- Favoritism among alternative production methods would serve no coherent policy goal:
- State governments should be subject to the employer payroll #tax on the same terms as private firms:
- A producer's VAT base should equal the value of its output less the value of inputs purchased from other producers:
- Proposals to impose an employer payroll tax on state governments have drawn sustained, but unfounded, criticism:
Download PDF The federal government confronts a long-run fiscal imbalance arising from the rapid projected growth of Medicare, Medicaid, and Social Security spending compared with revenue. The enactment of a value added tax is often mentioned as a possible solution to this imbalance. This article is the first part of a two-part examination of the contentious issue of how state governments' provision of goods and services to the public should be taxed under a VAT.1
In this article, I focus on the case in which the state government's output takes the form of consumption rather than inputs into private production. I also assume in this article that state taxes do not affect private production. The second article in this series will generalize the analysis to consider state governments' provision of production inputs and the interaction of a federal VAT with state taxes.
Within the framework at hand, I conclude that neutral design of a federal VAT requires the taxation of private firms' sales to state governments on the same basis as their sales to consumers as well as the imposition of an employer payroll tax (with a rate equal to the VAT rate) on state governments' wage payments. That policy provides neutral treatment of state governments compared with private firms and is also neutral between state governments' in-house production of goods and their purchase of goods from outside suppliers.
Any federal VAT that failed to include a payroll tax on state governments would be nonneutral. If there was no payroll tax and state governments' purchases from private firms were also untaxed, the federal VAT would systematically favor state government provision of goods over private provision. If there was no payroll tax, but state governments' purchases were taxed, the federal VAT would artificially and irrationally favor in-house production by state governments over purchases from outside suppliers. Although the imposition of an employer payroll tax on state governments as part of a federal VAT (or, equivalently, a federal retail sales tax) has drawn sustained criticism, the objections lack a sound economic basis.
The analysis proceeds in the following steps. I first consider the neutral design of an employer payroll tax rather than a VAT, observing that neutrality would require the application of the payroll tax to state governments on the same basis as private firms. Noting that a VAT is equivalent to an employer payroll tax in a simplified economy in which only labor is used in production, I then explain that neutral VAT design requires the imposition of an employer payroll tax on state governments in that economy. I proceed to show that this conclusion carries over to a realistic economy that includes capital as well as labor. Finally, I consider the objections that have been raised to the payroll tax.
Neutrality Under an Employer Payroll Tax
If an employer payroll tax were imposed on private firms, there is little debate that neutrality would require the application of the tax to state governments. If the payroll tax applied to private firms but not to state governments, the tax system would artificially favor goods produced in-house by state government employees. Those goods would be favored over goods provided by the private sector and would also be favored over goods purchased by state governments from outside suppliers for provision to the public.
I assume throughout this article that the federal tax system should be neutral between goods provided by state governments and those provided by the private sector. To be sure, state governments may provide too little of some goods, such as those that benefit politically disadvantaged groups or residents of other states, and too much of other goods, such as those that benefit well-organized special interests. In principle, the federal tax system could seek to correct those failures by taxing some state government goods more favorably --and others less favorably -- than private goods, but it is unlikely that the tax system is well suited to undertake that mission. I therefore assume that neutral federal tax treatment of state government output is desirable. Of course, federal grant programs can be used to encourage or discourage state governments' provision of particular goods, as necessary.
It seems clear that the federal tax system should be neutral between state governments' in-house production and their purchases from outside suppliers. Favoritism among those alternative production methods would serve no coherent policy goal.
A federal payroll tax should therefore apply to state governments on the same terms as private firms. That policy leaves the real operating costs of state governments unchanged from their level in a no-tax economy, where real costs refer to the amount of private goods that must be forgone to produce a fixed amount of state goods.
The point may be easiest to see under the (approximately valid) assumption that the introduction of the payroll tax does not change total employment and total output. In that case, the marginal product of labor is also unchanged, which implies that the introduction of the tax does not change the total tax-inclusive amount that a private firm is willing to pay to obtain a worker. The wage that a firm is willing to pay to the worker then must fall by the amount of the tax, causing the employer payroll tax to be fully shifted to workers in the form of lower real wages.2 Like private firms, state governments reduce their real wage payments to workers by the full amount of the tax, leaving no net effect on state budgets. Their real operating costs are the same as they were before the introduction of the payroll tax.
Of course, state governments face higher operating costs if they are subject to the employer payroll tax than they would face if they were exempt from it. If state governments were exempt from the tax, its introduction would actually lower their real operating costs. Because the application of the tax to private firms would drive down real wages throughout the economy, state governments would enjoy a reduction in their wage costs without having to make any offsetting tax payments. Subjecting state governments to the payroll tax on the same terms as private firms does not increase their operating costs, but merely denies them the cost reduction they would enjoy if they were exempt from the tax.
The employer (and employee) payroll taxes currently imposed to finance Social Security and Medicare Part A deviate from the neutral policy described above. Under section 3121(b)(7)(E) of the Internal Revenue Code, the Social Security payroll tax does not apply to wages paid by those state and local governmental units that opted out of Social Security before 1983. (The Social Security Act no longer permits such opt-outs.) Under Code section 3121(u)(2), those wages are generally subject to the Medicare payroll tax, except for wages paid to employees who have been continuously employed by the governmental unit since March 31, 1986.
It is widely and correctly recognized that the exemption offered by section 3121(b)(7)(E) is a deviation from neutral treatment. Indeed, a staple of Social Security reform plans is the prospective elimination of that nonneutrality through the extension of the Social Security payroll tax to all newly hired state government employees.3
The discussion so far has examined the neutral design of a federal employer payroll tax. I now consider the implications of this analysis for the neutral design of a federal VAT.
Neutrality Under a VAT in a Labor-Only Economy
I begin the VAT analysis by considering an economy in which labor, but not capital, is used in production. In that labor-only economy, the price of a good produced by a private firm equals the labor costs of its production. Similarly, the value added at each stage of production equals the wage payments made at that stage. For private firms, a VAT is therefore identical to an employer payroll tax.
Consider an example in which a manufacturer pays $10 in wages to produce a good. The manufacturer then sells that good for $10 to a refiner, who pays $3 in wages to refine the good. The refiner sells the good for $13 to a retailer, who spends $2 in wages to bring the good to market, where it is sold to a consumer for $15.
A credit-invoice VAT collects tax on $10 from the manufacturer. It then collects gross tax on $13 from the refiner, but allows credit for the tax paid on the $10 at the manufacturing stage. The refiner's net tax base is therefore $3. Similarly, the VAT collects gross tax on $15 from the retailer, but allows credit for the gross tax paid on the $13 at the refining stage. The retailer's net tax base is therefore $2. Each firm's net tax base is its value added, which is equal to its wage payments in the simplified labor-only economy. The three firms' tax bases sum to $15, the consumer price of the good. Because each firm's tax base is equal to its wage payments, the VAT is equivalent to an employer payroll tax for the private firms.
Now, assume that the retailer is a state government rather than a private firm.4 How should the state government be taxed to maintain neutrality with the VAT's treatment of private firms?
The previous analysis of an employer payroll tax provides a ready answer to this question. Under the mathematical axiom of transitivity, if A is equal to B and B is equal to C, A is equal to C. In the previous section, we saw that an employer payroll tax on state governments is neutral compared with an employer payroll tax on private firms. We have now seen that in the simplified labor-only economy, an employer payroll tax on private firms is equivalent to a VAT on private firms. It inexorably follows that an employer payroll tax on state governments is neutral compared with a VAT on private firms.
In the simplified economy, neutral VAT design calls for the imposition of an employer payroll tax on the state government retailer. Of course, the state government's purchase from the refiner is also subject to VAT; that the retailer is a state government rather than a private firm cannot change the VAT treatment of the manufacturer and refiner if those firms' payrolls are to remain in the tax base.
In the example, the state government would be subject to an employer payroll tax on its $2 of wage payments. There would be no credit for the VAT paid on the $13 of goods it purchases from the refiner. The resulting tax payment is identical to that which would result from taxing the state government on the correctly computed $15 value of its output and allowing it to claim credit for the VAT collected on the $13 purchased from the refiner.
Although the economics are unchanged from the preceding section, the conclusion that state governments should be subject to an employer payroll tax is less obvious in this context. When private firms and state governments were both subject to a 10 percent employer payroll tax, the neutrality was undeniable. But when private firms are subject to a 10 percent VAT and state governments are subject to a 10 percent employer payroll tax, the neutrality is less transparent. The two types of entities appear to face distinct, even dissimilar, taxes. But it is important to remember that the VAT and the employer payroll tax are equivalent in the simplified economy. To be sure, workers are generally thought to bear the burden of an employer payroll tax while consumers are thought to bear the burden of a VAT. In the simplified labor-only economy, however, there is no distinction between those two groups because workers produce the output and buy it with their wages.
At first glance, it may seem that the surest and simplest path to neutrality would be to subject state governments to the same VAT that private firms face, not to apply a superficially dissimilar employer payroll tax to them. Neutrality can indeed be achieved by imposing a correctly structured VAT on state governments; as discussed below, that VAT turns out to be identical to an employer payroll tax. Imposing an incorrectly specified VAT would not, however, result in neutrality.
A producer's VAT base should equal the value of its output minus the value of inputs purchased from other producers. For a private firm that sells its output at market prices, the value of the firm's output equals its sale proceeds. The situation is more complex, however, for a state government program that finances the provision of goods through compulsory tax payments rather than by selling them. A superficially appealing method would be to value its output at zero, the price at which it is provided to the public. If the government program made any purchases from outside suppliers, that method would treat the government program as having negative value added. That method would be incorrect, however, because goods provided by state governments do not cease to have value merely because they are financed by taxes rather than by user charges.
Instead, the VAT system should value the goods at cost, which is equal to wage payments in that simplified labor-only economy. If the state government produces efficiently, the output's value is equal to cost, on the margin. And if it does not produce efficiently, the tax system should not reward that inefficiency with a lower tax burden. The correctly computed VAT therefore turns out to be an employer payroll tax. Because state governments do not sell their output at market prices, it is necessary to tax them in a manner superficially different from private firms, through the imposition of an employer payroll tax rather than an explicit VAT on sale proceeds minus purchase costs.5 But there is no difference in substantive tax treatment.
As in the previous section, the application of the employer payroll tax leaves state governments' real operating costs unchanged from the no-tax economy. Because a VAT reduces real wages in the same manner as an employer payroll tax, state governments enjoy real wage cost reductions that offset their employer payroll tax payments.6 In a no-tax economy, a worker with a marginal product of 100 apples is paid a wage equal to the consumer price of 100 apples. In an economy with a 10 percent VAT, that worker is paid a wage equal to the consumer price of 90 apples because the VAT reduces the value of the worker's output to the firm by 10 percent.7
A federal VAT might zero rate some consumer goods, as foreign VATs often do. A firm that makes retail sales of a zero-rated good pays no tax on its sale proceeds, but it still claims credit for the gross VAT paid at the earlier stage of production. If the item in the previous example was zero rated, tax would still be collected from the manufacturer on a base of $10 and from the refiner on a net base of $3. But the retailer would have a net tax base of negative $13 because it would pay no tax on its $15 of sales receipts and would claim credit for the gross VAT paid on its $13 purchase from the refiner. The combined tax bases of the three firms would be zero, freeing the good of tax.
Zero rating necessarily introduces nonneutrality by favoring some goods over others. But neutral treatment between state governments and private firms is likely to still be desirable. Accordingly, state governments should also be accorded a zero rate when they provide goods that are identical to, or close substitutes for, goods that receive a zero rate when provided by the private sector.
In the example, therefore, if the goal is zero-rated, the combined tax base should remain equal to zero if the retailer is a state government rather than a private firm. That outcome can be achieved by sparing the state government from the employer payroll tax on its $2 of wage payments while allowing it to claim credit for the $13 paid on its purchase from the refiner, giving it a tax base of negative $13. Adding in the manufacturer's tax base of $10 and the refiner's net tax base of $3 yields a combined tax base of zero. Alternatively, the appropriate outcome can be achieved by sparing the state government from the employer payroll tax and zero-rating the refiner's $13 sales to the state government. The state government's tax base is then zero while the retailer's net tax base is negative $10 because the latter pays no tax on its sales while claiming credit for the $10 VAT paid on its purchase from the manufacturer. Adding in the manufacturer's tax base of $10 yields a combined tax base of zero.
As can be seen, exemption from the employer payroll tax is insufficient to replicate zero rating. The tax on the refiner's sale to the state government must also be removed or credited. If the refiner's sale remained subject to tax, the combined tax base would be $13 rather than zero. That policy would artificially induce state governments to do all production in-house, because the only way to achieve the full benefits of zero rating would be for the manufacturing, refining, and retailing to all be done by the state government.
As a general principle, neutrality between in-house production and purchases from outside suppliers requires that the employer payroll tax and the tax on the state government's purchases stand or fall together. In the normal case, both taxes should be imposed. In the zero-rated case, neither tax should be imposed.
From an administrative perspective, zero rating is much harder to apply to state governments than to private firms. Suppose that the VAT zero rates bread but not candy. When a private firm sells both bread and candy, the VAT system taxes the firm's receipts from candy sales, but not its receipts from bread sales. Credit for VAT paid at the preceding stage of production is claimed in both cases, so there is no need to track the inputs that go into the production of each good. But if a state government produces both bread and candy and distributes them free of charge, the VAT system must distinguish the wages paid to workers producing bread from the wages paid to workers producing candy, as the former, but not the latter, wages are exempt from the employer payroll tax. The VAT system must also distinguish outside purchases that are used to produce bread from outside purchases that are used to produce candy because the former, but not the latter, purchases have to be freed from the VAT, either by giving the state government a credit or by zero rating the outside suppliers.
An Economy With Capital
So far, the discussion has assumed a labor-only economy. I now extend the analysis to an economy that also includes capital. In that economy, a VAT clearly differs from an employer payroll tax. Because the two taxes continue to have similar effects on the allocation of economic resources, however, the analysis is largely unchanged.
In general, a VAT can be viewed as having two components. Define each firm's business cash-flow to be its value added minus its wage payments. (Because a firm's value added equals its business receipts minus its business purchases, its cash-flow equals its business receipts minus both its business purchases and its wage payments.) By definition, a tax on value added is equivalent to a tax on wage payments plus a tax on business cash-flow. In other words, the VAT is economically identical to an employer payroll tax plus a cash-flow tax. In the labor-only economy, of course, there was no business cash-flow and the VAT was equivalent to an employer payroll tax alone.
The decomposition is helpful because the payroll and cash-flow components of the VAT have profoundly different effects. As explained below, the cash-flow component does not alter the allocation of resources in the economy, so neutrality does not require that it be imposed on state governments. That turns out to be fortunate, because imposition of the cash-flow tax would be infeasible. The effect of the VAT on the allocation of resources across different sectors arises solely from its payroll component. Neutrality therefore continues to require that this component apply to state governments on the same basis as private firms, which is done by imposing an employer payroll tax on state governments.
Although a full treatment is impossible here, economists have long recognized that a cash-flow tax does not distort economic decisions.8 Unlike a tax on business income, a tax on business cash-flow does not penalize investment on the margin. The difference is that investment is expensed under the cash-flow tax, but is depreciated under the income tax. Under the cash-flow tax, the tax savings from the upfront expensing deduction are equal in expected present value to the tax on the future cash flows generated by a marginal investment that yields only the minimum rate of return demanded by savers.
Although the cash-flow tax does not tax investment on the margin, it raises revenue from investment in two ways. First, it raises revenue from the capital in place at the time the cash-flow tax is adopted, reflecting investments made at earlier dates. The cash flows generated by that capital are taxed, but there is no offsetting expensing deduction because the investments occurred before the tax system and its expensing policy were in place. Second, it raises revenue from any investments that yield above-normal returns. For those investments, the present value of the tax on the future cash flows exceeds the tax savings from the upfront expensing deduction.
The collection of that revenue does not cause the cash-flow tax to discourage new investment. The first revenue source is an unexpected burden on investments already in place and does not apply to new investments. The second revenue source taxes away part, but not all, of the extra return, above and beyond the minimum rate of return demanded by savers, generated by some investments. Because these investments continue to yield more than the minimum demanded return after the tax, they will still go forward.
Because the cash-flow tax does not change investment behavior, it is not shifted. A cash-flow tax on private firms therefore falls on investors who own capital when the tax is implemented and those with access to investment opportunities that yield above-normal returns.9 In general, those investors are likely to be affluent.
In summary, the application of the cash-flow tax to private firms raises revenue from the affluent without imposing economic distortions. The imposition of that tax is therefore good policy. Regarding private firms, then, it is better policy to impose a VAT, which is an employer payroll tax plus a cash-flow tax, rather than an employer payroll tax alone.
In contrast, there is little reason to impose the cash-flow tax on state governments. Because the cash-flow tax imposes zero marginal tax on new investment, investment is allocated in a neutral manner, with or without the tax. The failure to impose the cash-flow tax on state governments simply means that the federal government forgoes collecting revenue from the capital held by state governments on the date the VAT is adopted and from any above-normal returns earned by state governments. Collecting that revenue would unnecessarily burden state and local taxpayers or recipients of state and local services, who are generally less affluent than private investors.
It is fortunate that the cash-flow tax need not be imposed on state governments because imposition would not be feasible. The imposition of the cash-flow tax would require accurate measurement of the value of state government output, including police and fire protection -- a hopeless task. In the labor-only economy, the value equaled the wage costs of producing the output, but that does not hold in an economy with capital. Accurate measurement is impossible because the "cash-flow" from goods produced by state governments does not actually take the form of cash.
The decomposition of the VAT into its payroll and cash-flow components reveals that the first component, but not the second, should be applied to state governments. Imposing the payroll component only on private firms would have exactly the same nonneutral effects as in the labor-only economy, creating an artificial incentive for workers to be employed by state governments. In contrast, the cash-flow component need not and cannot be applied to state governments.
The correct approach, then, is to impose an employer payroll tax on state governments when a VAT is imposed on private firms. The employer payroll tax is only a portion of the VAT, but it is the portion that affects the allocation of resources across sectors and the portion that therefore should be applied to state governments. State governments can be spared from the remainder of the VAT, the cash-flow component.
The correctness of this conclusion becomes crystal clear when one considers the appropriate treatment of state governments under a Hall-Rabushka flat tax or a Bradford X tax. Each of those tax systems splits the VAT into its two components, with the payroll component imposed on workers, rather than firms. By itself, of course, that change in legal incidence has no economic significance. Imposing the tax on workers is useful, though, because it permits the flat tax to provide relief to low-paid workers through an exemption allowance and permits the X tax to provide comprehensive progressivity through an exemption allowance and graduated tax rates.10 As one might expect, and in accord with the above analysis, proposals for a flat tax or an X tax call for the cash-flow tax to be imposed on private firms, but not state governments. But the plans uniformly call for state government employees to pay the same wage tax as other workers. That treatment is absolutely uncontroversial. Indeed, any suggestion that state government employees be exempted from the tax paid by other workers would arouse immediate opposition as senseless discrimination. But if state governments' wage payments belong in the tax base under a flat tax or X tax, they must belong in the tax base under a unitary VAT. Splitting the VAT in two and changing the legal incidence of one component does nothing to change the criteria for neutrality between state governments and the private sector.
Proposals to impose an employer payroll tax on state governments as part of a federal VAT or retail sales tax have drawn sustained, but unfounded, criticism. Much of the discussion has occurred in the context of the FairTax plan, which would replace the individual and corporate income taxes, payroll and self-employment taxes, and the estate and gift tax with a 29.9 percent retail sales tax.
As is well known, important aspects of the FairTax plan are seriously flawed. Moreover, many of the claims made by the proposal's supporters are invalid or even nonsensical.11 I join many members of the tax policy community in opposing the FairTax plan. Unfortunately, however, some opponents of the plan have made invalid arguments against it. Invalid criticisms of the plan's treatment of state governments have been particularly common.
Under the FairTax bill,12 proposed Code section 2(a)(12)(A)(ii) defines the term "taxable employer" to include "any government except for government enterprises." Proposed section 2(a)(14)(B)(i) applies sales tax to "any service performed by an employee for which the employee is paid wages or a salary by a taxable employer." Under proposed section 102(d)(1), the person using the taxable services is generally liable for the payment of the tax. Under proposed section 103(b)(2), "In the case of wages or salary paid by a taxable employer which are taxable services, the employer shall remit the tax." Proposed section 703(a)(2) provides that "purchases by State governments and their political subdivisions of taxable property and services shall be subject to" sales tax.
Although the FairTax plan calls for a retail sales tax rather than a VAT, the economics are the same. Its application of a payroll tax to state governments is therefore the correct treatment.13
That provision has been repeatedly criticized as increasing the real operating costs of state governments. In a 2007 article, Bruce Bartlett asserts that the plan "raises [state governments'] spending without simultaneously raising their revenue. Realistically, their only choice is to increase their taxes to pay the FairTax on their spending, including that for police protection and other essential services." He goes on to estimate the magnitude of the alleged cost increase.14 Citing estimates that state governments would have to pay more than $300 billion in federal sales tax under the plan, Michael Graetz asserts that this money "would have to come from higher state and local taxes or reduced state and local spending. Somebody is going to have to pay."15 Hank Adler similarly complains that the plan "would cause an immediate increase in the cost of state government."16
As explained above, however, there is no cost increase because state governments' tax payments would be offset by the reduction in their real wage costs caused by the retail sales tax. It is particularly surprising that Bartlett and Graetz fail to recognize this point, as they emphasize that the sales tax would reduce real wages and rightly condemn the numerous FairTax supporters who deny that fact.17
Graetz further contends that the imposition of the payroll tax would create "a significant incentive for governments to contract out their functions to private enterprises."18 In reality, there would be no incentive for a state government to contract out, because doing so would merely trade a 29.9 percent payroll tax for a 29.9 percent sales tax. In contrast, the failure to impose the payroll tax would create a significant disincentive for governments to contract out.
Bartlett and Graetz both emphasize that the sales tax rate would have to be higher if state governments were not taxed and that the plan taxes state governments in order to lower the rate. That objection is hard to understand. Narrowing a tax base by artificially excluding some goods always requires the tax rate to be higher; broadening the base to make it more neutral is always a good way to lower the rate. Bartlett also notes that most foreign VATs exclude goods provided by sub-national levels of government from the tax base. Although that is correct, foreign VATs contain many base- narrowing provisions that good tax policy should not imitate.
Other commentators have provided a more careful analysis. As usual, the Urban-Brookings Tax Policy Center gets the economics right even as others go astray. When the center's Eric Toder, Jim Nunns, and Joseph Rosenberg specified a deficit-reduction VAT proposal in November 2011 that applied zero rating to goods and services provided by government, they correctly noted that governments' real operating costs would decline because of the real wage reduction caused by the VAT.19 When the same authors specified a version of Graetz's VAT plan in January 2012, they provided for an employer payroll tax on state governments and inclusion of purchases by state governments from private firms in the tax base. Departing from Graetz's own statements, they correctly observed that state governments' real operating costs would remain unchanged under that policy.20 George Yin, a staunch opponent of the FairTax plan, correctly observes that its taxation of state government avoids "an undesirable bias" in favor of government provision and that the imposition of an employer payroll tax is necessary "to avoid distorting the choice between hiring employees rather than independent contractors."21 William Fox and Matthew Murray correctly note that the FairTax achieves "neutrality between consumption of government and private goods and services."22
Economics aside, some observers have argued that the taxation of state governments would unconstitutionally violate their intergovernmental immunity from federal taxation. Graetz, for example, suggests that such taxation is "constitutionally questionable." Although the text of the U.S. Constitution does not restrict the federal government's power to tax state governments, the U.S. Supreme Court has quite properly inferred restrictions on that power, reflecting the status of the states as independent and indestructible parts of the constitutional system.
Yin devotes a lengthy footnote to a thoughtful review of the constitutional issue in the context of the FairTax plan.23 He notes that the Supreme Court's recent decisions, unlike its earlier decisions, allow the federal government to tax transactions between state governments and private parties, if the legal incidence of the tax is placed on the private parties.24 As he observes, though, the FairTax plan places the legal incidence of the tax on the state governments. A VAT would probably also place legal incidence on state governments. Of course, legal incidence has no economic relevance.25
Yin notes that the Supreme Court has indicated that direct taxation of state governments is permissible in some circumstances.26 He notes, though, that the Court said in dictum in 1946 that the federal government cannot tax state government activities that are unique to a state, such as having a statehouse.27 That concern, which the Court has not elaborated on in later decisions, would apply, at most, to a subset of state governments' activities.
Of course, state governments cannot be subjected to discriminatory taxation. As explained throughout this article, the taxation of purchases by state governments and the imposition of an employer payroll tax results in economically neutral treatment of state governments compared with private firms. The only potential problem is the appearance of discrimination arising from the fact that state governments face an employer payroll tax while private firms face a superficially dissimilar VAT. If appearance is the problem, the solution is to change the appearance. Rather than imposing a unitary VAT on private firms, the tax system should impose two separate taxes -- an employer payroll tax and a business cash-flow tax. State governments should be subject to the employer payroll tax on the same terms as private firms and should be exempt from the cash-flow tax. Neither component of the tax system would then discriminate against state governments, because the first component would treat them neutrally and the second would accord them favorable treatment.
This article, the first in a two-part series, examines the neutral treatment of state governments under a federal VAT. Focusing on the case in which state governments provide consumer goods rather than production inputs, the analysis shows that the neutral policy involves an employer payroll tax on state governments and the application of VAT to private firms' sales to state governments. The second article in this series will consider the case in which state governments provide production inputs and will examine the treatment of state taxes under a federal VAT.
Marginal Impact is a column from the American Enterprise Institute for Public Policy Research. Alan D. Viard is an AEI resident scholar. The views expressed in this article are those of the author and do not necessarily reflect the views of any other person or organization.
1 For brevity, I refer to "state governments" throughout this article, using the term to refer to state, local, and tribal governments.
2 Depending on the Federal Reserve's monetary policy response, the real wage reduction could occur through a reduction in nominal wages or through an increase in consumer prices. See infra note 7.
3 For one example among many, see the Bipartisan Policy Center, Debt Reduction Task Force, "Restoring America's Future: Reviving the Economy, Cutting Spending and Debt, and Creating a Simple Pro-Growth Tax System," Nov. 2010, p. 79 (recommending application of Social Security payroll tax to all state government employees hired after 2020), available at http://www.bipartisanpolicy.org/sites/default/files/BPC%20FINAL%20REPORT%20FOR%20PRINTER%2002%2028%2011.pdf
4 The second article in this series will examine the case in which the manufacturer or the refiner is a state government. As previously stated, this article considers only the case in which the state government's output is consumption, rather than an input into private production.
5 The same analysis, here and in the remainder of the article, applies to nonprofit institutions that follow a similar practice. Some state governments and nonprofits operate self-supporting commercial enterprises that sell output at market prices. If desired, those enterprises can be subjected to the explicit VAT rather than to the employer payroll tax.
6 The VAT's reduction of real wages illustrates the general principle that consumption taxation penalizes work in the same manner as income taxation. Although consumption taxation avoids income taxation's saving penalty, it shares its work penalty. For further discussion, see Robert Carroll and Alan D. Viard, Progressive Consumption Taxation: The X Tax Revisited (Washington: AEI Press, 2012), pp. 6-10, 13, 18-19.
7 As alluded to in note 2, the form of the real wage reduction depends on Federal Reserve policy. Because the price level is determined by monetary policy, the enactment of a VAT or any other tax does not increase consumer prices unless the Fed accommodates the tax by expanding the money supply. The Fed is likely to accommodate taxes that lower the equilibrium level of real wages paid by firms to workers, such as a VAT, retail sales tax, or employer payroll tax. If the Fed does not accommodate those taxes, the economy must reach equilibrium through a reduction in the nominal wages paid by firms. Because nominal wages are often slow to adjust downward, the process of reaching equilibrium in this manner can be protracted, with unemployment elevated until the process is complete. The Fed can avert those problems by accommodating the tax, allowing real wages to be reduced through price increases rather than nominal wage cuts. For further discussion, see Carroll and Viard, Progressive Consumption Taxation, supra note 6, pp. 166-169.
8 In economic terminology, it is a lump sum tax. For further discussion of cash-flow taxes, see Carroll and Viard, Progressive Consumption Taxation, pp. 30-32.
9 Unlike in the labor-only economy, the burdens of a VAT and an employer payroll tax do not fall on exactly the same groups. The VAT taxes all consumption whether it is financed from wages, capital held at the time the tax is adopted, or super-normal investment returns, while the payroll tax applies only to consumption financed from wages.
10 Imposing the payroll tax on workers also eliminates any need for the Federal Reserve to accommodate the tax by expanding the money supply and raising the consumer price level, as discussed in note 7. For a discussion of the logic behind splitting the VAT into two parts, see Carroll and Viard, Progressive Consumption Taxation, pp. 24-29.
11 The effort to collect a high-rate tax solely at the retail level (rather than at each stage of production as under a VAT) would be likely to encounter significant compliance problems, which would be greatly amplified by having state governments collect the tax on behalf of the federal government. Although many of the proposal's supporters describe it as revenue-neutral, it would actually lose significant revenue at the 29.9 percent tax rate. Many supporters also falsely deny that the proposal would reduce the real value of existing capital, reduce real wages, create pressure for the Federal Reserve to increase the consumer price level, and shift the tax burden from wealthy taxpayers to middle-income taxpayers.
12 The bill was introduced in the 112th Congress as H.R. 25 on January 5, 2011, and was routinely referred to the House Ways and Means Committee. No further action is likely.
13 The bill goes astray, though, by providing in proposed section 705 that the payroll tax would not apply to nonprofit institutions. See supra note 5.
14 Bruce Bartlett, "Why the FairTax Won't Work," Tax Notes, Dec. 24, 2007, p. 1241.
15 Michael J. Graetz, 100 Million Unnecessary Returns: A Simple, Fair, and Competitive Tax Plan for the United States (New Haven: Yale University Press, 2008), pp. 187-188.
Hank Adler, "FairTax -- Not Ready for Prime Time," Tax Notes, Jan. 14, 2008, p. 301.
17 Bartlett, supra note 14, at pp. 1247-1248; Graetz, supra note 15, at pp. 42-43.
18 Graetz, supra note 15 at p. 188.
19 Eric Toder, Jim Nunns, and Joseph Rosenberg, "Using a VAT for Deficit Reduction," Nov. 2011, pp. 9-11.
20 Eric Toder, Jim Nunns, and Joseph Rosenberg, "Using a VAT to Reform the Income Tax," Jan. 2012, pp. 10, 16-17.
21 George K. Yin, "Is the Tax System Beyond Reform?" Florida Law Review, 58(5), Dec. 2006, pp. 977-1042, at pp. 985-987. Yin expresses concern, at p. 987, that there may be a lingering bias in favor of state government production because state governments would not be subject to the cash-flow-tax component of the sales tax. As explained above, his concern is unnecessary because the cash-flow tax does not affect economic incentives.
22 William F. Fox and Matthew N. Murray, "A National Retail Sales Tax: Consequences for the States," State Tax Notes, July 25, 2005, p. 287, Doc 2005-14109 , or 2005 STT 141-1 . Fox and Murray correctly observe that the FairTax would not provide state governments the advantage that they currently receive from the federal income tax deduction for state taxes.
23 Yin, supra note 21, at p. 987 n.43.
24 See Helvering v. Gerhardt, 304 US. 405 (1938) (upholding federal income tax on wages of state government employee), overruling Collector v. Day, 11 Wall. 113 (1871); South Carolina v. Baker, 485 U.S. 505 (1988) (upholding federal government's power to impose income tax on municipal bond interest), overruling Pollock v. Farmers' Loan & Trust Co., 157 U.S. 429, 583-586 (1895).
25 For a discussion of the irrelevance of legal incidence in another context, see Alan D. Viard, "Railroad Taxes, Discrimination, and the 4-R Act: A Misstep by the Eighth Circuit," State Tax Notes, Jan. 9, 2012, p. 161, Doc 2011-24975 , or 2012 STT 5-1 .
26 See Baker, 485 U.S. at 523 n.14.
27 See New York v. United States, 326 U.S. 572, 582 (1946).