Here’s the draft of a work in progress, shared with the permission of the author.
All of the Above: How to Tax the Wealthy
by Victor Thuronyi
Substantially raising taxes on the wealthy not only raises technical issues but is a profoundly political question. While there is broad popular support for the idea that the wealthy should pay substantially more tax, this can translate into law only through the political system. Republican members of Congress have nearly uniformly voted in the opposite direction: in favor of reducing taxes on the wealthy. Even many Democrats need to be convinced. This means that a prerequisite for such a major change to our tax laws is Democratic control of Congress as well as a conversation among Democrats. It is possible that the Senate could flip after the 2020 elections, although 2022 looks more likely. If the Senate remains in Republican control, it would be extremely unlikely for Congress to be interested in substantially raising taxes on the wealthy. So the proposals discussed here are not immediately
While various politicians have proposed one scheme or another, this paper concludes that effectively taxing the wealthy is best done through a combination of measures. Hence the title of this paper: “All of the Above.”
A start would involve repealing key aspects of the 2017 tax law, including the deduction for 20% of income of an incorporated business (a deduction the lion’s share of which goes to the wealthy), and raising the corporate tax to 28% (affecting the wealthy via their stock ownership).
The 2017 law virtually did away with itemizing deductions, except for wealthier taxpayers. It is proposed here to allow everyone to itemize, but to subject specific itemized deductions to floors. The $12,200 standard deduction ($24,400 for married) would be converted to a credit at a 24% rate. Charitable contributions would be subject to a floor of 2% of AGI and SALT (state and local taxes) to a floor of 5% of AGI. This would give space to states to maintain or increase high tax rates on the wealthy. At the margin, state taxes would be deductible. Interest expense would be subject to a 1% floor. Because this floor increases with income, this change would target the mortgage interest deduction away from the wealthy and more to families who need it.
The alternative minimum tax would be repealed in order to simplify the system, and most of its adjustments incorporated into the regular tax rules.
The estate and gift tax changes made in 2017 would be reversed; alternatively, the estate and gift taxes could be converted into an inheritance tax, with a lifetime exemption of $1 million.
A 74% rate bracket would be added on taxable income over $2 million ($2,400,000 for a married couple). Combined with other proposals, and assuming a (deductible) state tax of 15%, this would result in about a 90% tax. While this seems high, it would still allow someone earning millions of dollars who earns an extra million dollars to keep $100,000 of those extra earnings after tax. Alternatively, and perhaps preferably, the desired degree of progressivity could be achieved via an SET (see below).
A wealth tax would not be introduced, since virtually the same tax burden could be achieved by other proposals (mostly those taxing gains as they accrue). A wealth tax is problematic because of concerns about constitutionality and valuation.
A supplemental expenditure tax (SET) would be instituted at progressive rates, beginning at fairly low rates on expenditure over about $100,000 per person and rising to 50% for expenditure of the very wealthy. The tax would be calculated mostly on the basis of information already currently collected for income tax purposes, on a cash-flow basis. This tax is simpler to calculate than the income tax.
Capital gains and dividends would be taxed as ordinary income and accrued capital gains would be taxed at death or if property is transferred by gift (including transfers to charity). Other changes concerning investment income would include. eliminating preferences for qualified small business stock and opportunity zones, taxing inside build-up on life insurance policies, and taxing section 1031 (like kind) exchanges. Gains on marketable assets would be taxed as they accrue, and on other assets for taxpayers above a specified threshold a look-back approach would be used to make this tax equivalent to a tax paid at the time of accrual.
Funding for the IRS would be increased so that the IRS could hire sufficient personnel to audit wealthy taxpayers effectively, as well as providing guidance to all on how to comply with the tax laws.
Summary of Proposals
- Repeal the deduction for 20% of unincorporated business income (section 199A).
- Convert the standard deduction ($24,400 for married and $12,200 for single) into a credit at a 22% rate ($5,368 for married and $2,684 for single) that is available to all regardless of itemizing.
- Subject itemized deductions to floors: SALT (5% of AGI), charitable (2% of AGI); medical (7.5% of AGI); interest (1% of AGI); casualty and theft losses and miscellaneous itemized deductions: restore pre-2017 law.
- Repeal the Alternative Minimum Tax and make most of the AMT rules part of the regular tax.
- Repeal NIIT and cap medicare tax (only if top rates are raised).
- Convert the estate and gift tax into a tax on heirs with a lifetime exemption of $1 million; alternatively, restore the pre-2017 law estate tax threshold and tighten estate and gift tax loopholes.
- Increase the corporate tax rate to 28%.
- Add a 74% rate bracket on taxable income over $2 million ($2,400,000 for a married couple, which would be four times the threshold for the 37% rate). Alternatively, achieve the desired level of progressivity via the SET.
- Do not institute a wealth tax, since virtually the same tax burden can be achieved by taxing unrealized gains.
- Institute a supplemental expenditure tax (SET), which is a progressive tax on personal consumption, with a threshold of about $100,000 per person and a top rate of 50%.
- Tax capital gains and dividends as ordinary income.
- Tax accrued capital gains when property is transferred by gift or at death (including transfers to charities); tax gains on a mark-to-market basis for marketable assets and, in the case of the wealthy apply a look-back tax for other assets.
- Eliminate preferences for qualified small business stock and opportunity zones. Tax inside build-up on life insurance. Tax section 1031 exchanges.
- Increase funding for the IRS so as to enable more intensive auditing of wealthy taxpayers.
In recent months, a number of politicians have put forward ideas for how to tax the wealthy. These proposals need to be put into a political context. Republicans have control of the Senate. As long as that remains the case, any substantial tax increase on the wealthy will be highly unlikely. On the other hand, if Democrats are able to get even a small majority in the Senate, this could be enough for them to push through a tax increase on the wealthy, as long as there is a Democratic President (this could be done through budget reconciliation, which the Republicans used to pass the 2017 tax bill; although Democrats should hold hearings and public mark-ups, unlike the Republicans). While 2020 looks challenging for Democrats to take control of the Senate, there is a good chance for Democrats to take control of the Senate by the 2022 elections. So the question of how to tax the wealthy is not theoretical, but may take some time to be relevant to legislation that can actually get enacted.
Even if Democrats control both Houses of Congress and the Presidency, would they want to increase taxes on the wealthy as much as this paper recommends? There have been several periods of time in the past when Democrats did have this degree of control, and they have modestly increased taxes on the wealthy during some of these periods, but not nearly as much as is proposed here. Most people might want them to, but would there be enough political will to buck wealthy donors? The period of time between now and the achievement of Democratic control can be used to test whether Democrats can in fact come on board to a progressive tax agenda.
The attitude of the President is also critical. Tax legislation is inherently controversial, since it affects people directly and very specifically in their pocketbooks. Experience shows that substantial tax legislation can be enacted only if it is a top presidential priority. With many issues competing for the next President’s attention, much depends on how the President sets priorities and how successful they are in getting legislation enacted. And – perhaps stating the obvious – if President Trump is re-elected, he will have virtually no interest in substantially raising taxes on the wealthy.
As others have pointed out:
- Over the past several decades, concentrations of income and wealth have approached historically high levels. Particularly for those at the very top, wealth is generally due not so much to hard work alone as to good luck and social arrangements that are taken advantage of to extract wealth from others. This high concentration of wealth also corrodes the political system.
- Instead of addressing inequality of income and wealth by raising taxes on the wealthy, Republicans in Congress moved in the opposite direction in 2017 by enacting a fiscally irresponsible and poorly thought through tax giveaway for the wealthy.
- There are a number of options to better tax extreme wealth, including enacting an annual wealth tax, marking unrealized capital gains to market, taxing unrealized capital gains in the final income tax return of the decedent, strengthening the estate tax, taxing capital gains and dividends as ordinary income, and providing additional funding to the IRS to strengthen auditing of higher-income taxpayers.
While the wealthy pay a large share of income tax, this is because they hold a lion’s share of income and wealth; taking all taxes into account, the U.S. tax system is only mildly progressive, and is even regressive at the very top, with effective tax rates on the very wealthy actually lower than the rates for those near the top.
This paper argues that the best approach to taxing the wealthy is to use multiple instruments. Reliance on a single instrument to tax the wealthy is likely not to be very effective in reducing inequality, because the wealthy will plan around it (they can hire very smart people to do so). A combination of measures would be much more effective. Admittedly, there are compliance and administrative costs to using multiple measures to tax the wealthy, but these costs are not very high, nor would they have a significant negative economic impact. Some of the cost will involve providing additional resources for the IRS, but these are relatively small; the total budget for this agency is $12 billion, which is a pittance compared with the hundreds of billions spent on the Pentagon, and additional resources needed to tax the wealthy are only a small fraction of this budget. Some of the costs involve additional compliance costs for the wealthy; some additional accountant and legal time would be involved, but the ultra-wealthy can easily afford this. Moreover, a number of the changes I propose involve simplification, so there will be offsetting reduced compliance costs for many individuals and businesses.
The income tax is complex and many other changes besides the ones discussed here could usefully be made; I do not purport to cover all changes to the tax code that could involve a higher tax burden on the wealthy. I do note that there are a lot of provisions in the income tax designed to police the distinction between capital gains and ordinary income, and many of these could be eliminated if capital gains were taxed at the same rate as ordinary income and accrual taxation of marketable gains were implemented. But this article does not explore all the issues involved, which are very technical and would require an extensive exploration.
When I refer to the wealthy in this paper, I am talking about a fraction of the top one percent. As David Leonard has pointed out, the 90th to 99th percentiles in the income distribution have seen their incomes roughly keep up with growth of the economy. It is the top one percent who have benefited substantially more, and this is even more true of the top ten percent and one percent of the one percent. (If there are about 300 million people in the U.S., one percent is 3 million, and one percent of the one percent is 30,000. That group is the main focus, and to some extent it is even smaller, if we are talking about billionaires. At the same time, some of the measures discussed here, such as those concerning itemized deductions and taxation of accrued gains on publicly traded instruments, would have a broader scope.
- Rolling back 2017 tax changes
Qualified business income deduction.–One tax change made in 2017 involves both increased complexity and lower rates for the wealthy. This is the deduction for qualified business income (section 199A). This deduction should simply be eliminated. There is no good rationale for taxing income of an unincorporated business at lower rates.
Itemized deductions.—The large increase in the standard deduction enacted in 2017 should be reversed, and the relationship between standard and itemized deductions should be reworked. The conceptual basis of the standard deduction is straightforward: either you could itemize your deductions or (as a simplification measure), instead of keeping track of all your expenses, you could take the standard deduction, which was in effect an estimate of what your itemized deductions might have been. This rationale calls for setting the standard deduction at a level that approximates the itemized deductions for most people. There are other reasons for the standard deduction, such as making the system more progressive. Over time, the relationship between itemized deductions and the standard deduction changed, so that fewer and fewer people ended up itemizing. The 2017 law dramatically accelerated this trend by sharply increasing the standard deduction and thereby curtailing the number of itemizers (the number of itemizers was estimated to decline by more than half). Like everything else with this piece of legislation, there was no public input for this change since no hearings were held. If there had been input, presumably one of the things that would have been pointed out is that the change runs counter to the justifications for itemized deductions. For example, if the deduction for medical expenses is allowed on the basis that it makes the tax system fairer for those with high medical costs; it makes little sense to condition the deduction on factors that have nothing to do with medical costs, such as whether the taxpayer is a homeowner with a mortgage. Currently, someone with high medical costs but no other itemized deductions faces a substantial barrier to deducting those costs. Not only are the medical costs subject to a threshold of 10% of AGI, but for an individual taxpayer there is the additional need to exceed the $12,200 standard deduction. This leads to the bizarre situation that someone who owns a home with a mortgage and therefore itemizes deductions can effectively deduct much more of their medical expenses than a similarly situated renter. This makes little sense. Similarly, homeowners with a mortgage whose deductions for interest and property taxes exceed the standard deduction can deduct all their charitable contributions, while those who lack other itemized deductions effectively cannot deduct their charitable contributions. The current rules target the incentive to contribute to charity on the basis of arbitrary factors. The higher standard deduction also perversely favors deductions by wealthier taxpayers rather than those with lower incomes. For example, “less prosperous prospective homeowners are now less likely to benefit from the home mortgage interest deduction, since they are less likely to itemize”.
A solution to this arbitrariness would involve getting rid of the concept of the standard deduction. Keep the amount of the standard deduction, but give it to everyone. This could take the form of a per-taxpayer credit, equivalent to 22% of the standard deduction amount. The 22% rate was selected because the vast majority of taxpayers fall into the 22% bracket or lower, so this change from a deduction to a credit would be neutral or advantageous for most.
In addition, each itemized deduction would be given its own floor where appropriate. Floors are already currently in place for medical expenses, miscellaneous itemized deductions, and casualty losses. Student loan interest also has its own limitations and is deductible regardless of whether you itemize.
The big itemized deductions which are currently not subject to limitations are charitable contributions, state and local taxes (SALT), and interest (non student loan). For the charitable contributions deduction, I’d suggest a floor of 2 percent of AGI. For interest expense, maybe 1 percent of AGI, and for SALT I’d suggest 5 percent of AGI. (These are my initial rough guesses and can be refined on the basis of data analysis.) The fairly big floor for SALT is needed to minimize the revenue consequences and avoid undermining progressivity.
The numbers should be tweaked depending on what revenue estimators find when they run the numbers: how many itemizers (of various types) would we end up with? What does the distribution look like? Under this system someone could end up itemizing only one or two deductions but not others, since there would be a separate floor for each.
The overall effect of the proposal would be to benefit those who end up itemizing, although some people who itemize under current law might end up paying somewhat more tax than under current law because of the floors. This is why this idea should be regarded as a preliminary sketch rather than a fully worked out proposal, because the latter can be developed only by doing further quantitative analysis.
A key effect of the SALT cap is that at the margin state and local taxes are not deductible in most cases. This makes it difficult for taxing jurisdictions to raise or maintain their tax rates if they wish to do so to fund programs such as education and public infrastructure. If a deduction for SALT is allowed at the margin, state and local government have more fiscal space, particularly to increase taxes on the wealthy. State and local taxes would be recognized as reducing the capacity to pay federal income tax of those who pay high state taxes. If the AMT is also repealed (as proposed below), at the margin state and local taxes would be fully deductible, which was not even the case under prior law. This is important in terms of taxing the wealthy, because it would give fiscal space to States to increase taxes on the wealthy.
Under this proposal, more people would be itemizing charitable contributions and at the margin more deductions would be available. Given the large standard deduction, the average taxpayer currently gets no tax benefit from charitable contributions at all. Under the proposal, people would get a deduction if they make charitable contributions, but only if the charitable contributions are substantial. This would provide an incentive for the average person to make generous contributions. For the wealthy, instituting a floor on charitable contributions deductions based on AGI would make the system more progressive, and would incentivize the wealthy to make more generous donations. The proposal would rebalance the system, which is currently tilted severely in favor of charitable contributions by the wealthy.
Estate tax changes.—The 2017 law increased the estate tax threshold. This change should be eliminated. In addition, the estate tax should be strengthened or replaced with an inheritance tax, as explained in Section VII below.
Corporate tax rate.—Given that the wealthy disproportionally own corporate stock, raising the corporate rate will affect tax incidence on the wealthy. Some reduction from the previous level of corporate tax probably made sense, but the rate can be raised; 28 percent might be a reasonable rate.
International provisions.—I leave it to others to discuss the complex details of the changes to the international tax regime enacted in 2017. It should be possible to revise them to come up with a tighter system that will not provide incentives to locate investments overseas.
- Alternative Minimum Tax
The alternative minimum tax (AMT) is a complicated set of rules which would be unnecessary if other reforms proposed here were adopted; it should be repealed. Repealing the AMT would provide a simplification benefit. The AMT is designed to make sure that wealthy taxpayers pay at least a minimum amount of tax; this goal would become redundant if the reforms proposed here were adopted, since the wealthy would be taxed in other ways. One of the main AMT triggers has been the SALT deduction; if that deduction were restructured as proposed here, the AMT would not be needed. At the margin, SALT should be deductible and the AMT runs counter to that.
In other cases, the rules of the AMT should be made part of the regular tax rules and applicable to everyone. These include: mining exploration and development costs, percentage of completion for long-term contracts, the AMT interest expense rules, the incentive stock option rules, and the tax preference adjustments in sections 57 and 58 of the Internal Revenue Code.
A traditional way of looking at tax rates in economics has been rather abstract. High rates tend to discourage work effort and so there tends to be a trade off between economic efficiency and vertical equity which can be achieved by higher rates.
This way of looking at things, however, ignores the reality of what our economic structure has become. One of the aspects of the huge increase in inequality that we have seen since 1980 is the massive increase in compensation of higher level corporate executives. It is simply not normal for corporate executives to be paid at multiples of pay of average workers that we have now. Considering a typical historical ratio between CEO pay and average worker pay at about 20:1, multiplying an average salary of roughly $50,000 we come up with an income of around $1 million which might be considered as providing sufficient salary differentiation on the basis of historical experience. The implication is that any higher income is excessive and not needed to provide incentives or reward. There was a half-hearted effort to regulate executive compensation in the Internal Revenue Code, but it failed. It was supposed to encourage incentive-based compensation but it ended up backfiring and accelerating the tendency to higher compensation. The 2017 law repealed exceptions to this rule, thereby denying a deduction for executive compensation in excess of $1 million, and imposing a 21-percent excise tax on compensation over $1 million paid to executives of non-profits. It remains to be seen what the effects of these 2017 changes will be, but in any event they will not preclude ultra-high levels of compensation; they will only make this compensation a bit more expensive for those entities subject to these rules. While this change was one of the few positive aspects of the 2017 law, my guess is that its impact on amounts received by highly-compensated executives will be modest. Nothing prevents companies from absorbing the additional 21-percent tax. The high level of executive compensation is something that the rest of us pay for, in the form of high consumer prices. For example, one of the factors in the high price of drugs is the perceived need of pharmaceutical companies to charge high amounts for their drugs so that they can generate profits for shareholders and high executive compensation.
To change the system, rather dramatic changes in tax rates are needed, for example, a 70% tax rate on income over $1 million and 90% on income over $2 million. Looking at the tax returns for 2016, for those with total income between $1million and $1.5 million, the composition of the major items was: employee compensation 46%, capital gains 13% and partnerships and S corporations 27%, and for those with total income between $2 and 5 million, the composition was employee compensation 35%, capital gains 21% and partnerships and S corporations 31%. So executive compensation is a large chunk of the income of those in this income range (and capital gains are quite likely linked to that, because a lot of executive compensation takes the form of stock options and other benefits involving corporate stock), with partnership and other business income representing another big chunk. Just from these aggregate figures, we can’t tell how much of the business income is investment income of corporate executives and how much is the main source of wealth for the taxpayers concerned (e.g. those engaged in real estate or oil production).
While it would be possible to impose 70 and 90 percent tax rates just on executive compensation, this would involve problems of definition and anti-abuse rules to include various arrangements that clever taxpayers would devise. A broader and more effective approach, which also recognizes the legitimacy of taxing at a high rate very high amounts of income, no matter what the source, would be to simply add 70 and 90 percent rates to the rate schedule. Alternatively, it might be possible to achieve the desired degree of progressivity via the SET (see below).
Raising the rates in this manner would involve a number of issues. Even if corporate rates were raised to 28 percent, a big disparity between the top individual rates and the corporate rates might require anti-avoidance rules to deal with tax planning by using corporations. How much of a problem this is depends on what other changes are made. It is less of an issue if capital gains and dividends are taxed as ordinary income, gains are taxed on an accrual basis (with a catch-up tax for nonmarketable property) and capital gains at death are taxed.
Income tax rates would not need to be raised to 70 and 90 percent if other measures are taken, in order to reach effective rates at this level. For example, suppose that a taxpayer receiving compensation in excess of $1 million is subject to a marginal State income tax rate of 10%. Suppose also that an SET is enacted with a marginal (tax inclusive) rate for this taxpayer of 50%. If this taxpayer receives $100,000 of compensation above the $1 million threshold, the following taxes would be due:
State income tax $10,000
Federal income tax (assuming that the state income tax is deductible, with a threshold of 5% of AGI as proposed above, the state income tax deduction would be $5,000, and the tax under existing law would be 37 percent on taxable income of $95,000) $35,150
SET. If the individual uses the entire after-tax proceeds of $54,850 for personal consumption, the SET would be $27,425.
Total tax $72,575.
Therefore, if the SET is enacted with a top rate of 50%, there would be a rate in excess of 70% on compensation over $1 million for those living in a state with a marginal rate of 10% or above even if Federal tax rates were not increased from levels under existing law. This result would not necessarily be undermined if the individual did not use the proceeds for consumption, since consumption tax would be due eventually. Eventually, amounts that were accumulated might be subject to inheritance tax or SET at the level of the person inheriting the money.
To get to a 90 percent or so tax rate on executive compensation on amounts over $2 million, the top Federal income tax rate would have to be 74 percent (in the above example, this would generate State income tax of $10,000, Federal income tax of $70,300 and SET of $9,850, for a total tax of $90,150).
In terms of married vs. single taxpayers, the current top rate bracket kicks in at $500K for a single taxpayer and $600K for married taxpayers filing jointly. So for example if two unmarried individuals each have taxable income of $1m, then they each pay a 37% rate on $500K; in other words between them $1m is taxed at rates below the 37% rate. If they marry, then only $600,000 of their taxable income is taxed below the top rate of 37%. So there is a marriage penalty in this case. This could be carried over for the proposal. The proposed $2 million bracket might be for single people, and for married filing jointly the threshold might be set at $2.4m (four times the threshold level for the current 37% top rate). At the same time, a lower rate could be provided in the case of a couple each of whom have earned income. In other words, there should not be a substantial marriage penalty if two highly compensated persons marry. So, for example, if each member of a married couple earns $2 million, none of their income would be subject to the 74 percent tax bracket.
There is a proposal to impose a 10% surtax on adjusted gross income (AGI) over $2 million. It does not seem that a tax on AGI is a good idea. While some deductions are allowed in getting to AGI, there are numerous deductions that are legitimate which are not included in AGI. One example is investment interest. Someone with substantial investments that are debt-financed would pay tax on gross investment income under the proposed surtax, and that does not strike me as fair. There are other deductions which would apply to a small number of people, but that does not mean they should be denied. An example is a deduction under section 1341 for amounts repaid which were received under a claim of right. This deduction might apply to only a handful to people subject to the surtax, but I see no good reason to disallow it. The big ticket items are deductions for charitable contributions and SALT. My preference would be to allow both. In the case of SALT, allowing this deduction would provide the States will leeway to tax the wealthy. In the case of charitable contributions, allowing the deduction would benefit charities. I realize that this benefit would be skewed to charities favored by the wealthy but on the other hand another aspect of my proposal would democratize the charitable contribution deduction by making it available to all, not just those who itemize generally. I think there is a value in allowing taxpayers to deduct contributions to charities of their choice, instead of trying to control the results.
If top rates were raised substantially, it would be possible to repeal the Net investment Income Tax (NIIT) and to cap the Medicare tax, instead of trying to prevent avoidance of these taxes, which mostly ends up making the system more complex. Instead of imposing special taxes on a narrow base, it makes more sense to impose higher rates on a broad base.
- A wealth tax
Senator Elizabeth Warren has proposed a wealth tax on the super wealthy:
“The Ultra-Millionaire Tax taxes the wealth of the richest Americans. It applies only to households with a net worth of $50 million or more—roughly the wealthiest 75,000 households, or the top 0.1%. Households would pay an annual 2% tax on every dollar of net worth above $50 million and a 3% tax on every dollar of net worth above $1 billion. Because wealth is so concentrated, Saez and Zucman project that this small tax on roughly 75,000 households will bring in $2.75 trillion in revenue over a ten-year period.”
This proposal has been criticized on several grounds. First, the Supreme Court might strike it down as an unconstitutional direct tax. No one can predict what the Supreme Court would do. There is precedent either way, so it will come down to a judgment call. Given that the Court has a conservative majority, the possibility that they will strike the tax down cannot be dismissed.
Imputed income tax as alternative to wealth tax
Reconceptualizing the wealth tax as an income tax would involve integrating wealth- based taxation into the income tax by imposing a tax on imputed income for those with very large wealth. Congress could determine an appropriate level of imputed income for those with wealth above the specified threshold. This amount could be subject to tax under the usual income tax rates (which would include the higher rates discussed in section IV above).
To avoid double taxation, realized investment income should be subtracted from the imputed amount, with the result of taxing the higher of realized investment income or imputed income from wealth. In principle, it would be possible to tax imputed income only, and ignore realized income, but that would open up avoidance opportunities. A taxpayer might have a low basis in investment property and contribute to the increase in value of that property through entrepreneurial activity. In such a case, realized income might exceed imputed income, and it would be appropriate to tax the realized income, as we do under current law. Also, if it turns out that over a taxpayer’s lifetime, actual income from investments (measured by taking into account capital gains at death) exceeds the assumed 4% return, then it would be appropriate to tax people on the difference.
A challenge for the approach of taxing the greater of realized or imputed income is that the return on some assets is a mix of returns to capital and the taxpayer’s effort. This will be the case for any business income subject to self-employment tax, but could also more broadly include any businesses in which the taxpayer materially participates. There is no ideal solution, but one approach would be to segregate these assets and exclude them from operation of the imputed income rule, so that for such assets, tax would apply to realized income, as defined under current law.
One way of understanding why a tax on imputed income makes sense is to consider the analogy of endowment funds. Managers of charitable endowments routinely assume a return of 4 or 5 percent as a reasonable amount that they can spend from endowed funds each year without dipping into the capital of the endowment. This has worked over the long term. Of course, an alternative approach for a charity would be to spend the amount of actual income earned in the preceding year. But the problem with this is that actual returns measured on a relatively short one-year basis can be quite volatile, while a long-return of 4 or 5 percent, while not guaranteed, is much more stable. Congress could make a reasonable determination to tax the very wealthy on this basis under the income tax, and this approach should be constitutional, falling under the 16th amendment’s power to impose a tax on income. There is also precedent in Europe for thinking about a wealth tax as in nature of an income tax. It would not make sense to tax most people on this basis, for example because this might involve imposition of a tax in a year in which the taxpayer suffers an actual loss. However, for very wealthy persons, who do not need to finance their consumption out of their wealth, taxing them on a longer-term basis analogously to how endowment funds operate may make more sense.
Based on historic long-term returns on the S&P 500, a return of as high as 8-10% might be warranted, although this does include inflation. Further thought would have to be given as to an appropriate rate. A key question is whether this should be restricted to a real return or should include inflation.
Wealth tax vs. imputed income tax
The imputed income tax would be more immune from constitutional challenge than the wealth tax.
It would also be a bit simpler to administer, because some assets (namely business assets where the taxpayer materially participates) would be excluded from the scope of the tax. These assets are hard to value.
The imputed income tax would hew more closely to the income tax, rather than being a new tax on wealth, although in practice the effects and revenues could be similar, depending on how rates are set.
If we keep in mind that we already have a wealth tax, because we tax the inflationary portion of capital income (see discussion in section VIII below), then this might be considered to be enough, without the need for an additional wealth tax. The significance of this point depends on top income tax rates. If these are increased, then the inflation-based wealth tax, which already exists under current law, goes up.
An imputed income tax and the wealth tax would have very similar effects, except that the imputed income approach would tax relatively lightly those with substantial amounts of realized investment income, because the scheme would in effect credit these amounts.
Consider the following alternatives:
- A wealth tax
- A tax on imputed return
- A tax on accrued but unrealized gains.
The difference between these alternatives is not major. The wealth tax is the same as a tax on imputed income, except that since the wealth tax is an add-on tax, while the imputed income tax would replace the tax on realized income, the difference lies in that the imputed income tax would impose less of a burden than the wealth tax on those with relatively high amounts of realized income. Another way of putting it is that the two are equivalent in how they tax unrealized gains. Since unrealized gains are the lion’s share of the action for the very wealthy, the differences between the wealth tax and an imputed income tax are not dramatic.
Next consider the difference between an imputed income tax and a tax on unrealized gains. Here the difference is that the imputed income tax reaches an assumed long-term return, while the tax on unrealized gains reaches the actual market return. There does not seem to be a compelling argument in favor of the former; in fact, actual return is arguably more fair, since it looks at actual results.
The conclusion is that there is not much policy difference between taxing wealth and taxing unrealized gains. The two are very close. How to decide between the two? It turns out that even administration is not much different as between the two. In the case of marketable securities, there is no difference. In the case of other assets, the wealth tax requires annual valuation, while the mark-to-market tax involves a lookback tax at the time of realization. But the wealth tax could be structured the same way, namely in the case of assets that are hard to value, the wealth tax could involve a catch-up tax that would be imposed at the time of realization. So neither as a matter of overall policy nor as a practical matter is there much difference between a wealth tax and an unrealized income tax. The main difference is that the unrealized income tax is not subject to constitutional difficulties to the extent that the wealth tax is.
State-level wealth tax
Another option for a wealth tax would involve imposition of the tax at the state level. Roy Ulrich argues that States should not hesitate to impose a wealth tax out of fear that wealthy citizens would leave for another state. However, several states have struggled recently with the politics of taxing the wealthy, and there is undeniably a concern that wealthy citizens might leave. Moreover, no state has adopted a wealth tax, even though there is nothing holding a state back from doing so. Why should a State impose a wealth tax, if tacking a point or so onto the income tax rate is so much easier? If we are serious about a wealth tax at the state level, this needs to be encouraged with a federal credit. There is precedent for this approach – there was a credit for state death taxes. The credit need not be 100% but it should be for a substantial part of the state tax. (The credit could be given against federal income tax liability.) If a substantial credit were offered, a state would be foolish not to impose a wealth tax, since they could get revenue without imposing much of a burden on their taxpayers. Although this approach is a bit more convoluted than imposing a wealth tax at the federal level, it would not raise problems of constitutionality, because there is no constitutional prohibition on states imposing a wealth tax. It would also be possible to combine the two ideas of a federal wealth tax and a state wealth tax. The approach would be to enact a federal wealth tax but to provide that, if the tax were held unconstitutional, then the same legislation would grant a credit for a state wealth tax. States could in this case enact wealth taxes as a contingency matter, with the state tax coming into effect only if the Federal tax were held unconstitutional, and a state could in addition enact a state wealth tax that would piggyback onto the federal tax at a lower level. For example, suppose that the Federal rate were 2%. The Federal legislation could say that in the event that the Federal tax is held unconstitutional then a credit would be allowed for a state wealth tax up to the 2% level. A state could enact a 2% contingent tax that would come into effect only if the Federal tax failed. And if it wanted to, it could enact a state surtax, for example 10% of the federal wealth tax liability, which would be payable in any event. In this way, those states choosing to do so could collect some revenue from the wealthy right away, with the prospect of a revenue bonanza if the Supreme Court held the federal tax unconstitutional. As with extending Medicaid under the Affordable Care Act, a state could choose to sit out the opportunity, but then it would also lose out on revenue if the federal level tax were struck down by the Supreme Court.
A state wealth tax based on a federal credit sounds a bit convoluted, and it is, but there is precedent for this approach in the estate tax area.
This would have the consequence of the wealth tax revenues going to the States rather than the Federal government in the event of an adverse court ruling, but this may not be a bad thing given fiscal tightness in the States. Putting more resources in the hands of the states in this way could allow the federal government to cut back on other payments to the states and would allow states to use the revenues to fund education, infrastructure, and other important priorities.
A second objection is that a wealth tax is difficult to administer because it requires valuation of assets. While this is true, on the other hand the relatively small number of taxpayers will mean that audit coverage can be extensive enough to deal with valuation issues.
The concern about valuation is not trivial. European countries with a wealth tax have had problems with valuation. Some commentators have gone further and asserted that wealth taxes are unworkable, based on the European experience. In fact, although European countries have had trouble with the wealth tax, their experience does not show that the tax is unworkable. European wealth taxes have been quite different from the tax that Warren envisages. In Europe, the wealth taxes were designed to reach a substantial portion of the population. For example, Germany had over 1 million wealth tax payers in 1993. Wealth tax thresholds in Europe have been much lower than what Warren proposes. This affects the nature of the tax. The larger number of taxpayers means that administrative and compliance costs tend to be high. In addition, political pressure to create various exceptions, such as for small businesses, complicate the tax and lead to uneven valuation, which is what brought the tax down in Germany and other countries. Where valuation was more uniformly based on market value, as in Switzerland, the tax has been more successful. In addition, wealth tax rates in Europe have been fairly low; for example, in Switzerland they range from 0.13% to 0.94 percent. So European countries really have not tried a tax along the lines of that proposed by Warren, one with a very high exemption and also fairly high rates. Therefore, the European experience is of marginal relevance to evaluating the Warren proposal.
That said, one can draw a lesson from the European experience that uneven valuation (namely deviation from fair market value) undermines the fairness of the tax. Reducing the number of taxpayers from the 75,000 that Warren has proposed would help assure better audit focus so that uneven valuation would not be as much of an issue. This is where the SET could play a role. For those with wealth in the range of $50 million or below, one could expect consumption at a high enough level as a percentage of their wealth that the SET could raise amounts even higher than would be raised by the wealth tax at the proposed rates and exemption level. It should therefore be possible to raise the wealth tax exemption level above the proposed $50 million, and still raise the same level of revenue from the wealthy, if the wealth tax were combined with an SET. Valuation is not a problem under the SET, since the SET is based on consumption cash flow, not the value of business or investment assets.
There seems to be an assumption that if a wealth tax (or a tax on unrealized gains above a threshold) were imposed, we would be stuck with the same tax unit as is used for the income tax. But this is not the case. There are serious marriage penalty/marriage bonus problems with using the married couple as a unit. Consider the billionaire who marries someone with little wealth, and uses a pre-nup. Even though the billionaire shares little wealth with the new spouse, if the married couple were the unit, there would be a substantial tax savings. Instead, the tax unit for the wealth tax could be each individual. Sure, under this approach a billionaire could save tax by transferring a substantial amount to the billionaire’s spouse, but this would require a real division of wealth.
Similarly, if the tax took the form of a tax on unrealized gains rather than wealth, the tax could be imposed in respect of property owned by any individual who owns more than the specified threshold amount.
Trusts present another tax unit challenge. If a very high rate were imposed on those with income exceeding $2 million, there would be a problem with trusts because trusts could be used to avoid these high rates. It would be overbroad to impose the top rate on all trusts, since only a very limited number of individuals would be subject to this rate. A solution could be to tax at the top rate those trusts that are related to individuals subject to the top rate of tax. This could be by way of such an individual being a donor of the trust or a beneficiary. This would require writing related-party rules that go beyond the current-law rules treating individuals as owners of trusts in specified instances.
Imputed income on personal use property
One of the potential advantages of a wealth tax is that it reaches wealth that is invested in personal use property. This tax base can also be reached through the income tax. Those with wealth above a specified threshold could be taxed on imputed income from property available for personal use, such as housing, artwork, private planes, yachts, and the like. A return can be imputed on the cost (or fair market value) of such property.
For the vast majority of taxpayers, it would be an administrative burden to pay such a tax, which is why it should be restricted to a relatively small group. The tax unit can be determined as discussed immediately above, with property subject to this rule defined as property owned by, or available for use by, this person.
As Batchelder and Kamin point out: “Theoretically, an accrual tax could apply to qualified retirement accounts, tax-exempt bonds, primary residences, and charitable transfers over which the donor retains some control.” If a wealth tax is not adopted, it can be mimicked by taxing accrued gains also in qualified retirement accounts, but only for individuals whose wealth exceeds a specified threshold. Likewise, the trust rules could reach assets held by a donor-advised fund or private foundation.
- Taxing Capital Gains
Lower tax rates on capital gains allow the compensation of fund managers taking the form of “carried interest” to be taxed preferentially. Compensation of corporate executives also involves a lot of capital gains, since compensation in the form of stock grants and options is common.
Len Burman has advocated taxing capital gains the same as ordinary income, as have others. Such a regime was part of the Tax Reform Act of 1986, although that regime did not last and also failed to follow through and realize all the simplification benefits that could be gained by establishing uniform taxation.
Capital gains and dividends, which are currently eligible for preferential rates, are received overwhelmingly by the wealthy. Taxing capital gains and dividends as ordinary income would be an important element of taxing the wealthy more effectively as well as simplifying the system, since various rules policing the conversion of ordinary income into capital gains would not be needed.
One difficulty with doing this is that if income tax rates are high, then a lock-in effect results. Lock-in can be avoided by taxing gains on an accrual basis, i.e. without a realization requirement. For publicly traded stock, this is feasible, but it is more difficult for other assets like partnership interests and real estate. One approach for assets that are difficult to value is to use a lookback rule for these assets.
Senator Wyden has proposed a regime that would eliminate preferential rates for long-term capital gains and dividends, and, for taxpayers with income or assets above specified thresholds, would tax tradable property on a mark-to-market basis and use a lookback rule for gains on nontradable property. The thresholds are $1 million in income and $10 million in assets.
It may make more sense to apply the regime more universally, applying a threshold only for nonpublicly traded assets, on the theory that mark-to-market taxation for traded assets should not be too burdensome. For publicly traded assets, in respect of those with smaller amounts of wealth, most of these gains are realized in retirement savings accounts. So it is not clear why there is a big concern about taxing unrealized gains for those with smaller amounts of wealth: there is not likely to be so much of these gains, and administration is not so difficult since the reporting can come from financial institutions. As for non-publicly traded assets, the catch-up tax would come at realization or death and again the administration difficulty is not so great. An exemption could be provided for these, so as to reduce administrative burdens for those with smaller amounts of such gains. Therefore a system that is much more universal than the system proposed by Senator Wyden may in fact be less fraught, since there would not have to be such a focus on policing the threshold.
The Wyden proposal does not include inflation adjustment, or even discuss the issue. The absence of inflation adjustment is not necessarily a bad idea. If all capital income is taxed on the same basis, then the tax on the inflationary component of income is essentially a wealth tax. If inflation is low and stable and tax rates are not too high, this may make sense. On the other hand, if ordinary income rates were to be substantially raised and inflation were to rise or fluctuate, then this wealth tax might become problematic. Adjusting capital income for inflation would introduce some complexity, although this complexity is not overwhelming; Latin American countries have done this in the past, when they had very high rates of inflation.
Taxing capital gains on property transferred by gift or inheritance would be an important accompaniment to a higher tax rate on capital gains. This is because lock-in is exacerbated if gains are tax-free on assets held until death. In the case of capital gains at death, one could provide for an exemption amount so that those with a modest amount of assets would not have to calculate basis.
Gains on property transferred to charity should also be taxed on transfer. Under current law, there are various limitations on transfers of appreciated property. Eliminating this rule would result in an overall simplification.
Other tax preferences relating to capital gains should also be eliminated, including the exclusion of capital gains on qualified small business stock, and the tax benefits for investments in opportunity zones. Inside build-up on life insurance policies should also be taxed as it accrues. Gains on section 1031 exchanges should be taxed.
- Taxing Inherited Wealth
One approach to taxing inherited wealth would involve rolling back the changes to the estate tax exemption made by the 2017 law, as well as closing loopholes in the estate and gift taxes. The annual gift tax exclusion should also be tightened. Although the annual exclusion is not used nearly to the extent that it could be to avoid estate and gift tax, it can be taken advantage of by careful tax planners, and should be tightened up. The main reason for an annual exclusion is to avoid having to keep track of gifts, such as for weddings, birthdays, and holidays. This purpose is not served by making the annual exclusion available for a transfer to a custodian or a trust, as it is under current law. Valuation discounts for transfers to a family limited partnership are used to reduce the amount subject to estate or gift tax. The possibility of using such discounts should be reduced. Maybe the annual exclusion should be limited to gifts of cash or tangible personal property, and the amount reduced.
An alternative, and preferable approach would be to restructure the wealth transfer taxes as a tax on the recipient of the gift or inheritance, as Lily Batchelder has proposed. Following Batchelder, I would suggest a lifetime exemption of $1 million, and including inheritances as part of gross income. Batchelder suggests a surtax of 15%, but whether this is needed depends on what the income tax rates are and whether an SET is adopted. To avoid excessively high rates, inheritances could be averaged over the previous four years. I would reduce the annual exclusion for gifts to $10,000. As discussed above, no exclusion should be available for a transfer to a custodian or a trustee. Such transfers are clearly part of a transfer of financial wealth, rather than a typical birthday present. A transfer to a section 529 plan with a single beneficiary should be subject to tax, for example. As Batchelder described in her article, a withholding tax would apply to transfers to trusts where there is more than one potential beneficiary.
McQuaig and Brooks also favor an inheritance tax, shifting the rhetoric from a “death tax” to a tax on privilege. They propose a lifetime exemption of $1.5 million, with progressive rates and a top rate of 70% for inheritances over $25 million. They argue that an inheritance tax approach would increase the political prospects for taxing capital gains at death, since the tax on capital gains would fall on the decedent, while the inheritance tax would fall on the heir.
- A Supplemental Expenditure Tax
The debate in the academic literature on income vs. consumption taxation has focused on whether the income tax should be replaced by a consumption tax. Discussion centered on which tax was better. Further reflection on the mechanics and politics of replacing an income tax with a consumption tax suggests that this is a false choice. It is quite unlikely that the income tax would ever be replaced by a consumption tax. This is partly because the transition arrangements for such a replacement would be quite difficult to fashion and would be controversial, particularly those having to do with the corporate income tax. A more fruitful approach is to think of a consumption tax as being implemented as a supplement to the income tax. A supplemental expenditure tax (SET) then ends up not being an antithesis to the income tax, but rather a complement. The SET would allow the income tax to remain as a mainstay of a progressive tax system, imposed at progressive rates, but rates that are not unduly high.
The SET can be an important element in taxing the wealthy. Of course, the ultra wealthy (billionaires) do not consume large portions of their wealth. Therefore, the SET cannot be the only vehicle for taxing them. However, an SET at high progressive rates can be part of the picture. If one views the longer term, the ultra wealthy can only do three main things with their wealth: they can consume, they can transfer to charity, or they can transfer to relatives or other individuals during their lifetime or at death. The only one of these that escapes the SET is a transfer to charity. Once wealth is transferred to others, they will eventually consume, so that wealth will be subject to SET.
The SET discussed in this paper is a standard personal expenditure tax based on cash flow (i.e. cash receipts, with a deduction allowed for net investments). Such a tax would be vastly simpler than the current income tax. Taxpayers would determine their SET liability using the same information as for the regular income tax, with a few modifications. The general approach is cash flow. Thus, items of income are taken into account when received. A deduction is allowed for any investments when made. This includes business investments. However, some items are left out of account to simplify administration and compliance. Thus, certain borrowing would be excluded (primarily mortgage debt), a deduction would be allowed for net savings, and includable receipts would be somewhat broader than under the income tax. Personal deductions would generally be the same as under the regular income tax (see the discussion on itemized deductions).
Instead of replacing the income tax, the SET would be levied in addition to the income tax. It would entail a large personal exemption, and thereby be paid only by a relatively small segment of taxpayers.
Because an SET would be new, there would be concerns that defining the tax base might be difficult and confusing. These concerns are misplaced. It is feasible to define the SET base in a fairly simple way, and this tax would not impose an undue compliance burden, especially if it is structured with a large exemption so that it would apply only to the wealthy (say, the top one percent or so). The key to success in this respect is to set a fairly high threshold so that the great majority of taxpayers do not have to deal with this tax. The threshold might be set in the range of $100,000 to $200,000. Some additional enforcement resources should be allocated to the IRS so that the SET can be effectively audited and policed, for example in respect of consumption financed by liquidating assets located abroad. The design of an SET is discussed in detail in the appendix.
- IRS Funding
IRS funding has been dramatically cut back over the past several years and should be restored. While there are indications that funding could be increased by the current Congress, this would just bring the IRS back to the nominal funding level as of 2010. This should be adjusted for inflation and reflect increased IRS responsibilities for administering the Affordable Care Act, the 2017 tax Act, and to undergo deferred maintenance of computer systems.
The cut in funding has gone hand in hand with reduced audit levels for the wealthy. To the contrary, audit levels for wealthy individuals need to be increased. There is a substantial tax gap, i.e. the difference between the amount of tax actually paid and the amount that should be paid. Increased IRS auditing efforts should reduce the size of this gap, thereby increasing both revenues and tax fairness.
- Effective Dates and Transition
The economic effects of any tax change depend critically on the rules for bringing the new rules into force. While most tax legislation is enacted prospectively, i.e. applicable to taxable years beginning after the date of enactment, it can have an effect on existing wealth. In the case of measures designed to tax the wealthy more heavily, the policy intention is to impose a burden on existing wealth holdings. However, care should be paid to the transition rules in order to avoid unfairness as between taxpayers in similar positions. This is the principle of horizontal equity.
Perhaps the most controversial issue will be transition rules for taxing unrealized gains. I argue above that taxing unrealized gains is close enough to a wealth tax to be a preferable alternative, given concerns about constitutionality. However, the similarity between a tax on unrealized gains and a wealth tax depends on what transition rules apply. If pre-enactment gains are exempted, then a wealth tax has quite a different effect than a tax on unrealized gains, because much of existing wealth would be exempted from the new scheme. So if the intention is to reach existing wealth, it is important not to provide an exemption for pre-enactment gains. Instead, those gains could be taxed in installments, in order to avoid a huge up-front tax. For example, pre-enactment gains could be taxed over 10 years. If income tax rates were increased as suggested above, with a 74-percent rate on income over $2 million, the tax on pre-enactment gains of the wealthy would be substantial.
As far as the SET is concerned, no relief for existing wealth holdings is warranted, as the intention is to reach consumption out of wealth existing on the date of enactment. Therefore general transition rules are not needed; the appendix discusses some specific transition rules that might be appropriate for housing and consumer durables.
New rules on taxing capital gains the same as ordinary income and taxing gains at death generally do not require transition rules. Some relief might be contemplated for those who die shortly after the effective date and who have accumulated large gains over a lifetime. They would face a much larger tax bill than if they had died a bit earlier, which seems arbitrary. This could be dealt with by providing an exemption which phases out over time. The case for such an exemption might be especially compelling if new high rates are enacted and if the gains cause the taxpayer to fall into that higher rate bracket. Another option to deal with this would be averaging, under which a large gain in the taxpayer’s final year could be spread out over several years for purpose of determining the applicable rate of tax.
The reverse problem involves taxpayers who realize capital gains at current low rates in order to avoid paying tax at the new higher rates. Again, this becomes especially problematic if much higher top tax rates are proposed. One option would be to apply the new rates to gains realized after the date that the proposal was made, which might be the date that the President submits the proposal in the budget document.
Similarly, if the estate tax is replaced with an inheritance tax, there will be an incentive to make gifts before the effective date at least to use up the unified credit remaining to the taxpayer, which under current law can be substantial. Transition rules can subject such large gifts made after the announcement date to the inheritance tax.
Although the subtitle of this paper is “all of the above”, my conclusion is that, among the proposals discussed in this paper, the one that may not be needed is the wealth tax, given that very close to the same tax liability can be achieved by taxing unrealized gains, taxing capital gains as ordinary income, and raising the top rate. In addition, the SET functions as a wealth tax to some extent, since it reaches consumption out of wealth. The SET could reach important items like political contributions. If the combination of proposals discussed here is enacted, it would be virtually impossible for the wealthy to avoid paying substantially more tax, even if they engage in some tax planning. This combination could be more effective than the wealth tax and on the whole easier to administer and less subject to constitutional challenge.
This paper discusses a number of proposals. While they would ideally all be enacted, the reality is that they may not all be, even if a decision is made to raise taxes on the wealthy. Some of the proposals go together, though. One of these is to tax capital gains as ordinary income. This would be desirable from an equity standpoint, and would also make the tax laws simpler and harder to avoid. For example, instead of enacting special rules to tax carried interests as ordinary income, if preferential rates for capital gains were removed, carried interests would automatically be treated the same as ordinary income. However, it would be problematic to treat capital gains as ordinary income without also dealing with gains at death and preferably taxing accrued gains as well. Otherwise, there would be a lock-in problem. So all the proposals on capital gains go together: eliminating preferential rates for long-term capital gains, taxing accrued gains on marketable assets, a catch-up tax on other property, and taxing gains on property transferred by gift or at death. This set of policies is what I would prioritize.
Further progressivity could be achieved by enacting an SET or increasing the tax rates at the top. I would prioritize the SET because it is a way of making the system more progressive without the additional pressure that higher tax rates at the top would involve. An example of this pressure is the difference in tax rates on corporations and individuals. If the corporate rate were 28 percent and the top individual rate were much higher, individuals subject to the top rate would be tempted to use the corporate form. In principle, the problem would be mitigated by a catch-up tax on realized gains, assuming that such a tax is enacted and imposed broadly enough. The SET would raise no such concerns, because that tax would not be due until the taxpayer incurs consumption expenses, so there would be no incentive to use the corporate form to avoid the SET. Achieving greater progressivity at the top via the SET, together with other measures advocated here, might be preferable to imposing very high marginal income tax rates, given the incentives to avoid tax that those very high rates would induce.
Appendix: Details of Supplemental Expenditure Tax (SET) Design
- In General
The SET would be a cash-flow tax (i.e., cash receipts, with a deduction allowed for net investments). Such a tax would be substantially simpler than the current income tax, although it does involve some design issues and new elements as outlined below.
In a broad sense, the SET is very similar to the income tax, except that includable receipts are defined more broadly than under the income tax and the SET does not tax income until the income is consumed (generally, investment is deductible).
Taxpayers would determine their SET liability using the same information as for the regular income tax, with a few modifications. The general approach is cash-flow. Thus, items of income are taken into account when received. A deduction is allowed for any investments when made. This includes both financial and business investments. Borrowing would generally be included in taxable receipts, a deduction would be allowed for net savings, and includable receipts would be somewhat broader than under the income tax.
Even though the tax base is personal expenditure, the base would be legally defined as income less specified deductions.
Appropriate levels for the SET exemption, or for the rates, will depend on the whole tax policy picture. The SET rates and exemption can be determined at the end of the process of designing a tax reform bill so as to attain the desired distributional and revenue results. In very general terms, however, I would envisage setting the SET exemption at a rather high level (probably between $100,000 and $200,000), so that only a small percentage of taxpayers would pay this tax.
The SET would be a return-based tax that would be implemented with an additional schedule on an income tax return and would be administered as part of the income tax.
As with the income tax, the SET would apply to individuals who are citizens or residents of the U.S. It would not apply to corporations or partnerships. Distributions from those entities would, however, be included in the SET tax base of the distributees. This would not require particularly complex calculations, since all that is needed is the amount of cash distributions, an amount that is relevant for regular income tax purposes as well.
Income would be defined much as under the regular income tax. Thus, income would include wages, interest, dividends, royalties, and the like. A major difference from the income tax is that the proceeds of sales are fully taxed when received, i.e., it is not just the gain that is includable in taxable receipts but the entire sales proceeds. These are accounted for on a cash basis, so, for example, installment sales would be taxed as cash is actually received, not at the time the sale takes place.
Fringe benefits present much the same issues under an expenditure tax as under the income tax. Accordingly, one could expect the same solutions, i.e., if a particular item is taxed as a fringe benefit under the income tax, it would be taxed in the same way under the SET. An example would be use of an employer-provided automobile.
In general, special deductions would be the same as for the income tax.
A deduction would be allowed for life insurance premiums, whether for term insurance or insurance that has an investment component. In other words, all life insurance would be treated like savings. The reason is to avoid having to make distinctions among different kinds of life insurance policies. (All life insurance has a certain degree of investment value.)
Correspondingly, life insurance payouts would be taxable to the beneficiary of the policy when received.
As a general rule, all borrowing proceeds are included as taxable receipts and a deduction is allowed for interest and principal paid on loans. If the borrowing proceeds are used for investment, an offsetting deduction is allowed.
The only exception to the general rule is for home mortgages, auto loans, and loans for other consumer durable items, purchased with debt secured by the item. In these cases, the taxpayer will be taxed on amounts used to pay off the loan, because no deduction would be allowed for principal or interest payments made. This approach times taxation closer to actual consumption of the consumer durable.
Forgiveness of loans the proceeds of which were included in taxable receipts would not be taxed. In contrast, forgiveness of mortgage debt, auto loans, and loans to acquire consumer durables would be taxed (since the loan proceeds were not taxed). Note, however, that there is unlikely to be much SET liability by reason of taxation of loan forgiveness, given the high threshold. Most people in a position to have their consumer debt forgiven will not be subject to the SET in the first place because of the threshold. If the amount of loan forgiven is large, then provision could be made in the law for spreading the taxable amount over several years, in order to allow taxpayers to make use of the threshold. Otherwise, a taxpayer generally below the threshold might get bumped up into being taxable in the year when a large mortgage loan is forgiven.
In principle, cash could be tracked, but the simpler approach is not to do so. This means, for example, that when an investment asset is liquidated and cash proceeds are obtained, the cash is included in the SET base at the time of receipt as personal expenditure. The particular time that the taxpayer uses the cash to pay for consumption items is irrelevant.
The suggested treatment of cash will allow taxpayers to engage in a certain amount of self-help averaging. If a taxpayer wants to increase the SET base for a year, this can be done by liquidating an investment and receiving cash. By contrast, a transfer of cash into an investment account will reduce the tax base for that year. Checking accounts can be treated the same as cash, so there would be no need to report balances in these accounts, or additions to or withdrawals from them. In addition to this legitimate averaging opportunity from liquidating investments, there is an opportunity to evade tax by failing to declare cash receipts and then transferring the cash into an investment account. The result would be a reduction of the tax base, beyond what could be accomplished under the income tax by failing to declare the cash income. Manoeuvres like this should raise a red flag for audit, but highlight that audit capacity does need to be there in order for the tax to succeed. In the case of investments held in a brokerage account, there would be no evasion opportunity, because sales would be reported to the IRS. (There would be no additional reporting requirement, as these transactions are already reported.)
To understand the treatment of owner-occupied housing under the SET, consider first the typical case of a home that is mortgage-financed. A purchase money mortgage used to buy a residence would be left out of debt account (in other words, the borrowing proceeds are not taxable and repayments are not deductible). In the case of someone buying a home with cash or putting up a substantial down payment, it would be unfair to treat the entire amount as expenditure for SET purposes in the year that the house is purchased. (Bunching all this expenditure into one year would tend to place the taxpayer into a higher tax bracket than usual.) The remedy is to allow the taxpayer to amortize the expenditure over some lengthy period, say twenty to thirty years (interest should be charged on the outstanding balance; in effect, the taxpayer would be put on the same footing as if a mortgage had been used.). The taxpayer should be allowed to notionally pay off all or part of the outstanding balance at any time, thereby including this amount in the SET base. This would put the taxpayer on a similar footing to someone who financed with a mortgage, who could achieve this tax result by paying off all or part of the mortgage. It would be advantageous for a taxpayer to do this in any year where there is an unused exemption amount under the SET.
If a taxpayer purchases a consumer durable, the transaction does not lead to a substantial amount of consumption for the year in an economic sense, given that annual consumption should include only the value of use of the durable for the year in question, not the entire value of the durable. However, from a legal point of view, absent a special rule, the entire purchase price is part of taxable expenditure for the year, because the SET treats the entire consumption as occurring in the year of purchase. As with housing, the case of consumer durables purchased with debt can be dealt with by ignoring the debt-financed part of the transaction. The debt could be excluded from receipts, with no deduction allowed for loan repayments. The result will be that loan repayments will be taxed as consumption as they are made. This is the same rule as applies for debt-financed owner-occupied housing. As with housing, one could also amortize the cost of substantial (such as those costing over $10,000) consumer durables purchased with cash over a period of years.
An argument can be made for averaging under an SET. In a number of situations the taxpayer may incur substantial expenses for reasons largely beyond the taxpayer’s control. These may be items such as medical expenses, legal fees, or tuition. If these expenses cause taxable expenditure to be higher than normal, the tax consequence may be considered unfair. An averaging rule could address this concern. Such a rule would, however, introduce complexity to the system. The complexity would involve both definitional issues as well as administrative burdens for both taxpayers and the tax administration in keeping track of carryovers from one year to the next. The added complexity of an averaging rule could be minimized by limiting the rule to expenditures that are quite large as a portion of taxable expenditure. The simplest approach would be to include no averaging rule.
The specific form of SET proposed raises an averaging problem that is somewhat different from that under a broader consumed-income tax. The large annual threshold (probably on the order of $100,000 to $200,000) means that taxpayers with relatively low amounts of consumption in a given year “waste” that year’s exemption. This could be dealt with by allowing taxpayers to file the information on expenditure with their return even if they are not subject to the SET for the year in question, and carry over the unused exemption. Although this would involve a recordkeeping burden, it is not major, particularly for taxpayers with relatively simple financial affairs. If this option were not allowed, taxpayers would have the incentive to accelerate consumption into low-expenditure years (e.g., by purchasing consumer durables rather than investments), and this distortion would not make sense as a matter of policy. Administration of the rule might be simplified by limiting the amount that could be carried over, as well as limiting the period of time for the carryover (otherwise, returns that are many years old might have to be audited in the tax year when the carryover is used, at which point much of the applicable information might no longer be available).
Rules would be needed to avoid double taxation in the case of citizens or residents who earn amounts from foreign sources and pay foreign tax. This can be done either by allowing a credit against SET for foreign income tax-paid, or exempting from the SET base amounts of consumption that are financed by foreign-source income. Assuming the foreign tax credit in its current form, for purposes of the foreign tax credit, the SET should be considered as part of the income tax for purposes of the foreign tax credit limitation. Under this approach, the foreign tax credit does not pose any difficulties for the SET. Admittedly, the result ends up being rough and ready, particularly where there is a substantial amount of taxable expenditure financed out of income of previous years. The foreign tax credit limitation formula does not take into account whether that previous income was domestic or foreign-source, and what tax rates it bore. Likewise, if the taxpayer incurs foreign income tax but saves a substantial portion of the current year’s income, with the result that the current year’s domestic tax is low, the formula reduces the foreign tax credit available. To calculate the foreign tax credit more precisely in a way that coordinated the different approaches of foreign and domestic tax law would, however, introduce needless complexity.
An individual subject to SET at a high marginal rate might have a tax incentive to retire abroad, if intending to continue at a high consumption level (or if the individual engaged in a high level of savings during the individual’s earning years, which the individual intends to consume during retirement). It may be appropriate to provide rules requiring expatriating individuals to continue to pay SET for a period of years, particularly where the amounts involved are substantial. The policy issues are similar to those for an exit tax under the income tax, and one would expect the SET rules to track the exit tax rules for the income tax.
I assume that gifts and bequests will not be taxed to the donor. In other words, they will not be treated as part of the donor’s consumption. The donor would accordingly receive a deduction for a cash gift. This – combined with a generous exemption – creates an obvious tax avoidance opportunity. A wealthy individual could transfer assets to his or her children, who could use them to purchase consumption goods and services. To cut off this opportunity, it would make sense to include in the SET base of the parent any consumption expenditures by minor children. In most cases, there will not be anything to report, because most children typically do not liquidate financial assets in order to pay for their consumption. The proposed rule would not apply once the child attains majority. For this situation, an anti-avoidance rule would probably be needed providing that a purported gift will be disregarded to the extent that the gift is used to provide a consumption benefit to the donor. This rule will of course not catch everyone; this weakness in the expenditure tax represents a good argument why we don’t want to rely on an expenditure tax exclusively, and why it is a good strategy to combine the SET with the income tax. Even if the donor could avoid SET by making gifts, manoeuvres of this kind would not avoid income tax.
Assuming that a deduction is allowed for gifts, rules will be needed to police the boundaries of this deduction. For example, gifts to corporations and other entities that do not qualify for the charitable deduction under the income tax should not be deductible. Importantly, this would include political contributions. The resulting inclusion of political contributions in the donor’s tax base is a strength of the SET. The only deductible gifts should be those made to individuals. Even these need to be restricted, since one would not want to allow deductions for gifts to a politician, or gifts to a person who provides services to the taxpayer.
12. Housing and Other Personal Use Property
In the longer term, the principal residence, if purchased after the effective date, will be entirely tax-paid. To the extent purchased with cash, the cost will be included in the SET base in the year of expenditure (or would be spread over several years; see above discussion of averaging for special rules that might be provided). To the extent financed with debt, no SET deduction would be allowed for repayments of principal or interest. Upon sale, the entire proceeds should be exempted (see discussion of transition in §1.04below for treatment of housing purchased before the effective date). The same treatment should apply for sales of other personal-use property.
Rules will be needed to deal with property that is purchased with a mixed personal use and investment purpose. This kind of property consists of either real estate or personal property such as antiques, collectibles, and art. A simple but tough rule would be to treat all such property as consumption expenditure. On disposition, the simplest rule would be to exempt the proceeds from tax. Any tax on the gain is excessive as a matter of consumption tax principles (except to the extent attributable to sweat equity).
This approach will require treating as personal expenditure property any property that is in fact used for personal purposes, as well as property held for investment or used in a business if it constitutes fine art, a collectible, or an antique. Real estate (such as a vacation home) that is available for use by the taxpayer or members of the immediate family (spouse or children 18 and under) would be treated in the same way as personal expenditure property. Proceeds from rental should be exempted. So, for example, if a vacation home is purchased partly with cash and partly with debt, the debt would be excluded (as consumption debt) and the cash payment for the home included in the SET base. Suppose that rental income is used to pay interest on the debt, property taxes, repairs, and so forth. All of these amounts would simply be ignored for SET tax purposes. There would be no need for the taxpayer to keep track of the number of rental days or the amount of rental income received. In other words, there would be a (possibly modified) yield exemption treatment for this kind of asset. (The same principle can be applied to artwork, a yacht, race horse, or other similar property which is treated as personal expenditure property: any income from renting the property can be ignored.)
One fairly obvious anti-abuse rule would be to provide that a purported gift to a third party will be disregarded to the extent that the gift is used to provide a consumption benefit to the donor.
Another abuse situation would consist of the purchase of a yacht, car, airplane, real property, or similar item by a corporation, trust, or other entity. The corporation might be owned at least in part by the potential user of the property, who might then lease it from the corporation. If the user of the property had bought it himself, the expenditure would have been part of the SET base. In principle, it would be possible to police the amount of rental charged, but this is unlikely to be effective because of potential disputes about the fair value of the rental, particularly in situations where the property is also rented to others for part of the time. A possible anti-abuse rule would impute to the user the purchase of personal-use property by a corporation or other entity (including an individual acting as an accommodation party). The purchase amount could be included in the expenditure tax base of the user in the year of purchase. An exception would be made for bona fide rentals by publicly held companies (e.g., if someone rents a car from a company engaged in automobile leasing). While the suggested anti-abuse rule would be harsh, this would be justified because there would be little bona fide non-tax reason for entering into such an arrangement. The existence of a tough anti-abuse rule of this kind should stamp out these kinds of transactions, with the result that it will not be necessary to actually apply the rule very often.
If the existing income tax were completely replaced by an expenditure tax, then there would be a need for transition relief. The classic case is that of the taxpayer who has saved up during a working life and is just about to retire when the expenditure tax is introduced. Suppose that the taxpayer’s savings are in high-basis assets. If the income tax continued, the taxpayer could draw down these assets without additional tax. However, under an expenditure tax, this taxpayer would face paying tax again. This situation would call for giving the taxpayer relief for consumption financed out of tax-paid assets.
In the case of the SET, however, the transition situation is somewhat different. The SET is designed to be an additional tax, imposed in addition to the regular income tax (hence it is called a supplemental expenditure tax). Its incidence could be intended to fall partly on existing wealth, and partly on wealth accumulated after the effective date, to the extent that either is consumed. This seems fair. The income tax continues. Taxpayers holding wealth at the time of introduction of the SET will benefit from any reduction of income tax rates that take place at the same time that the SET is introduced. The taxpayer in the above example would not pay any more income tax on assets that are liquidated to finance consumption. The SET payable would therefore not be in duplication of income tax already paid. Imposing a one-time tax burden on existing capital would, in other words, be part of the politically accepted strategy.
While general transition relief should therefore not be needed, a few specific transition rules will be required to avoid unfairness in particular cases.
One such rule involves consumer durables, particularly housing. In the case of someone buying a house after the effective date with borrowed funds, there would be no particular problem. Given that the loan would be kept out of account, the result is that interest and principal on the loan would be included in the tax base as the loan is repaid. This would provide an advantage to those who already own housing, but the advantage would be limited: no deduction for interest on existing housing would be available for SET purposes. The unfairness would apply to those who have saved up but not yet purchased a house as of the effective date. If no transition rule were provided, they would be seriously disadvantaged in comparison with someone who had purchased a house with cash just before the effective date of the SET. To address this, an exemption could be provided (subject to an appropriate limitation) for the purchase of a principal residence within a specified period (e.g., one year) after the effective date in the case of someone who does not own such a residence. (The exemption would apply only to amounts paid in cash. Any amounts in excess of the exemption limit would be eligible for averaging via amortization of the purchase price as explained above.)
A special rule will also be needed for disposals of the principal residence after the effective date, in the case of a residence purchased before the effective date. Assume that under current law, gain on the disposition of the principal residence is excluded. Suppose someone sells a principal residence qualifying for the gain exclusion after the effective date. This could apply for SET purposes as well.
Consideration should also be given to a transition rule for those who, before the effective date, purchased an unusually large house. Such individuals would be advantaged as compared with those who buy housing with income earned after the effective date, since the latter would be taxed on these amounts. If no account is taken of the existing asset as of the effective date, there will be an undue preference for these individuals. Accordingly it would make sense to include in the SET base an estimated rental value in the case of homes worth more than a specified amount. I recognize that this would involve some valuation issues, and it would be possible to get along without this rule, but some such rule would seem to be appropriate as a matter of fairness.
Apart from amounts invested in a principal residence (subject to a possible limitation as discussed above), the SET would constitute a levy on existing capital. The burden of this tax would, however, depend on the taxpayer’s consumption choices: it would become due only for consumption at a luxury level. As long as the taxpayer (or the taxpayer’s heirs) spent at or below the level represented by the SET exemption, no tax would be due.
Initial cash balances as of the effective date would be taxed (with an appropriate de minimis exclusion). Cash balances for this purpose would include whatever type of checking account and other bank balances that are treated the same as cash (i.e., not taken into account) for SET purposes generally.
not without difficulty, the above transition rules are far more modest than the
transition rules that would likely be required if the existing income tax were
completely replaced by a consumed-income tax. The difficulty of transition is
often cited as one of the principal problems of a cash-flow tax. The SET would largely avoid these problems.
 Republican-held seats in the swing states of Pennsylvania, Wisconsin, Florida, North Carolina, Georgia, Iowa, and Missouri will be up for election in 2022, and Democrats will have to defend no Senate seat in a state won by Trump. On the other hand, if there is a Democratic President, there is a tendency for an incumbent President to lose seats in a mid-term.
 See, e.g., Lily Batchelder and David Kamin, Taxing the Rich: Issues and Options (Sept. 11, 2019); Alexandra Thornton and Galen Hendricks, Ending Special Tax Treatment for the Very Wealthy (Center for American Progress, June 4, 2019) https://www.americanprogress.org/issues/economy/reports/2019/06/04/470621/ending-special-tax-treatment-wealthy/
 See Linda McQuaig and Neil Brooks, Billionaires’ Ball (2012).
 See id. (McQuaig and Brooks).
 See Kimberly Clausing, Fixing the Five Flaws of the Tax Cuts and Jobs Act (May 28, 2019) (the deficit-financed TCJ tax cuts make it more difficult to respond to the next recession, worsen inter-general equity, and make it more difficult to pay for urgent priorities such as social security and healthcare, and investing in green energy, education, and infrastructure).
 See Michael Graetz, Foreword—The 2017 Tax Cuts: How Polarized PoliticsProduced Precarious Policy, Yale Law Journal Forum, Oct. 25, 2018.
 See McQuaig and Brooks, at 209-210.
 See David Leonhardt, How the Upper Middle Class in America is Really Doing, New York Times, Feb. 25, 2019.
 See Daniel Shaviro, Evaluating the New US Pass-Through Rules,  British Tax Rev. 49; David Kamin et al., The Games They Will Play, 103 Minn. L. Rev. 1439, 1473-74 (2019).
 Martin McMahon, Erwin Griswold Lecture, 71 Tax Lawyer 421 (2018) comes to similar conclusions. Alan Viard, An Economic Analysis of the TCJA’s Larger Standard Deduction, Tax Notes, Apr. 1, 2019, at 79, also comes to the conclusion that the large increase in the standard deduction makes no policy sense.
 See Kimberly Clausing, supra note 5, at note 54.
 Kimberly Clausing, supra note 5.
 For example, Kimberly Clausing suggests a rate of 28 percent.
 See Chuck Collins, Is inequality in America irreversible? 7 (2018).
 Internal Revenue Code section 162(m).
 See, Internal Revenue Service, Statistics of Income (available on IRS website).
 See David Kamin et al., The Games They Will Play, 103 Minn. L. Rev. 1439 (2019).
 Josh Bivens, Restraining the power of the rich with a 10 percent surtax on incomes over $2 million (Economic Policy Institute 2019).
 Website of Warren for President.
 See Victor Thuronyi, Commentary: The European Experience with a Wealth tax, 53 Tax L. Rev. 693, 694 (2000); Moris Lehner, The European Experience with a Wealth Tax, at 649-52, 680-81.
 See Roy Ulrich, A Wealth Tax for the States (2015) (available at SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2552968).
 Internal Revenue Code section 2011, repealed in 2014.
 For example, David Brooks asserts: “”The Nordic countries tried wealth taxes of the sort Elizabeth Warren is proposing, and all except Norway abandoned them because they were unworkable.” New York Times, A19 (July 2, 2019).
 See Moris Lehner, The European Experience with a Wealth Tax, 53 Tax L. Rev. 615 (2000).
 Germany estimates administrative and compliance costs at EUR 150 million each, i.e. 4% each of tax revenue. See Lehner, supra note 24.
 See Lehner, supra note 24.
 See Batchelder and Kamin, supra note 2.
 See McQuaig and Brooks, 165-173.
 See Len Burman, The Labyrinth of Capital Gains Tax Policy (1999).
 Senator Ron Wyden, Treat Wealth Like Wages (Senate Finance Committee: Sept. 2019). Senator Cory Booker has endorsed essentially the same plan: https://corybooker.com/issues/economic-security-and-opportunity/opportunity-and-justice-for-workers/
 See Victor Thuronyi, Inflation Adjustment, in Tax Law Design and Drafting (International Monetary Fund: 1996).
 See Richard Schmalbeck, Avoiding Federal Wealth Transfer Taxes, in Rethinking Estate and Gift Taxation (William Gale, James Hines, and Joel Slemrod eds. 2001), at 120, 121..
 See Richard Schmalbeck, Avoiding Federal Wealth Transfer Taxes, in Rethinking Estate and Gift Taxation (William Gale, James Hines, and Joel Slemrod eds. 2001), at 120-23.
 See Schmalbeck at 132-36.
 Lily Bathclder, What Should Society Expect from Heirs? The Case for a Comprehensive Inheritance Tax, 63 Tax Law Review 1 (2010).
 See Lily Bathclder, “Taxing Privilege More Effectively: Replacing the Estate Tax With an Inheritance Tax,” discussion paper, Hamilton Project, Brookings Institution, June 2007.
 See McQuaig and Brooks at 239.
 See McQuaig and Brooks, 256 n. 38.
 Compare Niels Johannesen et al., Taxing Hidden Wealth, for a discussion of the challenges in including foreign bank accounts in the tax net. This study shows that there are still significant amounts of unreported foreign bank accounts held by U.S. citizens, particularly the wealthy.
 See Joint Committee on Taxation, Overview of the Tax Gap (JCX-19-19, May 8, 2019).
 See Batchelder supra at 21-26 for a detailed discussion of the choice between a wealth tax and accrued income tax.
.Graetz, supra note xx, at 1611-13.
. Including a second home, as well as collectibles and the like described in below.
. Compare Kaldor (1955) who proposed including in the expenditure tax base the annual rental charge on housing. The approach suggested here differs from Kaldor’s in that it applies only to pre-effective-date housing, and applies with a threshold, so that only more expensive houses are affected. The value of a second home should also be included in the calculation.