for International Tax Policy Forum Conference, Who Pays the Corporate Tax in an Open Economy, December 18, 2007.

⁹ Alan J. Auerbach, "Who Bears the Corporate Tax?" in James Poterba, ed., Tax Policy and the Economy Vol. 20, ed., Cambridge, MA: MIT Press, 2006, pp. 1-40 and Arnold C. Harberger, "Corporate Tax Incidence: Reflections on what is Known, Unknown, and Unknowable" in John W. Diamond and George R. Zodrow, eds., Fundamental Tax Reform: Issues, Choices, and Implications, Cambridge, MA: MIT Press, 2008.

¹⁰ Michael P. Devereux, Ben Lockwood, and Michela Redoano, "Do Countries Compete over Corporate Tax Rates," Journal of Public Economics, 92:5-6, 2008, pp. 1210-1235, and Jennifer C. Gravelle, "Corporate Tax Incidence: Review of General Equilibrium Estimates and Analysis" Congressional Budget Office Working Paper 2010-03, 2010.

¹¹ Julie Anne Cronin et al. "U.S. Treasury Distributional Analysis Methodology," Office of Tax Analysis Working Paper, 2012.

¹² Jennifer C. Gravelle, "Corporate Tax Incidence: Review of General Equilibrium Estimates and Analysis" Congressional Budget Office Working Paper 2010-03, 2010.

¹³ Under the differential allocation model, there is an average worldwide tax on capital. Each country's deviation from this average represents a subsidy or a tax on corporate profits. This deviation from the worldwide average is the tax that is then allocated to capital and labor using the incidence assumptions from the traditional general equilibrium model. This approach emphasizes the importance of relative, rather than absolute, tax rates in the ability for capital to flow abroad in an attempt to avoid the burden of the corporate income tax. See Jennifer C. Gravelle, "Corporate Tax Incidence: Review of General Equilibrium Estimates and Analysis" Congressional Budget Office Working Paper 2010-03, 2010.

¹⁴ See Congressional Budget Office, The Incidence of the Corporate Income Tax, 1996, Congressional Budget Office, Effective Federal Tax Rates, 1979-1997, 2001, and Congressional Budget Office, Trends in the Distribution of Household Income between 1979 and 2007, 2011.

¹⁵ Congressional Budget Office, The Distribution of Household Income and Federal Taxes, 2008 and 2009, 2012.

¹⁶ Julie Anne Cronin, "U.S. Treasury Distributional Analysis Methodology," Office of Tax Analysis Working Paper, 1999.

¹⁷ Julie Anne Cronin et al. "U.S. Treasury Distributional Analysis Methodology," Office of Tax Analysis Working Paper, 2012.

¹⁸ The Joint Committee staff briefly distributed corporate taxes completely to capital owners in the early 1990s, partially reflecting the short-term aspect of the five-year budget window in effect at that time. See Joint Committee on Taxation, Estimated Budget Effects of the Finance Committee Chairman's Proposed Mark of the Revenue Reconciliation Provisions (JCX-7-93), June 17, 1993, and Joint Committee on Taxation, Methodology and Issues in Measuring Changes in the Distribution of Tax Burdens (JCS-7-93), June 14, 1993. However, the Joint Committee staff has not distributed corporate taxes since the replacement of the five-year budget window with a ten-year budget window.

¹⁹ Should corporate tax reform occur, the Joint Committee staff may revisit this assumption to allow the baseline distribution of corporate tax incidence to reflect the fact that substantial tax changes may take time to reach their steady-state equilibrium.

²⁰ Alan J. Auerbach, "Who Bears the Corporate Tax?" in James Poterba, ed., Tax Policy and the Economy Vol. 20, Cambridge, MA: MIT Press, 2006, pp. 1-40.

²¹ As an example of the uncertainties in analyzing the passthrough sector, while a C corporation may know fairly exactly its tax rate, a partnership needs to know the tax rates for each of its partners to determine the overall average tax rate on partnership income. This depends, among other things, on whether partners are affected by the alternative minimum tax or by the passive loss rules, which in turn depends on other income sources about which the partnership may know nothing.

²² For example, a proposal changing depreciation schedules may reach a long-run equilibrium more slowly than an immediate corporate tax rate change, as it takes time for equipment purchased under the old tax regime to be phased-out. Similarly, a firm with a multi-year collective bargaining agreement with its workers may be slower to renegotiate wages in response to a tax change than a firm that reassesses wage schedules on a more frequent basis.

²³The JCT Revenue Estimating Process, January 30, 2013 (available at https://www.jct.gov/publications.html?func=startdown&id=4500) provides more information about the individual tax model.

²⁴ The Joint Committee staff uses data from the Medical Expenditure Panel Survey, which includes information about income and health insurance benefits, to impute the value of employer health insurance to individuals on the individual tax model. Ideally, other forms of employee compensation, such as employer contributions to section 401(k) plans, pension plans, and other non-taxed benefits should also be included. However, data limitations prevent the incorporation of these additional income sources because they are not readily matched to individuals on the individual tax model. The exclusion of these additional elements of compensation may impact where in the income distribution labor's share of the corporate tax burden falls, to the extent that the distribution of excluded compensation elements differ from the distribution of those included. However, because labor's burden of the entire corporate tax is distributed to individuals based on observed compensation levels, the exclusion of these elements of compensation does not change the total amount or share of corporate taxes borne by labor for any proposed corporate tax change. Although there may be justifications for treating minimum wage workers differently in a distributional analysis of business taxes, because their wages cannot be adjusted downward in response to tax increases, data limitations make such a distinction impractical to implement.

²⁵ While this approach is appropriate for determining the ultimate owners, the indirect nature of the ownership level does impact the tax level since corporations only receive a 70-percent deduction for dividends received. Alan J. Auerbach, "Who Bears the Corporate Tax?" in James Poterba, ed., Tax Policy and the Economy Vol. 20, Cambridge, MA: MIT Press, 2006, pp. 1-40. Given the limitations of data for tracing all intermediate ownership links, the Joint Committee staff abstracts away from this concern when estimating the tax distribution.

²⁶ An alternative would be not to distribute tax-exempt organizations' share of corporate taxes. However, this requires assuming that the taxes borne by tax-exempt organizations have no effect on the resources available to domestic individuals. The Joint Committee staff sees this as a tenuous assumption. Given the limited information available about stakeholders of tax-exempt organizations, the Joint Committee staff distributes their share of corporate taxes consistent with other capital owners' share.

²⁷ Conversely, foreign corporate taxes of corporations abroad are borne in part by U.S. citizens. However, as the Joint Committee staff is only distributing U.S. taxes with the intent of understanding the distributional impact of U.S. legislative changes, no attempt is made to estimate or distribute foreign taxes to U.S. taxpayers.

²⁸ Department of the Treasury, the Federal Reserve Bank of New York, and the Board of Governors of the Federal Reserve System. Report on Foreign Portfolio Holdings of U.S. Securities as of June 30, 2011, 2012.

²⁹ The Joint Committee staff acknowledges that this may be a strong assumption. For example, John Graham and Alok Kumar, "Do Dividend Clienteles Exist? Evidence on Dividend Preference of Retail Investors" The Journal of Finance, 61, 2006, pp. 1305-1336, observe that high income individuals have a greater preference than low income individuals for stocks with low dividend yields. They attribute this result to the uneven tax treatment of dividends and capital gains. In an environment where capital gains are taxed at a lower rate than dividends, individuals in high marginal tax brackets prefer to receive returns on investments as capital gains rather than dividends. Such a dividend preference by income would bias our distribution of capital holdings downward in the income distribution. However, since 2003 capital gains and dividends have generally been taxed at the same marginal tax rate. As such, the concerns of tax preferences by income group should be mitigated, although the ability to defer recognition of capital gains may still result in such clienteles as Graham and Kumar describe.

³⁰ An alternate approach employed by other organizations, including the CBO, is to combine dividend receipts and realized capital gains. However, the Joint Committee staff believes such a hybrid approach likely introduces biases from the substantial disconnect between assets and realized capital gains that outweigh the additional information that capital gains holdings provide to the analysis. The inclusion of taxable realized capital gains results in a higher share of corporate taxes being borne by the top of the income distribution, which diverges from the distribution of equity-based taxable capital assets observed in either the asset sale data or the dividends data.

³¹ The end-of-year fair market value is determined for each year in the ten-year budget window based on an extrapolation of the reported amounts in IRS tax data using the Joint Committee staff individual tax model. The distribution of capital assets held in these accounts in each year for estimating the corporate income tax distribution are based directly on this extrapolation using the actual reported end-of-year values along with Joint Committee staff and CBO estimates of economic growth.

³² Joint Committee on Taxation, Overview of the Definition of Income Used by the Staff of the Joint Committee on Taxation in Distributional Analyses, (JCX-15-12), February 8, 2012, p. 3.

³³ Don Fullerton and Gilbert E. Metcalf, "Tax Incidence" in Alan J. Auerbach and Martin Feldstein, eds., Handbook of Public Economics, Vol 4, Amsterdam: North Holland, 2002.

³⁴ A small portion of the $10 billion corporate tax increase is not presented on the distribution table since taxpayers with negative expanded incomes bear a small portion of the tax burden but are excluded from the distribution results. As a result, the total tax distributed to domestic persons with positive expanded incomes and to foreign persons is below the $10 billion tax increase.

³⁵ Similar to that seen for the distribution of the corporate income tax, a small portion of the $10 billion tax increase on passthrough entities is not presented on the distribution table since taxpayers with negative expanded incomes bear a small portion of the tax burden but are excluded from the distribution results.