A recent report from the British Institute for Fiscal Studies argues: “There is a large, unjustified and problematic bias against employment and labour incomes and in favour of business ownership and capital incomes.” FT opinion columnist Martin Wolf concludes that “the tax treatment of returns to investment is a mess: incentives depend on the asset type, source of finance and legal structure, and range from large subsidies to large penalties. There is also a strong incentive to turn employees into self-employed and so expand the “gig economy”. These statements also apply to the United States.
Messy, inequitable, and inefficient taxation was also the subject of T. R. Reid’s A Fine Mess, which I reviewed previously. Reid makes a strong case for a broad based, low rate income tax structure (BBLR.) In a 2018 afterword to the book, Reid concludes that the Tax Cut and Job Act passed by the U.S. Congress in December 2017 “largely ignored the fundamental principle of BBLR”; thus, the “fine mess” got worse.
The motivation for this posting comes from two sources: the increasing divergence of taxes on labor relative to the taxes on capital and the wide array of different tax rate treatments of income based upon its source. The applicable tax structure for income from capital depends upon the organizational structure that generated the income (the form of the business), the funding source underpinning the capital (debt, equity, or retained earnings), specific features that do or do not qualify for special treatment (equipment, buildings, software, research and development), and where the official headquarters are located (e.g., USA, Ireland, Luxemburg, or the Bahamas.) Not surprisingly both horizontal and vertical equity principles are egregiously violated. (For those unfamiliar with these principles, see the afterward to this posting.)
Based on the work of the MIT Task Force on the Work of the Future, Daron Acemoglu and his colleagues observe that the average tax rate of labor in the U.S. has been about 25% over the last three decades while the average tax rate on software and equipment has fallen from 15% to about 5%. They argue that these rate differences encourage firms to substitute automation for labor rather than support technologies that complement labor, and thus, would increase labor productivity and compensation. One result is that labor’s share of national income in the U.S. has fallen from 70.2% in 2000 to 62.6% in 2019. (Bureau of Economic Analysis, FRED)
Income from capital arises in a variety of forms (as categorized by the National Income and Product Accounts): net interest, corporate profits, partnership rental income, and proprietor’s net receipts. Furthermore, there exists a variety of types of businesses that generate income from capital including sole proprietorships, limited liability partnerships, limited liability corporations, chapter S corporations (those with fewer than 100 shareholders), and chapter C corporations (those subject to the corporate income tax.) The applicable U.S. income tax structure varies by the type of business as well as whether the income is treated as earnings, interest payments, dividends, or capital gains. Ideally, analysis should differentiate among both the types of businesses and the forms of capital income; however, a reliable information source to conduct such an analysis proved too challenging for this blog post but should be central to any tax reform research agenda.
One can easily differentiate, however, among personal, corporate and social insurance (payroll tax) revenue sources. In addition to the above chart, the following table provides a breakdown of how these sources of federal revenue have changed between 2000 and 2019. In total, federal revenues have grown by 71.1% over the past two decades, and personal income taxes have been stable as a share of federal revenues. In contrast, payroll taxes have grown relative to income; while corporate income taxes have fallen markedly relative to corporate income. Clearly, such changes encourage the substitution of capital for labor.
|Personal Income||Corporate Income||Social Insurance Receipts|
In addition, between 2000 and 2019, real GDP only grew by 2% per year, much lower than the 3.5% annual average over the previous 40 years (bea.gov). So why should capital be taxed at a lower rate than labor?
My analysis closely followed that provided by Leonard Burman and Joel Slemod in the second edition (2018) of Taxes in America: What Everyone Needs to Know. Analysts cite at least three reasons (listed below) for the favorable tax treatment of income from capital gains. None hold up to serious scrutiny if overall economic well-being is the primary social objective; furthermore, the economic distortions generated by such treatment have yielded increased income equality and reduced income growth.
- The encouragement of risk taking and entrepreneurial activity
- Protection from inflation
- Encouragement of asset turnover
Capital gains accrue to all asset holders, not just those who have put in years of sweat equity and eventually generate income. In particular, the favorable tax treatment of carried interest, when labor compensation comes in the form of capital gains rather than as wages or salaries, seems to have no specific justification.
Inflation protection can best be obtained through indexing; that is, the purchase price can be linked to the consumer price index (CPI) or the personal consumption expenditure deflator (PCE.) Then the difference between the nominal value of the gain and the change in the price index would constitute the change in the purchasing value of the underlying asset, and, thus, be the appropriate income to tax.
Some analysts are concerned about the lock-in effects of a tax system that only applies when assets are sold. Unless the tax rate is quite high, however, the potential for delayed realization of the gain, and thus, lock-in, seems very limited.
Burman and Slemrod (page 46) conclude that “… first, lower capital gains tax rates encourage tax shelters, which are inefficient. Second, the vast majority of capital gains are realized by people with very high incomes. Tax breaks on capital gains disproportionately benefit the rich and undermine the progressivity of the tax system.”
How about the separate tax on corporate income? Since corporate income is taxed as it is generated and also when it is passed to shareholders as dividends or capital gains, should it be taxed twice (that is, both when the income is generated and when it accrues to households?) According to Burman and Slemrod (p. 65), 70% of these assets are held in non-taxable accounts (held by retirement funds and life insurance companies); so, such double taxation does not apply. Furthermore, corporate taxation is unrelated to the appropriate tax bracket for the underlying shareholders; thus, what matters is the total tax rate not the notion of two separate rates. Depending upon one’s state of residence wage and salary income could be taxed three times: as state and federal income as well as by the federal payroll tax. The number of times a source of income is taxed tells us absolutely nothing about either its equity or efficiency implications. To determine equity, one would need to know how the burden is distributed across people by income group and across income groups.
In addition to the equity effects of the corporate tax, economists argue that the tax can be shifted and, thus, who bears the eventual burden of the tax need not align with what legislators intend. If the tax could be levied purely on economic profits (that is, returns (rents) above the competitive return on capital), then the tax would be efficient as the profit maximizing level of production (thus output) would be unaffected by the tax rate since the profit maximizing level of production would be the same both before and after the tax. However, some of the tax could be shifted forward to consumers – assuming that consumers reduce consumption (but not completely) when the tax leads to higher prices. Secondly, a portion of the tax could be shifted onto labor as many laborers are employed by companies that pay the tax, and if the tax leads to reduced output, then it likely would lead to reduced employment. Thirdly, some of the tax might be shifted to non-corporate capital in the form of reduced returns as companies may choose to produce as chapter S corporations or partnerships rather than standard chapter C corporations. Finally, some of the tax might be shifted to foreign investors to the degree they continue to hold US corporate equity.
There are a variety of factors that affect the distribution of the burden of the corporate income tax, but that discussion goes beyond the purpose of this posting. It is worthwhile, however, to cite Burman and Slemrod’s conclusion (page 72): “The bottom line is that, alas, we simply do not know who will benefit from cutting the corporate tax rate.”
Given the many ways that capital income taxation generates both inequities and inefficiencies, I conclude that it should not be treated more favorably than labor income.
Afterward: Horizontal and Vertical Equity Principles
Horizontal equity requires that people in similar economic circumstances be taxed similarly. For example, two households with the same total income but from different sources should be taxed the same. Existent tax systems, which favor capital over labor income, violate this principle.
Vertical equity applies the ability to pay principle: those with higher income (from whatever source) should be taxed at least proportionately more than those with lower income. Payroll taxes in the United States violate this principle since the tax on wage and salary income goes to zero for income greater than $142,800 (2021), and since they do not apply to labor compensation taken as capital gains. Those who receive labor compensation as carried interest (almost exclusively high income recipients), avoid both the payroll tax and the higher tax rate applied to wage and salary income – clear violations of both horizontal and vertical equity.