In this chapter we discuss the leading types of consumption taxes—the retail sales tax, the value-added tax, the flat tax, the Bradford X tax, and the personal expenditures tax—and explain our preference for the X tax. Part of this discussion is drawn from Viard (2011a, 185–95).
The retail sales tax is the type of consumption tax that is most familiar to Americans, and the VAT is the type with the most widespread international use.
Retail Sales Tax. A retail sales tax is easily recognized as a consumption tax because, in its pure textbook form, it is imposed only on retail sales to consumers. Sales from one business firm to another are excluded from the tax base. Although a general sales tax has never been used at the federal level in the United States, forty-five states and the District of Columbia currently impose sales taxes, as do many local governments. Unfortunately, as we will discuss in chapter 10, most state and local sales taxes diverge significantly from this pure textbook design, exempting significant amounts of consumer purchases while taxing a substantial amount of sales between business firms.
The retail sales tax is a real-based tax, which means that it applies only to the sale of real goods and services and does not tax, or provide deductions for, financial transactions such as loans and stock purchases. Also, firms do not deduct payments of wages or other employee compensation in computing their sales tax base.
The discussion of the sales tax offers a useful opportunity to clarify the distinction between the tax-exclusive and tax-inclusive methods of quoting tax rates. For example, consider a sales tax system that imposes a $20 tax on a consumer good that costs $80 before tax so that the total price paid by the consumer is $100. The tax-exclusive rate is 25 percent, because the $20 tax is 25 percent of the $80 net-of-tax price. This method of quoting the tax rate is called tax-exclusive because the tax payment is excluded from the tax base to which the rate is applied; here, the $20 tax payment is excluded from the $80 tax base to which the 25 percent rate is applied. The tax-inclusive rate in this example is 20 percent because the $20 tax is 20 percent of the $100 total price. Under the tax-inclusive method, the tax payment is included in the base to which the rate is applied; here, the $20 tax payment is included in the $100 base to which the 20 percent rate is applied.
Sales tax rates are usually quoted in tax-exclusive form, and income tax rates are usually quoted in tax-inclusive form. (An individual who pays $20 tax on $100 before-tax income is usually viewed as paying a 20 percent tax rate, not a 25 percent rate on her $80 after-tax income.) Because the tax-exclusive rate is always higher than the corresponding tax-inclusive rate,3 this practice artificially makes sales tax rates look higher than income tax rates. Nevertheless, deviating from common practice by quoting sales tax rates in tax-inclusive form can cause confusion.
At the federal level, the most prominent sales tax proposal is the FairTax plan, put forward by Americans for Fair Taxation. The plan would replace the individual and corporate income taxes, payroll and self-employment taxes, and the estate and gift tax with a retail sales tax featuring a 29.87 percent tax-exclusive (23 percent tax-inclusive) rate. The FairTax plan has repeatedly been introduced in Congress but has never emerged from committee. The FairTax bills in the 112th Congress, H.R. 25 and S. 13, have attracted sixty-seven sponsors in the House of Representatives and nine sponsors in the Senate.
Value-Added Tax. The VAT can be viewed as a modification of the sales tax. Durner, Bui, and Sedon (2009) report that more than 145 countries have a VAT; they tabulate each country's tax rate and the year in which the tax was introduced.
A VAT, like a retail sales tax, applies to goods and services sold to consumers. But unlike a retail sales tax, which is collected once on the final sale to a consumer, a VAT is imposed and collected at every stage in the production and distribution of a good or service. This collection structure helps prevent the tax from being evaded at the retail level.
Like the retail sales tax, the VAT is a real-based tax that disregards financial transactions. Also like the sales tax, it does not allow a deduction for wages or other employee compensation. Like sales tax rates, VAT rates are usually quoted in tax-exclusive form.
For simplicity, the discussion below focuses on a subtraction-method VAT. (As we will explain in chapter 10, most actual VATs use the distinct, but highly similar, credit-invoice method.) Under a subtraction-method VAT, the tax base for each firm is receipts from sales of real goods and services minus purchases of real goods and services, including capital goods, from other firms. Sales minus purchases measures the firm's valued added, which is the contribution of the firm to the overall value of output.
For the economy as a whole, the base of a VAT is sales of real goods and services to consumers, because sales from one business to another are subject to offsetting inclusion and deduction and therefore do not comprise part of the net tax base. So, the aggregate VAT tax base is equal to consumption, which is also the aggregate sales tax base.
Figure 2-1 illustrates the relationship of the VAT to the sales tax in an economy with two firms and two individuals. Firm A produces a machine that it sells for $100 to Firm B and pays $70 of wages to Jones, its employee. Firm B buys the machine for $100, pays $40 of wages to its employee Smith, and produces $150 of consumer goods. Jones buys $90 of the consumer goods, and Smith buys the remaining $60.
Under a retail sales tax, tax is collected from firm B on the $150 of consumer goods that it sells. Under a VAT, tax is collected from firm A on the sale of the $100 machine and from firm B on its $50 value added ($150 sales to consumers minus $100 machine purchase). Because the sale of the machine nets out, the VAT has the same $150 aggregate tax base as the sales tax.
The Regressivity Problem. As we will discuss in chapter 10, a number of recent proposals call for the adoption of a VAT alongside the income tax. Except for the FairTax plan mentioned above, however, policy makers have shown little interest in completely replacing the income tax with a sales tax or VAT.
The most important reason for this lack of interest is that such complete replacement would result in politically unacceptable regressivity. Without any modification, the sales tax or VAT is regressive because all consumers pay tax equal to a fixed fraction of their consumer spending. The importance of the regressivity concern has been amplified by the rise in economic inequality during the last few decades; the Congressional Budget Office (2011c) reports that the top 20 percent of households received 59.9 percent of national income before taxes and transfers in 2007, up from 49.6 percent in 1979.
The regressivity problem can be addressed to some extent by providing rebates (as the FairTax plan does), expanding transfer payments, or exempting necessities while taxing luxuries at higher rates. Unfortunately, these solutions fall short of fully addressing the problem. Rebates and transfer payments provide only a limited offset to regressivity, leaving a politically unacceptable share of the fiscal burden on the middle class. Taxing different goods at different rates creates complexity and economic inefficiency and also provides only a limited offset to regressivity.
Accordingly, any sales tax or VAT that is adopted is likely to serve as only a partial replacement of the income tax, with an income tax retained to provide progressivity. Given our interest in complete-replacement options, we set the sales tax and VAT aside for most of this book, although we will discuss in chapter 10 how a partial-replacement VAT could be designed if complete replacement cannot be achieved.
We therefore turn to more progressive consumption taxes, beginning with the flat tax and the X tax. These taxes, which differ only in their rate schedules, both use the same tax base, which is obtained by splitting the VAT tax base, value added, into two parts. Before discussing the rate schedules of the flat tax and the X tax, we discuss the two-part VAT tax base design that both taxes share, explaining why, despite its initial appearance, this design yields a consumption tax rather than an income tax.
To alleviate the regressivity of the VAT, the Hall-Rabushka flat tax and the X tax split the VAT base, value added, into two components, wages and business cash flow, and tax them separately.
Household Wage Tax and Business Cash-Flow Tax. Robert Hall and Alvin Rabushka (1983) proposed to split the VAT into two taxes, one imposed on business firms and the other imposed on households. Business firms compute value added, as they would under a subtraction-method VAT, but deduct their wage payments. The resulting tax base is called business cash flow. Households are taxed on their wages but not on their investment income. The total tax base is the same as under a VAT and therefore the same as under a retail sales tax; the only difference from a VAT is that wages are taxed to workers rather than to firms.
Throughout this book, we will refer to the components of the two-part VAT as the "household wage tax" and the "business cash-flow tax." The tax rates under both components of the two-part VAT are generally quoted in tax-inclusive form, as income tax rates are usually quoted. This practice makes the tax rates look artificially lower than sales tax and conventional VAT rates, but does not lower the true level of the rates.
Figure 2-1 shows the application of the two-part VAT in the simple economy. Firm A is taxed on its $30 cash flow ($100 value added minus $70 wage payment), and Firm B is taxed on its $10 cash flow ($50 value added minus $40 wage payment). Smith and Jones are taxed on their wages.
Although its aggregate tax base is equal to national consumption, the two-part VAT looks more like an income tax than a consumption tax, at least at first glance. After all, it includes a household tax on wages, a major component of income, and it does not appear to include any tax on consumers. Adding to the confusion, the name of the two-part VAT's first incarnation, the "flat tax," offers no clue that it is a consumption tax. Many early descriptions of the flat tax were ambiguous about the nature of the tax; Zelenak (1999, 1180–82) describes how the status of the flat tax as a consumption tax has been obscured in the popular debate. Further confusion occurs when supporters of a flat-rate income tax refer to that quite different levy as a flat tax or even, as in Laffer (2010), as a "true flat tax." The name of the two-part VAT's later incarnation, our preferred "X tax," avoids misleading income-tax connotations, but also fails to reveal that it is a consumption tax.
No Tax Penalty on Saving. One way to see the proper classification of the two-part VAT is to introduce it into the Patient-Impatient example that we used in chapter 1. If wages and business cash flow are both taxed at 20 percent, do Patient and Impatient ultimately receive the same treatment that they experienced under the 20 percent consumption tax, or do they instead receive the treatment that they experienced under a 20 percent income tax?
Little information can be gleaned by considering Impatient. Under both the income tax and consumption tax considered in chapter 1, he paid $20 tax on his wages and consumed the remaining $80. The two-part VAT also yields that same result, because Impatient pays $20 tax on his wages. Moreover, his consumption results in no firm-level tax being collected. If we imagine a firm paying workers $80 to produce the consumption goods sold to Impatient, the firm is taxed on the $80 sale proceeds and deducts the $80 wage payment, leaving no net tax.
The litmus test is the treatment of Patient. Does the 20 percent two-part VAT cause her to consume $144 tomorrow (the income-tax result) or $160 (the consumption-tax result)? Under the two-part VAT, Patient pays $20 tax on her wages today and lends the remaining $80 to the firm. So far, this looks like an income tax, because Patient has paid tax on her wages, even though she has not yet consumed anything. The picture changes, though, as we walk through the remaining steps.
With the $80 obtained from Patient, the firm buys a machine with a before-tax cost of $100. Under the two-part VAT, a $100 machine costs only $80, because the firm immediately deducts the $100 cost of the machine under the business cash-flow tax, reaping a $20 tax savings. If the firm produces the machine itself, it deducts the $100 wages paid to its workers; if the firm buys the machine from another firm, it deducts the $100 purchase cost.
The $100 machine yields a $200 payoff tomorrow, on which the firm pays $40 of business cash-flow tax. With the remaining $160, the firm pays Patient her $80 principal plus an $80 return. Patient then consumes the entire $160; she makes no tax payment because investment income is exempt from the household tax.
As with the direct consumption tax considered in chapter 1, Patient bears only a 20 percent burden, the same as the burden on Impatient, and is not penalized for her decision to save. Because the defining characteristic of the consumption tax is that it does not penalize saving, this outcome confirms that the two-part VAT is a consumption tax. It is easy to see why the household component of the two-part VAT does not penalize saving, as the tax applies only to wages and exempts income from saving. The question of why the firm-level component of the two-part VAT, the business cash-flow tax, does not penalize saving is a little more complicated, with the answer turning on the fact that this tax allows investment to be expensed.
The Importance of Expensing. Although a tax on a firm's net income, such as the current corporate income tax, penalizes saving, a tax on the firm's business cash flow does not. The difference between the two approaches is the manner in which the firm deducts the costs of its business investments. Under a net income tax, the firm depreciates the cost of the investment over its useful life; under a business cash-flow tax, as under a VAT, the firm immediately deducts, or expenses, the investment costs. This subtle difference has dramatic implications.
Under a 20 percent firm-level net income tax, the firm would have been able to buy only an $80 machine with Patient's $80 of savings because there would be no tax deduction at the time of purchase. When the machine yields $160 tomorrow, the firm would deduct the full $80 cost as depreciation because the machine would then be worthless. The firm would pay $16 tax on its $80 net income, leaving only $144 to be paid over to Patient for her to consume. A 20 percent firm-level tax on net income therefore penalizes saving in the same way and to the same extent as the 20 percent individual income tax we considered in chapter 1.
The two-part VAT avoids that outcome because it taxes firms on business cash flow rather than net income, allowing firms to expense their investments up front. Under the business cash-flow tax, the firm immediately deducts the cost of the $100 machine and reaps an immediate $20 tax saving, which is why it can buy a $100 machine with only $80 of Patient's funds. When the machine delivers its $200 payoff, the entire cash inflow is taxable, so the firm pays $40 tax, leaving $160 to be paid to Patient.
Although the investment triggers a $40 tax on the firm tomorrow, it also triggers a $20 tax saving for the firm today, which is precisely the effect that saving had on Patient's tax liability under the direct consumption tax discussed in chapter 1. Here, as there, the immediate tax savings and the subsequent tax payment cancel out in present value, leaving no net tax penalty on investment.
This neutrality result applies to expensing in general. By definition, the future cash flows from a marginal investment have a present discounted value equal to the cost of the investment. Provided that the tax rate remains constant (always a prerequisite for the neutrality of consumption taxation), the tax on the marginal investment's future cash flows has the same present value as the tax savings from the up-front expensing deduction. As a result, the marginal effective tax rate on saving and investment is zero.
Is the Tax Paid by Consumers? Although the two-part VAT has an aggregate tax base equal to national consumption and does not penalize saving, readers may still resist the conclusion that it is a consumption tax. After all, it looks like a tax on workers and firms rather than a tax on consumers. In figure 2-1, the sales tax and VAT, although remitted by the firms, are universally understood to ultimately be paid by the final consumers (Firm B's customers), with Jones ultimately taxed on $90 and Smith on $60. The two-part VAT has the same $150 aggregate tax base, but the allocation of tax payments across individuals and firms looks completely different.
The Patient-Impatient example features a similar discrepancy in the timing of payments under the two-part VAT. Today, Impatient is taxed on $100, Patient is taxed on $100, and the firm deducts $100, leaving a net aggregate tax base of $100, which equals today's aggregate consumption, all of which is done by Impatient. Tomorrow, the firm is taxed on $160 with no tax on Patient and Impatient, yielding an aggregate tax base of $160, which equals tomorrow's aggregate consumption, all of which is done by Patient. As in figure 2-1, the tax base is always equal to consumption in the aggregate, but the allocation of tax payments across individuals and firms does not line up with the allocation of consumer spending. Splitting the VAT in two appears to completely change who pays the tax, breaking any link between tax payments and consumption.
Economic theory reveals, however, that these differences in allocation are illusory. According to long-standing principles of public finance economics, the tax burden on a transaction depends on the combined tax of the parties to the transaction, not on the tax liability of either party. Moreover, the manner in which the two parties share the real economic burden of the tax does not depend on the amount of tax liability legally assigned to each party. In the Patient-Impatient example, the two-part VAT imposes zero tax on Patient's saving, just as any good consumption tax should. Yes, Patient pays tax on the $100 that she saves, but the firm to whom she lends her saving claims an offsetting $100 deduction. Moreover, the two-part VAT taxes the $160 consumed by Patient tomorrow, just as a consumption tax should. Although no tax is imposed on Patient, the full $160 is taxed to the firm that pays her the money that she consumes.
A True Consumption Tax. Appearances notwithstanding, the two-part VAT is a consumption tax. It has an aggregate tax base equal to national consumption and, assuming that the tax rate remains constant over time, it imposes a zero effective marginal tax rate on saving and therefore does not penalize saving. It also imposes a combined tax on each transaction equal to that imposed by a conventional VAT, although the legal allocation of the payment between parties to the transaction is different than under a conventional VAT.
Recall that the combination of two key features makes the two-part VAT a consumption tax. First, the household tax applies only to wages. All returns to saving, including interest, dividends, and capital gains, are exempt from the household tax. Second, firms are allowed to immediately expense, rather than depreciate, their investments. Both of these features are crucial to the design of the tax.
The above discussion establishes that a two-part VAT with a 20 percent tax on households' wages and a 20 percent tax on business cash flow is economically equivalent to a 20 percent conventional VAT. Because the conventional VAT is far easier to explain, though, there seems to be little to recommend the two-part VAT. Indeed, if the goal is to tax all consumption at the same proportional rate, the conventional VAT is the way to go.
The purpose of splitting the VAT in two, however, is precisely to tax different components of consumption at different rates, something that cannot be done under the conventional VAT. As explained below, both incarnations of the two-part VAT—the flat tax and the X tax—tax business cash flow more heavily than wages and tax different workers' wages at different rates, an arrangement that promotes progressivity.
Under the flat tax proposed by Hall and Rabushka (1983), firms are taxed at a single flat rate—say, 25 percent—on business cash flow. Workers are taxed at that same rate on wages, but only above a substantial exemption amount. This exemption amount ensures some degree of progressivity across workers and lowers the overall tax rate on wages, relative to the rate on business cash flow. Ironically, this makes the "flat" tax more progressive, and hence less flat, than a conventional VAT or sales tax, casting additional doubt on the utility of its name.
To further promote progressivity, Bradford (1986, 81–82) proposed that the Hall-Rabushka flat tax be modified to feature a full set of graduated rates for the household wage tax. Under this approach, the tax rate on business cash flow is relatively high (as explained in chapter 3, we will use an illustrative value of 38.8 percent throughout this book), and workers with the highest wages pay a marginal tax rate equal to that rate. But workers with lower earnings face lower rates, and those below the exemption amount continue to pay no tax. If desired, refundable tax credits can be provided to low-wage workers. Bradford (1988) referred to this modified approach as an "X tax."
Taxing higher-wage workers at higher rates clearly allows the flat tax and X tax to promote progressivity in ways that the conventional VAT does not. The other distinctive feature of the flat tax and X tax, though, is that business cash flow is taxed more heavily than wages. This is true even under the flat tax, which allows a fixed exemption amount under the wage tax but not under the business cash-flow tax. The X tax goes further, taxing all business cash flow at the top rate, a rate that applies to the wages of only the highest-paid workers. As explained below, this heavier taxation of business cash flow further promotes progressivity, because the burden of the cashflow tax largely falls on well-off households.
Role of the Business Cash-Flow Tax. In the current form of the Patient-Impatient example, the business cash-flow tax is irrelevant, and only the 20 percent household wage tax matters. Patient would have the same $144 consumption tomorrow if the 20 percent business cash-flow tax did not exist. As previously discussed, thanks to the expensing of investment, the business cash-flow tax results in a $20 tax saving when Patient saves today and a $40 tax payment when Patient consumes her investment proceeds tomorrow. The government enjoys no net revenue gain because the $40 tax payment it collects tomorrow is the same as what it could have obtained by investing the $20 tax payment forgone today at the 100 percent marginal return available in the economy. As discussed above, this zero-present-value feature is precisely why the business cash-flow tax does not penalize saving. Indeed, there would still be no net revenue gain and no saving penalty if the cashflow tax were imposed at a 50 or 99 percent rate, as the tax would then yield a $50 or $99 tax saving today, followed by a $100 or $198 tax tomorrow.
But the actual economy differs in important ways from this example. Under more realistic assumptions, the business cash-flow tax raises some revenue from saving, even though it still imposes no penalty on new saving on the margin. This revenue arises from two sources, existing capital that is in place when the tax is introduced and above-normal returns.
The first source of revenue is associated with the introduction of the tax. In the above example, suppose that there is no tax of any kind today and that the business cash-flow tax is unexpectedly introduced between today and tomorrow. Impatient consumes $100 today and is unaffected by the subsequent introduction of the tax. Patient saves $100 today, expecting to consume $200 tomorrow, but ultimately consumes only $160 due to the unexpected $40 tax liability.
The unexpected tax leaves Patient with only a 60 percent after-tax rate of return on her investment, which is actually lower than the 80 percent return she would have cleared under the income tax and far below the 100 percent before-tax rate of return. The problem is one of timing. If the business cash-flow tax is in place all along, Patient's decision to save gives her a $20 tax savings today, offset by a $40 tax tomorrow (all at the firm level, of course). But with the business cash-flow tax introduced in midstream, Patient gets the worst of both worlds. She pays the $40 tax tomorrow, but does not receive the $20 tax savings today because the business cash-flow tax is not in effect today to provide those savings.
So, the business cash-flow tax raises revenue from the savings already put in place before it is adopted and imposes an unexpected penalty on that saving. But this penalty applies only to savings that Patient has already done. The tax does not impose a penalty on new saving and therefore does not discourage future saving, unless it creates fears of future unannounced levies on savings already in place. The unexpected burden on past saving does raise fairness and other concerns, which we will discuss in chapter 8. Although most consumption tax proposals respond to these concerns by offering some transition relief, none of them, including our proposal, offer sufficiently generous relief to completely eliminate the tax on capital that is in place when the reform is introduced.
Second, some investments yield above-normal returns, also called rents or pure profits. In the above example, Patient's investment was assumed to be in a machine that yielded the 100 percent marginal rate of return. But some machines may yield a rate of return higher than the marginal machine, perhaps due to innovation or the exercise of market power. Suppose that Patient has access to a specific investment opportunity with a 120 percent return, a machine that yields $220 tomorrow. Under the business cashflow tax, her decision to invest triggers a $44 tax payment tomorrow. If she had not saved, the government would have collected $20 tax today, which could have been invested (at the 100 percent marginal yield available on additional investment) to yield $40. So, the government collects $4 of net revenue tomorrow, which is equal to 20 percent of the $20 above-normal return on Patient's investment.
Here, too, the business cash-flow tax appears, at first glance, to impose a penalty on this investment. After all, the tax reduces Patient's consumption by 22 percent, compared to the 20 percent reduction suffered by Impatient.
But the higher tax applies only to the above-normal returns, returns over and above the return on the marginal investment. There is no penalty on a marginal investment, one that yields a 100 percent return. And any investment that yields a before-tax return greater than the marginal return also earns an after-tax return greater than the marginal return. So long as investors are willing to buy machines that yield the marginal return of 100 percent, they must also be willing to buy any that yield more than 100 percent. The business cash-flow tax therefore does not penalize saving and investment on the margin. Yet, it raises revenue by taxing above-normal returns, those that exceed the required rate of return on the marginal investment.
In the presence of uncertainty, the cash-flow tax also has other effects, collecting positive tax from lucky investments and negative tax from unlucky ones. We defer a discussion of those effects, which are much less significant than they initially appear, to chapter 3.
Because the business cash-flow tax applies to savings that have already been done and to savings with above-normal returns that will continue to be done even in the face of the tax, it should not cause households to save less. Because the business cash-flow tax does not cause a change in behavior, it cannot be shifted to other people. The burden of the tax therefore falls on those who own capital at the time the tax is introduced and those who are able to invest at above-normal returns.
Because those groups are generally likely to be well-off, the heavy taxation of business cash flow under the X tax promotes progressivity. Under the X tax, therefore, high tax rates apply to high-paid workers, owners of existing wealth, and recipients of above-normal investment returns, while lower tax rates apply to lower-paid workers. This pattern makes the X tax a progressive two-part VAT.
The President's Advisory Panel on Federal Tax Reform (2005) adopted two different tax reform proposals in its final report. Although one plan would have merely reformed the income tax system, the other plan, called the Growth and Investment Tax Plan, would have largely replaced the income tax system with an X tax, featuring a 30 percent tax rate on business cash flow and the wages of the highest earners. This plan would have retained one vestige of the income tax system, a 15 percent flat rate tax on capital income. The panel considered, but did not adopt, a Progressive Consumption Tax Plan that would have completely replaced the income tax system with an X tax, featuring a 35 percent top tax rate. More recently, Hubbard (2011) urged policy makers to consider replacing the income tax with the X tax.
The PET offers another way to achieve progressive consumption taxation. Under this tax system, each household files an annual tax return on which it reports income, deducts all saving (deposits into savings accounts, asset purchases, amounts lent to others, and payments made on outstanding debts), and adds all dissaving (withdrawals from savings accounts, gross proceeds of asset sales, amounts borrowed from others, and payments received on outstanding loans). The resulting measure equals the household's consumption, which is taxed at graduated rates. The direct consumption tax considered in the Patient-Impatient example in chapter 1 can be viewed as a PET.
The PET was proposed by Kaldor (1955) and extensively analyzed by Andrews (1974). The tax briefly received attention in the policy arena in 1995 when Senators Sam Nunn (D-Georgia) and Pete Domenici (R-New Mexico) introduced the Unlimited Savings Allowance (USA) plan, which would have replaced the individual and corporate income taxes with a PET, accompanied, oddly enough, by a VAT. Due to various problems in its design, the USA plan never received serious consideration in Congress. Robert Frank of Cornell University (2005, 2008) and Edward McCaffery of the University of Southern California Law School (2002) advocate the PET. Andrews (1980), Seidman and Lewis (2009), and Thuronyi (2011) propose levying a PET on high-consumption households as a supplement to the current tax system. Landsburg (2011) also suggests consideration of a PET. A budget plan recently released by the Heritage Foundation (2011, 36–38) would replace income and payroll taxes with a PET, although the plan imposes a flat rate of 25 to 28 percent on consumer spending above an exemption amount, rather than using graduated tax rates.
Given that the X tax and the PET can each generate progressivity and serve as a complete replacement of the income tax system, which is superior? It is certainly easier to explain that the PET is a consumption tax than it is to offer the corresponding explanation for the X tax. That, along with other features, may give the PET an optical advantage. Yet, as discussed in the "Optics of the X Tax and the PET" box (page 39), some optical issues cut in favor of the X tax. In any case, we favor the X tax on balance because of the simplicity offered by its real-based nature. Of course, either the X tax or the PET would be a dramatic improvement over the income tax system.
Advantage of Real-Based X Tax. Because the key differences between the X tax and the PET arise from how the two taxes treat financial transactions, it is useful to review the relationship of real production and financial transactions. Firms and workers engage in real production by using labor and capital to produce goods and services, generating wages for workers and capital income for firms. Two sets of financial transactions determine which households ultimately receive the cash flow and income generated by the real production.
One set of financial transactions initially allocate the income and cash flow generated by production to households. Firms obtain funds from households by issuing stock and bonds, pay funds to households in the form of interest and dividends, and retain funds on behalf of stockholders. These transactions do not change the total cash flow or income generated by the firms' real production.
A second set of financial transactions, such as borrowing and lending, occur between households. These transactions further rearrange cash flow and income, but also result in zero aggregate net cash flow and income. For example, when a lending household receives interest income, the borrowing household incurs negative interest income (interest expense).
A real-based tax system tracks only the production activity of firms and workers, whereas a real-plus-financial tax system also tracks financial transactions. Either system can measure the aggregate income and business cash flow in the economy, but only the real-plus-financial system can measure the income and consumption ultimately enjoyed by specific households. Because the real-based system does not track capital income or business cash flow to the final recipients, it can tax those items only at a flat rate at the firm level. To employ graduated rates based on a household's annual income or consumption, as the PET and the current individual income tax do, we need a real-plus-financial approach that tracks all flows to the final household recipients.
At one extreme, the real-based nature of the VAT eliminates the need for any household tax returns, but it forces all consumption to be taxed at a single flat rate. At the other extreme, the real-plus-financial nature of the PET permits graduated tax rates tied directly to annual consumer spending, but requires households to file annual tax returns reporting a wide array of financial transactions. The X tax follows an intermediate strategy, adopting the real-based approach of the VAT, but taxing wages at the household rather than the firm level. The X tax achieves much, but not all, of the simplicity of the VAT and largely matches the progressivity of the PET.
First, the X tax is almost as simple as the VAT. Although households must file tax returns, they report only their wages, the type of income that is easiest to measure. Like the current individual income tax, the household wage tax would be collected through withholding. As they do today, most households would obtain the necessary information to report their wages from their W-2 forms. Households would not report interest or other capital income and would not deduct interest expense. Few households would need to make quarterly estimated tax payments. The X tax undertakes the relatively simple task of tracking wages to their final recipient, while avoiding the far more difficult task of tracking business cash flows to their final recipients.
Second, the X tax achieves progressivity, but in a less refined manner than the PET. The X tax system taxes wages at graduated rates tied to annual wage income and imposes a high flat tax rate on business cash flow, which largely accrues to the well-off. This progressivity is not as finely calibrated as that achieved by the PET, in which graduated rates are tied directly to annual consumer spending. For example, households that are not affluent may hold a little wealth accumulated prior to the tax reform and may earn a little in above-normal returns. Under the X tax, such households face the high flat tax rate on the business cash flow that they receive, even though this high rate is not appropriate for their economic circumstances. This outcome is avoided under the PET, which links tax rates directly to spending levels. But the PET achieves this finer calibration only by tracking all financial flows and requiring households to report all of their saving and dissaving. We view the additional refinement of the PET to not be worth the associated complexity. If a small number of disadvantaged households are adversely affected by the business cash flow tax, the best solution is to provide them with targeted relief within the X tax system.
Front-Loaded versus Back-Loaded Treatment of Saving. The difference between the treatment of financial transactions under the PET and under the household wage tax in an X tax system can be most easily understood with an analogy to the treatment of different types of tax-preferred accounts under the current tax system.
The PET treatment of financial transactions matches the current treatment of pensions and conventional IRAs, in which savings receive front-loaded tax breaks. Each household deducts inflows into these savings vehicles and pays tax on subsequent withdrawals. In contrast, the treatment of financial transactions under the household wage tax in an X tax system matches the current tax treatment of Roth IRAs, in which savings receive back-loaded tax breaks. Households do not deduct inflows and do not pay tax on withdrawals.
As economists have long known, the two approaches are equivalent for investments with marginal returns if the household remains in the same tax bracket over time. As we showed in our discussion above of the role of the cash-flow tax, the up-front deduction and the inclusion of the proceeds offset each other in present value. Because each household's financial transactions have zero expected market value, the present value of the outflows equals the value of the inflows. If the tax rate remains constant, a tax on the outflows must have a present value equal to the tax savings from deducting the inflow. As we will explain in chapter 3, this result also holds for risky investments.
Of course, the equivalence of front-loaded and back-loaded treatment breaks down if tax rates are not constant. The conventional front-loaded approach used by the PET is more generous if the household is in a higher tax bracket when it saves than when it withdraws; the Roth back-loaded approach used by the household wage tax in an X tax system is more generous if the household is in a higher tax bracket when it withdraws than when it saves.
The issue at hand, though, is not which approach is more generous, but whether the PET's use of the front-loaded approach at the household level provides social gains that warrant the complexity of reporting savings and withdrawals on tax returns. On balance, we think not. If the household remains in the same tax bracket over time, the front-loaded approach has no net effect and financial transactions are tracked for no purpose. If the household moves between different brackets over time, the front-loaded approach actually disrupts the neutrality that could be achieved with constant-rate consumption taxation, as taxpayers are encouraged to save in high-bracket years and to withdraw savings in low-bracket years.
The advantage of using the front-loaded approach at the household level is that it achieves a finer calibration of progressivity for households that have highly variable wages but relatively smooth consumption. This advantage does not seem to warrant the complexity of tracking financial transactions, particularly because, as we will discuss in chapter 3, the X tax can partially attain this advantage by allowing households with highly variable wages to average their wages across different years.
Although the household wage tax in an X tax system adopts a Roth back-loaded approach to saving, the overall X tax system, no less than the PET, adopts the conventional front-loaded approach, with a deduction for savings and a tax on outflows. Rather than implementing the deduction and the tax at the household level as the PET does, the X tax implements them at the firm level through the business cash-flow tax as firms deduct their business investments and are taxed on the proceeds of the investments. The business cash-flow tax is simple because it is imposed at a flat rate and does not track financial flows to individual recipients.
The choice between the X tax and the PET is a close call. But for the reasons described above, we judge the X tax to offer the better combination of simplicity and progressivity. Nevertheless, the X tax poses some challenges that must be addressed before it can be accepted as a replacement for the income tax.
Challenges Facing the X Tax. The X tax faces four major difficulties. In the remainder of this book, we will present solutions to these difficulties and discuss other aspects of X tax implementation.
One problem concerns individuals who work for firms while also providing capital to the firms, such as sole proprietors and some partners and S corporation shareholders. There is no simple way to divide payments that such individuals receive from the firm into wages and business cash flow. This division is critical under the X tax because business cash flow is taxed more heavily than wages. We will discuss this issue in chapter 5.
A second problem under the X tax concerns the significant number of firms that are likely to have negative business cash flow in particular years. The neutrality of the X tax requires that each firm immediately deduct its investment outlays, something that it may be unable to do if excess deductions are not refunded in cash. We will also discuss this issue in chapter 5.
A third problem concerns the tax treatment of financial institutions, whose real activities are often mislabeled as financial transactions in the marketplace. Because the X tax is a real-based tax, it runs the risk of not taxing these transactions properly. In chapter 6, we will present a relatively simple solution that separates real and financial activities in expected market value.
A fourth problem concerns international trade. The X tax does not readily fit into the trade rules adopted by the international community, which were written with the VAT in mind. We will address this issue in chapter 7.
In other chapters, we will discuss the transition to the new tax system and the treatment of pensions and fringe benefits, transfer payments, owner-occupied housing and consumer durables, nonprofit organizations, and state and local governments.
In addressing these challenges, we often exploit the flexibility offered by consumption taxation. As explained above, we generally prefer the basic X tax design, which includes a firm-level business cash-flow tax that imposes zero present-value tax on marginal investments, and generally excludes financial transactions, which also have zero present value. But because these features have zero present value, we can, when appropriate, deviate from the basic X tax design in specific areas without untoward consequences. For example, we choose not to extend the business cash-flow tax to the nonbusiness sector, and we choose to include financial transactions in the tax base for financial institutions and for transactions between U.S. firms and their foreign affiliates. This mixing and matching allows us to capture the advantages that the basic X tax design offers in most applications while avoiding the disadvantages that it poses in selected areas.
We conclude that the X tax is the best consumption tax to use for a wholesale replacement of the income tax system. In the next chapter, we will examine the degree of progressivity that can be achieved under the X tax.
BOX
OPTICS OF THE X TAX AND THE PET
The PET has an optical advantage over the X tax because it satisfies the political demand for a visible tax on households living off of capital income. Of course, no consumption tax system imposes a marginal tax on capital income from new saving; that is done only by income tax systems. But consumption taxes do tax above-normal returns and existing wealth held on the reform date. The PET makes those levies visible because households report their capital income on a tax return while deducting their new saving. In contrast, the X tax imposes no household-level tax on those with capital income, while imposing a highly visible tax on households with labor income. Under the X tax, above-normal returns and existing wealth are taxed much less visibly, at the firm level through the cash-flow tax.
On the other hand, the PET faces an optical challenge with respect to the tax treatment of borrowing. In accord with income tax principles, the proceeds of borrowing are currently not taxed. In contrast, the PET taxes households on the proceeds of borrowing, but then allows the borrower to deduct all subsequent payments, principal and interest, on the loan. Although those deductions cancel out the initial tax in expected market value (if the household remains in the same tax bracket), the tax on borrowing is still likely to be unpopular. The X tax avoids this optical problem by not tracking financial flows; although consumption by borrowers is taxed, the tax is collected at the firm level.
Also, the X tax may have an optical advantage over the PET because Americans generally prefer that business firms "remit" or "pay" part of the tax burden. Although individuals rather than firms ultimately pay taxes, the X tax's inclusion of a cash-flow tax on firms may be politically popular.