A substantial amount of production occurs outside of the business sector, including the services provided by owner-occupied housing, consumer durables, governments, nonprofit institutions, and household employees. In this chapter, we explore the treatment of these sectors under the X tax.
Although Mintz (1996, 471), the Tax Executives Institute (1992, 24), Conrad (2010, 471), and others argue that owner-occupied housing poses a difficulty for VATs and other consumption taxes, the reality is almost exactly the opposite. The proper income tax treatment of owner-occupied housing is very difficult to achieve, and no actual income tax system even approximates the correct treatment. In contrast, the correct treatment, in expected present value, can be easily achieved under consumption taxation.
Ironically, the current U.S. tax treatment of owner-occupied housing is thoroughly incompatible with the principles of income taxation, but is relatively similar to the treatment prescribed by the principles of consumption taxation. The move to correct consumption tax treatment under the X tax, therefore, involves making only minor changes to the current tax treatment.
As explained below, the appropriate treatment of owner-occupied housing can be achieved by taxing the sale of new homes and ignoring all resales. The sale of a new home is treated as the sale of a consumer good, and all the basic X tax rules are then applied without modification. As explained below, these rules should also apply to owner-used consumer durables.
An owner-occupied home is an investment asset that produces output in the form of housing services. The value of these services is equal to the imputed rental value of the home. Nevertheless, as explained below, treating the home as a consumer good yields the correct result in expected present value.
We first consider the treatment of housing under the income tax and then examine its treatment under the X tax.
Owner-Occupied Housing under Income Taxation. Under a textbook income tax, purchases of investment assets are not deducted, and the net-of-depreciation return is taxed. Accordingly, an ideal income tax allows no deduction for home purchases, taxes imputed rent net of maintenance costs and economic depreciation, and taxes accrued real capital gains net of losses. Mortgage interest payments, like other interest payments, are deducted.
Implementation of the correct income tax treatment of owner-occupied housing is impractical. The taxation of imputed rent and accrued capital gains would involve formidable measurement problems each year for each homeowner in the United States. Distinguishing maintenance expenditures, which should be currently deducted against imputed rent, from improvement expenditures, which should be capitalized and amortized, involves the standard complexities that income taxes face in distinguishing between current and capital expenditures, as we discussed in chapter 1.
Unsurprisingly, the current tax treatment of owner-occupied housing deviates sharply from the income tax principles outlined above. Notably, imputed rent is excluded from taxable income. Repair and maintenance expenditures are nondeductible because they constitute costs of earning tax-exempt imputed rental income. Most, but not all, capital gains are also excluded from taxable income; a taxpayer who has used a home as a principal residence for at least two of the previous five years may exclude up to $250,000 ($500,000 for a couple) of gain on the home's sale, with smaller maximum exclusions if the home was used as a principal residence for less than two years. Any gains that do not qualify for this exclusion are taxed, in nominal rather than real terms, at the time of sale, not as the gains accrue. Capital losses on owner-occupied homes are nondeductible, except that an itemized deduction is allowed for casualty losses that exceed 10 percent of adjusted gross income. Improvement expenditures are capitalized, but, rather than being amortized, are added to the taxpayer's cost basis and deducted from any future capital gains that do not qualify for the exclusion described above. An itemized deduction is allowed for interest payments on up to $1.1 million of mortgage debt on up to two homes. The Joint Committee on Taxation (2011) provides a thorough description of the current tax treatment of owner-occupied housing.
Under the current tax system, owner-occupied housing faces an effective tax rate of approximately zero, because imputed rent and most capital gains are not taxed. In contrast, business capital is generally taxed at significantly positive effective rates. This nonneutral tax treatment drives investment away from business capital and toward owner-occupied housing. Altering the allocation of capital in this manner is economically inefficient unless investment in owner-occupied housing provides spillover benefits to society that cannot be captured by the homeowner.
Some people argue that homeownership provides such spillover benefits by promoting social stability and neighborhood cohesion. Economists and others continue to debate the validity of this argument. But even assuming that the argument is correct and that some tax preference for homeownership is therefore desirable, today's tax treatment is still improperly designed. Although society may benefit from a person owning a home rather than renting, it is hard to see how society receives significant additional benefits from a person owning an expensive home rather than one of modest value. While current tax policy may prompt some taxpayers to become homeowners, it also prompts many others who would own homes anyway to buy more expensive homes. Indeed, the highest-income taxpayers, who would generally own quite expensive homes in any event, are in the highest tax brackets and therefore receive the largest percentage subsidy from the exclusion of imputed rent, giving them the strongest incentive to buy even more expensive residences.
Political difficulties aside, neutral treatment of owner-occupied housing is unlikely to be achieved under the income tax. To begin, neutrality requires the taxation of imputed rent. Although such taxation is not impossible (the Joint Committee on Taxation [2011, 30] reports that nine of the thirty-three OECD countries tax imputed rent to some extent), it is clearly administratively difficult. Economists sometimes propose that mortgage interest be made fully or partly nondeductible as a rough-and-ready offset to the exemption of imputed rent. Although such a policy probably would be beneficial, it would be imperfect because it would impose positive tax only on debt-financed home purchases while maintaining a zero effective tax rate on equity-financed purchases. It would therefore only partly correct the current system's bias in favor of home purchases and would actually introduce a new distortion by favoring equity-financed purchases over debt-financed purchases.
Owner-Occupied Housing under the X Tax. Fortunately, things are much more straightforward under the X tax, as there are two ways to provide appropriate treatment to owner-occupied housing: an ex post method and a prepayment method. In keeping with the real-based nature of the X tax, both approaches ignore financial transactions, so there are no tax implications of mortgage interest or principal payments for either the homeowner or the lender. Although the ex post method would be difficult to administer, the prepayment method is largely similar to, but slightly simpler than, the approach adopted by the current income tax system.
Under the ex post method, taxpayers deduct home purchases and repair, maintenance, and improvement expenditures. Imputed rent is taxed, as are the full proceeds from home sales. Homeowners are effectively subject to the cash-flow tax, with imputed rent treated as "cash" flow, so homes are treated in the same manner as business capital is treated under the X tax. Although this method is slightly simpler than the treatment prescribed by income tax principles because it need not measure accrued capital gains and losses or distinguish repair and maintenance costs from improvements, it retains the formidable difficulties of taxing imputed rent.
Fortunately, the prepayment method allows dramatic simplification while remaining faithful to consumption tax principles. This approach provides no deduction for home purchases or repair, maintenance, and improvement costs and imposes no tax on imputed rent or home sales. This method exempts homeowners from the cash-flow tax that applies to business firms. The only tax on homes is the tax on original construction that arises because the wages and business cash flow associated with the construction are taxed; a zero effective marginal tax rate applies to the imputed rental returns that the house subsequently generates.
This zero effective tax rate is essentially equal to that imposed by the current tax system. This equality is no coincidence; the prepayment method is identical to current tax policy, if we set aside the fact that the current system taxes a few capital gains from home sales. (The other apparent difference from the current tax system—the inability of the homeowner to deduct mortgage interest payments—is offset by the absence of tax on the mortgage lender, as further discussed below.)
But if it is undesirable for the current income tax system to apply a zero effective tax to owner-occupied homes, how can it be beneficial for the X tax to do the very same thing? The difference is that the X tax is designed to apply a zero marginal effective tax rate to all investment returns. The prepayment method delivers exactly the right result under the X tax, giving owner-occupied homes the same zero effective tax rate that the X tax applies to business capital, and thereby maintaining neutrality between the two types of investment. And, as we discussed in chapter 1, the zero effective tax rate also maintains neutrality between investment and current consumption.
In contrast, an income tax is supposed to impose a positive effective tax rate on all investment returns, and the current tax system generally does so for business capital. By using the prepayment method and applying a zero effective tax rate to owner-occupied homes, the current system improperly favors them over business capital. To achieve neutrality, the income tax would need to extend the positive tax rate on business capital to owner-occupied homes by taxing imputed rent. The X tax achieves neutrality by moving in the opposite direction, extending the zero effective tax rate that now applies to owner-occupied homes to business capital.
Because it avoids the taxation of imputed rent, the X tax prepayment method is as simple as the current tax system. Actually, it is slightly simpler, because it sweeps away the current system's vestiges of capital gains taxation and the associated capitalization of improvement costs and, as discussed below, eliminates the casualty loss deduction. But unlike the current tax system, the X tax achieves this simplicity without favoring owner-occupied housing over business capital.
Owner-Occupied Housing and the Business Cash-Flow Tax. Although the prepayment method results in owner-occupied homes facing the same zero effective tax rate as business capital, the two types of investment are not treated the same; business capital is subject to the cash-flow tax, and owner-occupied homes are exempt from it. A steel manufacturer that builds a factory claims an up-front deduction for its cost, whereas a homebuyer receives no up-front deduction for his purchase. On the other hand, the steel manufacturer is taxed on the output generated by the factory, whereas the homeowner pays no tax on the imputed rent produced by the home. This differential treatment causes no divergence in effective tax rates, however, because, as we have emphasized from chapter 2 onward, the business cashflow tax imposes a zero effective marginal tax rate on new investments. On a new marginal factory investment, the tax savings from the manufacturer's up-front deduction offset, in expected present value, the subsequent taxes on its output, and the business cash-flow tax therefore imposes no net burden.
Exempting owner-occupied homes from the business cash-flow tax promotes simplicity because the "cash" flow consists of imputed rent, which is difficult to measure. What are the other implications of this exemption? As we explained in chapters 2 and 3, the business cash-flow tax has three effects. First, it taxes capital in existence when the reform is adopted, subject to transition relief. Second, it taxes above-normal returns. Third, under uncertainty, it gives the government a proportional share in the risks associated with the future cash flows. Exempting owner-occupied homes from the business cash-flow tax therefore means that the value of existing homes escapes tax, as do above-normal imputed rental returns, and that the government does not share in the riskiness of future imputed rental returns on homes. We defer the existing-homes issue to the discussion of transition below, but we now assess the other two effects.
The failure to tax above-normal returns on owner-occupied homes is largely unavoidable, given that such returns manifest themselves as high imputed rents. If such rents cannot be accurately measured, then above-normal returns cannot be detected and taxed. Moreover, such returns are surely of limited significance. Whereas many business firms exploit innovative ideas and market power to earn large above-normal returns, few homeowners are likely to be able to generate above-normal housing services.
The risk issue may appear to be more important. Suppose that two equally costly houses are constructed. Because the prepayment method taxes only the original construction, it imposes the same tax on each house, which is ex ante reasonable because the future imputed rents on the two houses have the same expected market value. But subsequent events may cause their actual imputed rents to sharply diverge. The neighborhood around one home may unexpectedly improve while that around the other home may unexpectedly deteriorate. Or one home may burn to the ground shortly after construction. The prepayment method leaves the full risk of such events on individual homeowners, whereas the ex post method applies the cash-flow tax to homes and thereby spreads part of these risks. The latter method both gives homeowners an up-front deduction for the house purchase and then taxes them on the actual imputed rents they later receive; homeowners lucky enough to earn high imputed rents pay extra tax, whereas unlucky homeowners who earn low imputed rents pay little future tax, leaving them with a net tax saving from their up-front deduction.
Although the risk spreading done by the ex post method may seem desirable, it has no real effect so long as homeowners can achieve the same risk spreading in private markets, as discussed in the "Zero Revenue from Taxation of Risky Returns" box in chapter 3 (pages 51–52). There is no reason to think that the government can observe and insure against the relevant risks more cheaply than the private sector. Nearly all homeowners privately insure against casualty losses such as their home burning down; although there is almost no private insurance against general economic developments, the same problems that preclude private insurance in this area, such as the difficulty of measuring imputed rents, are likely to make public insurance impractical or undesirable.
In view of the availability of private insurance, we recommend that there be no deduction for casualty losses. Of course, because the X tax is real based, insurance recoveries for such losses are tax exempt.
At first glance, it may seem that the government could achieve some risk spreading without having to measure imputed rents simply by taxing gains and allowing loss deductions on home resales. For reasons explained in the "Taxation of Home Resales" box (pages 154–55), however, we advise against such a policy.
Transition. By exempting owner-occupied homes from the business cashflow tax, the prepayment method spares from tax the future housing services provided by homes in existence when the reform is adopted. In other words, these homes escape the X tax transition burden that we discussed at length in chapter 8, a fact noted by the Congressional Research Service (1996), Bradford (1996b, 140), Diamond and Zodrow (1998, 25; 2008, 231), and Viard (2000). On balance, sparing homeowners from the transition burden seems appropriate.
We concluded in chapter 8 that the government should not opportunistically target existing capital for taxation, because doing so would create harmful expectations about future policy. We also concluded, though, that existing capital should not be fully spared the transition burden of the X tax. Instead, we recommended that firms be allowed to recover the value of their unclaimed depreciation deductions, which would offset part, but far from all, of the transition burden.
Given this treatment of business capital, completely sparing homeowners from the transition burden may seem unduly lenient. Poddar (2010, 450) and Mintz (1996, 471), among others, condemn this outcome. But we believe that this disparate treatment can be justified on four grounds.
First, properly imposing the transition burden would require measuring and taxing the homes' imputed rents in each future year or, alternatively, appraising and taxing the market value of all homes on the reform date. These administrative costs should not be incurred unless there is a very strong case for imposing the transition burden.
Poddar (2010, 455) outlines a rough-and-ready way to impose a transition burden on existing homes. His proposal would tax gains on home resales, with the proviso that, on the first resale after reform of a home purchased prior to reform, the entire sales proceeds would be treated as gain. Of course, no transition burden would be imposed unless and until the home was resold. The tax policy would penalize sales, as discussed in the "Taxation of Home Resales" box. If a transition burden is to be imposed, it is not clear that this is a better way to do it than taxing imputed rents. In any case, the remaining reasons persuade us that no transition burden should be imposed.
Second, tax reform will reduce the value of owner-occupied homes to some extent even if they are spared the transition burden. Although exempting homes from the transition burden allows their values to remain equal to their replacement cost, their values will still decline because the replacement cost of homes will fall. By lowering the effective tax rate on new business investment to zero, while keeping the tax rate on new home construction at its current value of zero, reform will prompt (an efficient and highly desirable) reallocation of new investment away from housing and into the business sector. Like most other assets, the replacement costs of homes rise as more of them are produced and fall as fewer are produced. Accordingly, the reduction in the demand for homes in the wake of reform will drive down the replacement cost of homes and hence the values of existing homes. Note that this decline is due not to the introduction of the X tax but to the repeal of the income tax system that has given homes their tax advantage.
The magnitude of the price decline is uncertain. Diamond and Zodrow (1998, 27; 2008, 233–34) note that studies have reached a variety of conclusions about the impact of switching to a consumption tax on the value of existing homes, ranging from a 20 percent decline to almost no decline. The large price declines are unrealistic, however, because they assume extremely high adjustment costs. In their own simulations, Diamond and Zodrow (2008, 241–44) estimate a decline of 1.9 percent with no adjustment costs, 2.7 percent under moderate adjustment costs, and 3.3 percent with larger adjustment costs. As Diamond and Zodrow (1998) and the Congressional Research Service (1996) note, the most expensive homes will experience the largest price declines, because their high-tax-bracket purchasers receive the largest housing tax preferences today, and the end of those preferences will therefore have the largest impact on their demand.
In any case, existing homes will decline in value to some extent due to the reduction in housing demand. Because the value of mortgage debt will not decline, the percentage impact on homeowner equity will be larger than the price decline. For example, Diamond and Zodrow (2008, 241–44) assume that homeowners initially carry debt equal to 35 percent of home value and note that their estimated home price declines of 1.9, 2.7, and 3.3 percent translate into reductions of 2.9, 4.2, and 5.1 percent, respectively, in homeowner equity. Of course, homeowners who are more heavily leveraged would be harder hit.
As we discussed in chapter 8, the situation is quite different for most business capital. Even as reform reduces housing investment, it increases business investment, except perhaps for a few types of business capital that receive highly favorable treatment under today's tax system. Because the replacement costs of business capital rise as more is produced, reform will increase the replacement costs of most business capital, which would boost the value of most types of business capital were it not for the transition burden. We mentioned another reinforcing factor in chapter 8; many types of business capital are taxed on a deferred basis under the current tax system and their deferred tax liabilities make their current values less than replacement cost. In the absence of the transition burden, the forgiveness of the deferred tax liabilities would push their value up to replacement cost, causing further appreciation beyond that caused by the rise in replacement cost. In contrast, there are no significant deferred tax liabilities for owner-occupied homes.
In short, in the absence of the transition burden, reform would boost the value of most business capital but lower the value of homes. It is reasonable to impose a transition burden, softened by relief, to block an increase in the value of existing business capital while dispensing with a transition burden that would amplify a decline in home values.
Third, homeowners are generally less affluent than equity holders of business firms, who, as discussed in chapter 8, will bear the transition burden on business capital. Fourth, on a related note, it would be politically difficult to impose the transition burden on homeowners.
The change in the treatment of mortgage debt under the X tax does not represent a tax increase on housing, at least in principle, because the denial of a deduction for the borrower is accompanied by a removal of the tax on the lender. As Hall (1996, 84) notes, though, there may be a net tax increase in practice, because not all interest income is actually taxed under the current system. Existing mortgages would be subject to the transition policy discussed in chapter 8, with current-law treatment (taxation of the lender and deduction by the borrower) continuing to apply until and unless the debt was renegotiated.
Possible Homeownership Preference. There may be strong political pressure to maintain some type of preferential treatment for homeownership under the X tax. As discussed above, a preference may be justified if homeownership promotes social stability. But any such preference should encourage homeownership rather than the purchase of more expensive homes. If a subsidy is provided, it should take the form of a flat tax credit for the purchase of a home. Although the credit could be conditioned on purchasing a home with some minimum value (perhaps regionally adjusted), it should not otherwise rise with home value. One possible model is the current federal income tax credit, which is essentially equal to a flat $5,000, for first-time District of Columbia homebuyers with moderate incomes.26 Unlike this credit, however, any homeownership preference under the X tax should not be limited to first-time homebuyers, confined to a particular geographical area, or vary with income. The preference should also not depend on the extent to which the home purchase is debt financed.
Consumer Durables. Under current law, consumer durables are generally taxed similarly to owner-occupied housing. Owners cannot deduct purchases and are not taxed on imputed service flows. Capital gains are theoretically taxable, but rarely occur and are even more rarely reported. Capital losses are not deductible, except that an itemized deduction can be claimed for theft and casualty losses exceeding 10 percent of adjusted gross income. Unlike home mortgage interest, interest on durables loans is not deductible.
Under the X tax, the prepayment method should apply to consumer durables in the same manner as to owner-occupied housing. Because the only changes from current law are repeal of the (theoretical) tax on capital gains and the deduction for theft and casualty losses, the switch to the new tax system should be straightforward. Like owner-occupied housing, consumer durables will escape the transition burden, but the value of existing durables will still decline due to income tax repeal; production of new durables will fall due to the loss of their current tax advantage over business capital, which will lower replacement costs and the value of existing durables.
Production by governments, nonprofits, and households can and should be subject to the wage tax, but the cash-flow tax generally cannot be applied to these institutions.
Application of Wage Tax. Employees of the federal, state, local, and tribal governments and nonprofit organizations should be subject to the household wage tax on the same terms as other workers. This policy would be similar to current law, as those employees now pay income tax on their wages on the same terms as other workers.27
We begin by considering the application of the wage tax to federal government employees. This case is slightly more complex than the others, because the entity that pays these workers' wages also collects the tax imposed on the wages. If the household wage tax was strictly proportional, say at a 20 percent flat rate, it would be economically irrelevant whether the tax applied to federal employees. Exempting federal employees from the 20 percent tax would, in equilibrium, cause a 20 percent reduction in their before-tax wages. For all actual and potential federal employees, the exemption and the before-tax wage reduction would be exactly offsetting, leaving unchanged their after-tax wages and the federal government's net-of-tax cost of employing them.
Because the wage tax is progressive, though, budgetary balance and economic efficiency are improved by taxing federal government employees. Although an exemption would still allow the government to pay lower before-tax wages in equilibrium, the reduction in the government's wage costs would be outweighed by its loss of tax revenue. While the equilibrium wage reduction would be linked to the tax bracket of the marginal federal employee, workers with tax brackets higher than that of the marginal worker would disproportionately move to federal employment because they would receive the largest tax savings from the exemption. Because their tax savings would be greater than the reduction in before-tax wages, the exemption would raise their after-tax wages and increase the federal government's net-of-tax cost of employing them. In short, exempting federal employees from the wage tax would distort the allocation of labor in the economy by artificially drawing high-tax-bracket workers into federal employment and would provide windfall gains to those workers at the federal government's expense.
Economic neutrality also requires the application of the wage tax to employees of state, local, and tribal governments and nonprofit organizations. If these entities' employees were exempt from the wage tax, the entities would be able to pay significantly lower before-tax wages in equilibrium. As a result, in-house production by these entities would enjoy a large cost advantage over production by business firms, including production contracted out to firms by these entities. There is little case for providing a tax subsidy to production by these entities, at least not one limited to their in-house production.
Moreover, even if a subsidy was desired, a wage tax exemption would be an inefficient way to provide it. The analysis parallels the above analysis for federal employees. While the reduction in before-tax wages would be linked to the marginal employee's tax bracket, high-bracket workers would flock to work for state, local, and tribal governments and nonprofits, and the average employee's tax savings would exceed the reduction in before-tax wages. So, the exemption would distort labor allocation by drawing high-tax-bracket workers into this sector. Only part of the federal government's revenue loss would translate into reductions in employment costs for these entities; the remainder would show up as windfall gains to high-bracket workers employed by these entities. A wage tax exemption would have the same flaw as the current income tax exemption for municipal bond interest, which artificially entices high-bracket investors to hold municipal bonds and provides windfall gains to those investors at the federal government's expense.
Employees hired by households, such as babysitters, lawn-care workers, and so on, should also be subject to the wage tax. Tax compliance by such employees would presumably be no worse, and no better, than under the current tax system.
Exemption from Cash-Flow Tax. We recommend that the federal, state, local, and tribal governments and nonprofit organizations be exempt from the tax on business cash flow. Their production activities would effectively be subject to the same prepayment method that we have recommended for owner-occupied housing and consumer durables.
To begin, application of the cash-flow tax to the federal government would be pointless, as the same entity would both pay and receive the tax. And, for all these institutions, application of the cash-flow tax would be administratively infeasible, because the "cash" flow would consist of the imputed value of the services these entities provide. Although we have judged the measurement of imputed rent on owner-occupied homes to be impractical, the task could be attempted with the aid of real estate professionals and, as discussed above, has been attempted in some other countries. But measurement of the imputed value of police and fire production and environmental regulation transcends the impractical and enters the realm of the impossible.
The economic implications of exempting these entities from the cashflow tax are similar to the implications, discussed above, of the corresponding exemption for owner-occupied housing and consumer durables. Because the cash-flow tax imposes a zero effective marginal tax rate on new investment, exemption gives these entities no competitive advantage. The exemption has the same three effects it did for owner-occupied homes; it causes these entities' above-normal returns to escape tax, spares these entities' existing assets on the reform date from the transition burden, and prevents the federal government from spreading the risk of these entities' returns.
Each of these effects is either unavoidable or desirable, or both. Because the returns generated by these entities' operations are impossible to measure, there is no hope of detecting and taxing above-normal returns or spreading risk. And there seems little reason to impose the transition burden on these entities. Imposing the transition burden on state, local, and tribal governments would merely transfer wealth from the American people in their capacity as state, local, and tribal taxpayers and service recipients to the American people in their capacity as federal taxpayers and service recipients. Imposing the transition burden on nonprofit organizations might burden the recipients of their services, an unappealing result.
If desired, the cash-flow tax could be applied to commercial enterprises operated by governments and nonprofits, because the cash flows of these enterprises can be easily measured. Some of these enterprises currently pay the unrelated business income tax. Nevertheless, we recommend that tax not be imposed on these enterprises. There is little reason to impose the transition burden on these entities, and it may be difficult to separate the commercial enterprises from the entities' other operations.
Similarly, the cash-flow tax would obviously not apply to household employers. For example, the purchase of a lawn mower would be treated as the purchase of a consumer good, even if the purchaser actually hires neighboring children to use it to mow the lawn; the purchaser would not file a business tax return on which he would expense the purchase of the mower, deduct the wage payments, and pay tax on the imputed value of the mowing services. As a result, existing mowers owned by household employers would escape the transition burden, an outcome that seems eminently satisfactory.
Under the above recommendations, different sectors of the economy will operate under different tax rules. The business sector will be subject to the cash-flow tax, whereas the nonbusiness sector, consisting of owner- occupied homes, consumer durables, governments, nonprofit organizations, and households will be exempt from it. We now discuss the complications that arise when assets cross the boundary between the two sectors.
General Principles. Because the cash-flow tax imposes a zero marginal effective tax rate on new investment, there is generally no tax incentive for investments to migrate from one sector to the other. It may seem at first glance that such an incentive exists for assets with above-normal returns. If such an asset is initially held by a business firm, wouldn't there be a tax incentive to sell it to, say, a nonprofit organization? Because the nonprofit, unlike the firm, is exempt from the business cash-flow tax, moving the asset into the nonprofit's hands appears to allow the subsequent above-normal returns to escape the tax that they would have faced if the asset remained in the firm's hands.
This analysis is flawed, though, because it overlooks the fact that the firm also pays cash-flow tax on the full proceeds of asset sales, including sales to nonprofits. If the firm retains the asset, it is taxed on the asset's future cash flows as they arise; if the firm sells the asset, it is taxed on the asset's sales price at the time of sale. Because the asset's price is equal in expected market value to its future cash flows, there is no tax incentive to sell. Although cash flows arising after the sale to the nonprofit are not taxed when they arise, that result is perfectly appropriate because the cash flows were pretaxed (in expected present value) at the time of the sale.
For the same reason, there is no tax disincentive to a firm purchasing an asset with above-normal returns from a nonprofit. To be sure, such a sale causes subsequent cash flows to be taxed as they arise, which they would not have been if the asset had remained in the nonprofit's hands. But the firm is allowed to deduct the asset's full purchase price at the time of purchase. The tax savings from that deduction offset the future taxes because the purchase price is equal, in expected market value, to the future cash flows. So, if such a purchase is desirable for nontax reasons, the X tax does nothing to inefficiently preclude it from occurring.
We need to consider a few special cases where things are a little more complicated.
Sales in Anticipation of Reform. The above analysis demonstrates the importance of an asset's location on the day that reform takes effect. If the asset is in the business sector on the effective date, its future cash flows are irrevocably subject to tax. If the asset stays in the business sector, the cash flows are taxed as they arise; if the asset is sold to the nonbusiness sector, the future cash flows are pretaxed at the time of sale. On the other hand, if the asset is in the nonbusiness sector on the effective date, its future cash flows are irrevocably shielded from tax. If the asset remains in the nonbusiness sector, the cash flows are not taxed; if the asset is sold to the business sector, the cash flows are taxed as they arise, but the taxes are fully offset by the purchaser's up-front deduction at the time of sale. In short, the transition burden applies to those assets, and only those assets, that are in the business sector on the effective date, regardless of subsequent sales. As explained above, there are no tax incentives for sales between the two sectors after the reform takes effect.
Without special rules, though, there would be an incentive for assets to migrate into the nonbusiness sector before reform takes effect. Fortunately, one of the transition rules specified in chapter 8 addresses this concern. During a transition period beginning on the date at which reform becomes a serious possibility, firms are allowed to expense asset purchases and are fully taxed on asset sales.
The next situation cannot be resolved quite as easily as this one.
Sales between Related Parties. The above analysis relied heavily on the asset's price being equal to the expected market value of its future cash flows. That equality holds for the market price, which is the price at which transactions between unrelated parties can be expected to occur. But it may not hold for the prices at which related parties choose to contract with each other, and still less the price at which a party transacts with itself.
For example, there is no net tax incentive or disincentive for a homeowner to sell his house at a market price to an unrelated company that will convert it to rental use; as explained above, the tax savings from the rental company's expensing of the purchase is equal to the expected market value of the taxes on the future cash flows that the home will generate as a rental property. But what happens if a taxpayer moves out of his house and begins renting it to other parties? The conceptually correct treatment is to allow the taxpayer to deduct the market value of the house, which his rental business has effectively purchased from his homeowner alter ego, and to tax the subsequent cash flows generated by the rental business. The problem, of course, is that the taxpayer has an incentive to overstate market value, which cannot be readily observed. Conversely, a taxpayer who ceases to rent out a house and begins living in it should pay tax on the full market value of the house, which his rental business has effectively sold to his homeowner alter ego. Here, the taxpayer has an incentive to understate market value.
Strict rules, strictly enforced, will be necessary to prevent abuse. In the first situation described above, the taxpayer should be required to use valuations that are commensurate with the past purchase price of the house and should also be required to retroactively adjust the valuation upward if subsequent cash flows are high relative to the valuation initially used. In the second situation described above, the value should be commensurate with the past rental cash flows. Although such rules may be too harsh in some cases and may deter some legitimate transactions, stringency is necessary to prevent abuse.28
We now discuss the interaction of the X tax with state, local, and tribal tax systems and the municipal bond market.
Deductibility of State, Local, and Tribal Taxes. The current federal income tax allows an itemized deduction for income and property taxes paid to state, local, and tribal governments. Taxpayers may elect to deduct sales taxes in lieu of income taxes. Firms may generally deduct their tax payments as business expenses. The propriety of this treatment continues to be debated, with some proposing to repeal the itemized deduction and some proposing limits on it, such as a ceiling or a floor on the amount deducted.
The appropriate federal tax treatment of such taxes may depend on how they are used. Allowing a deduction for taxes that finance a municipal golf course may be as inappropriate as allowing a deduction for fees paid to a private golf course. On the other hand, a deduction for taxes that finance productive infrastructure may be as appropriate as a deduction for business purchases. Taxes that finance transfer payments should probably not be deductible if the recipients are not taxed on the transfer payments, as they are not under the proposals we outlined in chapter 4.
We need not resolve these issues. The various treatments for these taxes that have been adopted, or proposed, under the current income tax could be adopted, with little change, under the X tax. Accordingly, this issue can, for the most part, be addressed separately from the question of whether to adopt our proposed tax reform.
It would be difficult, however, to allow a deduction for state and local individual income taxes imposed on investment income, given that such income would not be subject to individual tax under the X tax system. This issue would not arise if state and local governments conformed to the federal X tax design, as discussed below, and therefore ceased to impose individual income taxes on investment income.
Conformity of State and Local Tax Systems. Today, most states have an individual income tax. The state income tax base is usually defined, at least in part, by reference to either the current or a lagged version of the federal income tax base. A generally similar pattern prevails with respect to state corporate income taxes. It would be difficult for states to continue taxing corporate income or individual capital income after the X tax was adopted at the federal level. In view of the inefficiency of such taxes, that is all to the good. State taxes on capital income make even less sense than federal taxes on capital income. To maximize the simplicity and efficiency advantages of reforms, states should be offered financial incentives to reconfigure their taxes as household wage taxes and business cash-flow taxes. Of course, states would be free to set their own rate schedules and to choose which personal tax deductions and credits to allow on household wage tax returns.
Municipal Bonds. Under the current federal income tax, interest income on bonds issued by state, local, and tribal governments ("municipal bonds") is generally tax exempt. Some restrictions apply to bonds issued by these governments for private use. This exemption is obsolete under the X tax, because all interest is exempt from tax.
The removal of the tax on interest income throughout the economy should cause interest rates to decline in equilibrium, which would benefit all borrowers, including municipal bond issuers. Nevertheless, these issuers would lose the tax advantage that they have over other borrowers and would capture a smaller share of the pool of savings.
As mentioned above, however, the exclusion of municipal bond interest is an inefficient subsidy because it provides windfalls to high-bracket investors. If Congress wishes to continue subsidizing municipal borrowing, it can adopt a more efficient subsidy by making direct payments to issuers, as was done for the Build America bonds issued in 2009 and 2010.
The nonbusiness sector poses few difficult issues for the X tax. Exempting production in this sector from the business cash-flow tax promotes simplicity and is consistent with consumption tax principles.
BOX
TAXATION OF HOME RESALES
Under the prepayment method, there is no deduction for home purchases and no tax on home sales. One implication of this treatment is that home resales have no tax implications. In contrast, the ex post method would tax the seller on the full sales price and would allow a corresponding deduction for the purchaser.
Some proposals, such as those of Poddar (2010) and Conrad (2010), would modify the prepayment method to include a tax on home resales. The tax would apply only to the gain on the sale, although some proposals, such as Poddar's and Conrad's, would treat the entire sale proceeds as taxable gain on the first resale after reform of a home purchased before reform. The proposals vary in their treatment of resales that generate losses. No deduction would be provided for the purchaser.
The goal of such a policy is to tax above-normal and lucky returns on homes while avoiding the complications of taxing imputed rent. If imputed rents on a home were higher than normal and were expected to remain so, the home price would appreciate to an unusual extent. Under these proposals, the government would tax away part of this price appreciation if and when the homeowner sells the home. These proposals are undesirable, however, for two reasons.
First, because this policy does not measure above-normal and lucky returns correctly, it can overtax imputed rents. The fact that a home experiences a capital gain does not mean that the owner has received an above-normal or lucky return. If a home's imputed rent rises over time, its price also rises over time, even if the owner receives only a marginal return. For example, assume that the marginal rate of return is 5 percent, and consider a home for which the imputed rent is $10,000 today and rises 1 percent per year. This home sells for $250,000 today, and the price rises by 1 percent per year. Each year, the homeowner receives imputed rent equal to 4 percent of the home value and experiences a 1 percent capital gain, thereby earning the 5 percent marginal return. Under consumption tax principles, no tax should be imposed on these returns. The original $250,000 sale price, which the prepayment method taxes at the time of construction, fully reflects the present value of the future imputed rents; note that a stream of rents that begins at $10,000 today and rises 1 percent per year has a present value of $250,000 when discounted at 5 percent per year. Any additional tax would overtax the imputed rents and violate the neutrality of consumption taxation.
In fact, there is no way to accurately measure the size of the above-normal or lucky returns without measuring imputed rents, precisely the problem that this policy is meant to sidestep. A cautious version of the policy might tax only the appreciation in excess of the normal rate of return; assuming that imputed rent cannot be negative, a homeowner with a capital gain greater than the normal rate of return must have received an above-normal or lucky return. But if some lucky returns are to be taxed, neutrality requires that there be some offsetting relief for unlucky returns. Of course, providing a deduction to homeowners who sell at a loss would introduce further complications.
Another problem is that the tax would generate a lock-in effect similar to that induced by capital gains taxation today. Because lucky homeowners would pay this tax only if and when they sold their homes, many of them would be deterred from selling, disrupting efficient allocation of the housing stock and impeding mobility. If a deduction was provided for homeowners who sell at a loss, owners whose homes have declined in value would have the opposite incentive; they would be prompted to sell in order to claim the loss deduction, even if a sale would otherwise not be beneficial. Of course, any policy to tax gains would introduce administrative complications, particularly because home improvement costs would then need to be tracked and capitalized.
As discussed above, Congress has exempted the first $250,000 ($500,000 for couples) of capital gains on homes from tax under the current system. It would be peculiar to expand taxation of gains on homes when switching from the income tax to the X tax, given that such taxation has far less of a useful role to play under X tax principles than under income tax principles.