Treasury OTA Says Estate Tax System Can Be Made More Efficient

Taxes are typically levied to raise revenues to fund government programs. Yet at times the government provides tax relief to special groups or to certain activities. Through these preferential treatments, the tax code is used to alter the allocation of resources and shape economic activities. These preferences are often labelled as tax expenditures as they serve a function akin to that of governmental outlays.¹ Although most frequently associated with income taxes, tax expenditures can also be measured as a deviation from a baseline tax system applied to other taxes, including as estate taxes. Indeed, the Administration Budget included tax expenditure estimates for the estate tax in years 1994-2003.

This paper examines the extension of tax expenditure analysis to the estate tax. It describes current features of the estate tax system and summarizes the tax expenditure treatment of provisions under the estate tax. Using a broad definition of what constitutes a tax expenditure, the analysis illustrates the revenue effects of provisions that narrow the estate-tax base are large. In effect, the current system is characterized by high tax rates and a narrow base.

More specifically, and using the baseline employed in previous published Budgets, the deduction for charitable bequests represents by far the largest tax expenditure under the estate tax. For returns filed in 2014, the deduction reduced the gross estate by about eleven percent. Using a broader definition of tax expenditures, the unlimited marital deduction for spousal bequests plays a major role in narrowing the tax base. Indeed, it reduced the gross estate by close to 34 percent in 2014. When combined with the unified credit, which provides an exemption of $5 million (indexed), these provisions have the effect of reducing the effective estate tax rate from a statutory rate of 40 percent down to 10 percent.

The paper is organized as follows. Section 2 provides a brief description of the estate tax and its contribution to federal tax revenues. Section 3 identifies tax preferences under the estate tax and provides estimates of the foregone revenue. The intent here is to follow along the lines suggested by David Bradford, and view the analysis as simply comparing different tax rules. Section 4 discusses the implications of broadening the estate tax base. Section 5 concludes.

2. History of Tax Expenditures under the Estate Tax

Tax expenditures were initially developed in the context of income tax analysis. These expenditures were conceived of by Stanley Surrey as deviations from the treatment that would be offered by an idealized income tax. Although most frequently associated with income taxes (individual and corporate), the concept of measuring a tax expenditure as a deviation from a baseline tax system also can be applied to other taxes, such as estate taxes. Indeed, the 1974 Budget Act, which mandated accountability of tax expenditures in the federal budget, was very general in its definition of a tax expenditure. For each federal tax, the 1974 Act called for the documentation of preferences and the measurement of the revenue loss that each preference causes.

Nonetheless, for many years tax expenditures focused exclusively on individual and corporate income taxes. The 1968 annual report of the Department of the Treasury, which contained the first official tax expenditure analysis, considered only individual and corporate income taxes. That tradition was followed for many years by the presentation in the President's Budget and by the parallel presentation by the Congressional Joint Committee on Taxation.²

That changed in the early 1990s, when the Treasury published tax expenditures for the estate tax. Paving the way for this change was a revision of Treasury's concept of a baseline tax system against which tax expenditures were measured. Traditionally, the tax expenditure budget (TEB) employed a modified Haig-Simons (HS) measure of income as the baseline against which special treatment is measured. HS is a comprehensive, accrual based, measure of income as the sum of consumption plus the change in net worth. The reliance on a HS base, an income concept, made it difficult for many to consider the estate tax in preparing the annual TEB estimates. By 1982, however, Treasury had moved to a different, and perhaps more practical, baseline concept for tax expenditures. The new baseline, the "Reference Law," looks for general tax rules that apply to all taxpayers. What is most important about these rules is that they are general, not that they conform to any income tax ideals. Once these general rules are identified, departures from them for reasons other than simplification in administrating the tax code are considered as tax expenditures.³ The same Reference Law concept is easily extended to taxes other than income taxes such as the estate tax.⁴

Using the Reference Law baseline, the first estate tax expenditure budget was published in the annual Budget for Fiscal Year 1994, alongside those of the individual and corporate income taxes. The Budget continued to publish tax expenditures for the estate tax for nine years.⁵ But for the Fiscal Year 2003 Budget, the estate tax expenditure section was dropped, a change attributed at the time to a lack of agreement on an appropriate baseline to define estate tax expenditures.⁶ Estate tax expenditures have remained out of the Budget since then.

As with the income tax expenditure budget, there are good reasons for considering estate tax expenditures. The analysis can help inform tax policy by pointing out some implications of alternative tax rules. There might be interest, for example, in knowing how much revenue is lost and who benefits from the various exceptions and exclusions allowed under the estate tax. This of course would be relative to an estate tax base that was broader in the sense of not allowing those exclusions and exceptions. Such an analysis need not be included in the official presentation of the President's Budget. But preparing such analysis can provide helpful information in evaluating the estate tax.

In considering baseline problems for an estate tax analysis, it is worth remembering that the baseline for any tax expenditure estimate is somewhat arbitrary and frequently controversial. For income tax analysis, even HS has ambiguities. For example, consider the employer exclusion for contributions for medical insurance premiums and medical care which represents the largest tax expenditure.⁷ Should this exclusion be treated as a tax preference or do expenditures on employee health maintenance represent an investment in human capital that should be deductible as in the case of maintenance expenditures on plant and equipment?⁸

As emphasized by David Bradford (1989? Double check cite), perhaps a tax expenditure estimate is no more than a revenue estimate for changing the taxation of a given income or expense flow. If so, then it also is no less than this. It can be useful to have this information, as well as other information on the effects of one form of tax treatment vs another (e.g., the effect on the distribution of the tax burden, perhaps, or on the likely behavioral reaction to different treatments). But there has never been any agreement on the precise baseline tax system to use for the analysis nor has there has been agreement on the normative dimensions of the analysis. Indeed, many tax expenditures potentially represent good policy.

Against this background, problems with defining the baseline for an analysis of estate tax expenditures simply are not persuasive rationales for shunning the topic entirely. This would be true even if the problems were identified by critics, which they are not. Huge baseline problems have plagued interpretation and use of income tax expenditures. Certainly, estate tax expenditures fit into David Bradford's idea that a tax expenditure estimate is a revenue estimate for replacing one tax rule with another. To find such an analysis helpful, one need not think that the baseline's tax rules define the ideal estate tax. Rather, one must just see value in comparing estate tax alternatives in formulating tax policy.

3. The federal estate tax

3.1. Brief history

Congress had contemplated an inheritance tax in leading up to the enactment of the Revenue Act of 1916, an Act aimed at raising revenue to fund war preparations. Instead Congress opted for the estate tax variant of the inheritance tax. The choice reflected on the administrative simplicity of the estate tax over a pure inheritance tax.⁹

The newly introduced estate tax targeted the wealthy, and was generally viewed as means to preempt the erosion of the income tax, as well as enhancing the progressivity of the tax system and controlling wealth concentration.¹⁰ Not addressed at the time was the possibility of easily avoiding the estate tax through lifetime gifts. Individuals may transfer funds to their children or grandchildren in life, thereby minimizing wealth transfer taxes at death. Accordingly, a gift tax was introduced in 1932 in an attempt to reduce estate and income tax avoidance.¹¹ The stated purpose of the proposed gift tax in the House Committee report was "To assist in the collection of the income and estate taxes and prevent their avoidance through the splitting of estates during the lifetime of a taxpayer." As an example, by making gifts to beneficiaries or to trusts created on their behalf, income taxes can be reduced if the donor were in a higher tax bracket when compared with those of the recipients, and of course the estate tax would be avoided entirely.

Because transfer taxes can be reduced by leaving bequests to grandchildren rather than to the children, the Generation Skipping Transfer Tax (GSTT) was introduced in 1976. In general, this applied to transfers from trusts and did not affect direct transfers from donors to the grandchildren. The Tax Reform Act of 1986 extended the GSTT to all transfers, direct and indirect, fully phased in by 1989. With the exception of the rules that applied in 2010, the basic structure of estate and gift taxes has remained the same ever since.

The contribution of estate and gift tax revenues to federal receipts grew rapidly in the first decade of the inception of the transfer tax. As its scope narrowed, with attempts to repeal these taxes altogether in the 1920s, tax revenues from this source decreased. This trend was reversed in the 1930s when estate and gift taxes contributed nearly 10 percent of total receipts. After WWII, the income tax base was expanded, and the relative contribution of estate and gift taxes diminished greatly. With reductions in rates and base in 1982, the contribution of the estate tax further declined. In fiscal year 2015, Federal estate and gift taxes generated $19.2 billion in revenues, about one percent of federal revenues. Figure 1 provides the historical trend of estate and gift tax revenues and their contribution to total government receipts from the inception of the tax through 2015.

3.2 Details on the Current Structure of the Tax

A wide range of assets are included in the base of the estate tax, such as cash, bonds, stocks, farms and business, life insurance, and retirement assets, among others¹² . The value of businesses and farms is reduced by special allowances for estates meeting certain conditions. This special use allowance was capped at $1.1 million in 2015, or $750,000 indexed for inflation as of 1997, and was available to estates where 35 percent or more of the value of the estate was in the form of business and farm assets, and where the beneficiaries agreed to engage in these activities for a specific period of time. The gross estate is reduced by a set of deductions and exemptions in deriving the tentative taxable estate. The largest deduction is that for spousal bequests, followed by that for charitable bequests, and that for state estate and inheritance taxes.

The gross estate reduced by deductions, the tentative taxable estate, is expanded by adding lifetime taxable gifts. The tax rate schedule in Table 2 is then applied to the final taxable estate to derive the tentative estate tax. In 2015, a tax credit of $2,117,800 is applied which exempts tax estates with taxable values below $5,430,000. A tax credit for gift taxes paid during life is also provided so as to avoid double taxation of lifetime gifts. It should be noted that unlike the estate tax, the gift tax applies on a tax exclusive basis.¹³

Under limited circumstances, part of the estate tax liability can be deferred for close to 15 years from date of death. The fraction of taxes deferred is equal to the ratio of the value of the qualifying interest in a closely held business to the adjusted gross estate, provided that this ratio is in excess of 35 percent. Qualifying interest includes the value of proprietorships, and corporate stock or partnership interest if at least 20 percent of the voting stock or partnership assets is included in the estate, or if the corporation has no more than 15 shareholders. A less generous deferral is also available to certain estates, such as those with severe liquidity constraints, and is granted at the discretion of the Commissioner of the Internal Revenue Service.

3.3 Profile of estate taxpayers

About 11,800 estate tax returns were filed by estates in excess of $5 million in 2014, as shown in Table 2. The size of the estate here is defined as the sum of the gross estate plus cumulative lifetime taxable gifts; these gifts exclude transfers below the annual exclusion as well as gifts for education and medical purposes. About 65 percent of the estates were in the range of $5 to $10 million, reporting about $5 billion from a total of $169 billion in gross estates. In contrast, about 13 percent of the estates were in excess of $20 million reporting $91 billion in gross estates.

These individuals also made $20 billion in taxable gifts, with 45 percent of it made by the wealthiest of estates. When reduced by debts outstanding and estate expenses, the gross estate is reduced to $158 billion in net worth. When further reduced by the deductions for spousal and charitable bequests as well as state death taxes, but expanded by lifetime taxable gifts, the amount subject to tax, labelled as the adjusted taxable estate, is about $100 billion. The latter represents a reduction in the gross estate by more than one half for the wealthiest estate group.

The gross estate is reduced by $87 billion in deductions. These include deductions for estate expenses and debts, spousal bequests, charitable bequests, and state death taxes. About one half the estates were for married decedents reporting $56 billion in spousal bequests, accounting for over 60 percent of all deductions. About one half of these reported deductions were for transfers in the form of Qualified Terminable Interest Property (QTIP) trusts. In a QTIP, the surviving spouse typically has only a right to income for life from the trust with no control on the ultimate dissolution of the trust and disposition of the underlying assets.

About $18 billion in deductions for charitable bequests are reported. Unlike the income tax, charitable bequests are fully deductible. As with the marital deduction, as well as its QTIP component, the reported deductions are top heavy, as the bulk is claimed by those with estates in excess of $20 million. About $2.4 billion in deductions for state death taxes are claimed. These too are top heavy, as observed from the figures in Table 2.

The reported estate tax liability in 2014 was $16.4 billion. This reflects to application of the tax rate schedule in Table 1, and reducing the ensuing tentative tax liability by the unified credit of $2,117,800 which effectively provides an exemption of $5,430,000. Of the reported liability, $788 million were deferred and payable in installments. When compared to the net worth of the estate, i.e., gross estate less estate expenses and debts, the effective tax rate is 10 percent. Even for the wealthiest class, those with estates over $20 million, the effective tax rate is only 14 percent, a far cry from the statutory tax rate of 40 percent. This wedge between the effective and statutory tax rates can be explained by the various deductions and the size of exempted estate (by virtue of the unified credit).

4. Baseline definition and identifying tax preferences

A set of tax expenditure estimates for the estate tax can be motivated entirely on their own terms, as the effects of movements away from a (fairly) comprehensive estate tax.¹⁴ There need be no overriding, a priori normative support for the comprehensive estate tax one choses as the baseline. Rather, the analysis can be seen simply as one that examines the effects (e.g., revenue, beneficiaries) of changes in the estate tax.

We now turn to our definition of a comprehensive estate tax. This tax includes the broad outlines of the existing estate tax, but not what we take to be the existing tax's special rules for particular assets and taxpayers.

4.1. A Baseline Estate Tax

The reference tax rules for the unified transfer tax from which departures represent tax expenditures include:

Definition of the taxpaying unit.

The payment of the tax is the liability of the transferor whether the transfer of cash or property was made by gift or bequest

Definition of the tax base.

The base for the tax is the transferor's cumulative, taxable lifetime gifts made plus the net estate at death. This excludes all spousal transfers. Gifts in the tax base represent the cumulative sum of annual transfers in excess of the annual exclusion ($14,000 in 2015, indexed) to any donee. Further excluded from taxable gifts are payments on behalf of family members' educational and medical expenses, as well as gifts to political organizations.

Property valuation.

In general, property is valued at its fair market value at the time it is transferred. This is not necessarily the case in the valuation of property for transfer tax purposes. Because of fluctuations in the prices of assets in the marketplace, executors of estates are provided the option to value assets at the time of the testator's death or at six months later. For the tax expenditure identification and measurement, the value reported for tax purposes is chosen as the baseline. This is typically the lesser of the value established at death and that six months later.

Tax rate schedule.

A single graduated tax rate schedule applies to all taxable transfers. This is consistent with the name of the ''unified transfer tax'' that has replaced the former separate gift and estate taxes.

The tax rates vary from 18 percent on the first $10,000 of aggregate taxable transfers, to 40 percent on amounts exceeding $1 million.

A lifetime credit is provided against the tax in determining the final amount of transfer taxes that are due and payable. For decedents dying in 2015, this credit allows each taxpayer to make a $5,430,000 tax-free transfer of assets that otherwise would be liable to the unified transfer tax.

Time when tax is due.

Donors are required to pay the tax by April 15^th following the year gifts are made. The generation-skipping transfer tax is payable at the time the gift tax is due. The net estate tax liability is due within nine months after the decedent's death; the GSTT is due then as well.

4.2 Preferences

Given the baseline tax structure above, a number of preferences can be identified. The latter are provisions that have been typically included in previous published Budget reporting estate tax expenditures. These are reported below with a brief description of how they deviate from the baseline tax, followed by (static) revenue loss estimates.

a. Donations of conservation easements.

Bequests of property and easements (in perpetuity) for conservation purposes can be excluded from taxable estates. The use of the property and easements must be restricted to one or more of the following purposes: outdoor recreation or scenic enjoyment for the general public; protection of the natural habitats of fish, wildlife, plants, etc.; and preservation of historic land areas and structures. Conservation gifts are similarly excluded from the gift tax. No more than 40 percent of the value of land subject to certain conservation easements, up to a maximum amount of $500,000, may be excluded from taxable estates.¹⁵ The revenue consequences of this this exclusion are negligible as less than $2 million in reduced tax liability is computed from estate tax returns filed in 2014; about $5 million in exclusions were reported on tax returns.

b. Special-use valuation of closely-held farms and businesses.

Farmland owned and operated by a decedent and/or a member of the family may be valued for estate tax purposes on the basis of its ''continued use'' as farmland if: (1) the value of the farmland is at least 25 percent of the gross estate; (2) the entire value of all farm property is at least 50 percent of the gross estate; and (3) family heirs to the farm agree to continue to operate the property as a farm for at least 10 years. The special-use valuation rule available for family farms is also available for nonfarm family businesses. To be eligible for the special-use valuation, the same three conditions previously described must be met. Up to $750,000 of the difference in the market value and the "continued use" value, indexed for inflation as of 1997, can be excluded from the estate. About $90 million in exclusions were claimed by estate tax filers in 2014, which led to an estimated revenue loss of about $35 million.

c. Tax deferral of closely held farms and businesses.

The tax on a decedent's farm can be deferred for up to 14 years from the due date if the value of the farm is at least 35 percent of the net estate. For the first 4 years of deferral, no tax need be paid. During the last 10 years of deferral, the tax liability must be paid in equal annual installments. Throughout the 14 year period, interest is charged at a special, favorable rate. Estate tax filers in 2014 deferred $788 million in taxes, as reported in Table 3.

d. Charitable contributions.

Bequests to educational institutions, to health institutions, and to charitable, religious, and certain other nonprofit organizations, can be deducted from taxable estates. As reported earlier, the deduction claimed by filers in 2014 was $18 billion. Absent any behavior, its repeal would increase the size of taxable estates by 18 percent. The implied revenue loss from this deduction is about $7 billion.

e. State and local death taxes.

A deduction is allowed for State taxes in computing federal taxable estate. The deduction claimed by filers in 2014 led to an estimated $0.9 billion in foregone revenues.

While there should be little disagreement on defining the above provisions as tax expenditures, there are ambiguities in such designation in the case of a number of other provisions. Examples include the unlimited marital deduction, the exclusion of certain life insurance proceeds, and the tax exclusive nature of the gift tax. Nonetheless, while there may controversy over whether these tax reducing provisions are part of an idealized estate tax, repealing them would broaden the estate tax base and allow for a lower estate tax rate.

f. The unlimited marital deduction:

Consider the marital deduction which provides for an unlimited deduction for spousal bequests. Under the income tax, a surviving spouse is treated as a separate taxpaying unit after death. Gains accrued and accumulated during life on assets inherited from deceased spouse are stepped up at death. The surviving spouse is then able to realize these gains free of any capital gains taxes. In addition, spousal bequests are fully deductible in computing the taxable estate. Thus, inherited assets can be sold by a surviving spouse without being subject to either income or estate taxes. The step up in basis is identified as a tax preference item under the income tax. Should a similar treatment be extended to the marital deduction under the estate tax?

Prior to 1982, the deduction for spousal bequests was generally limited to one-half the net worth of the estate.¹⁶ The deduction was first introduced in 1948. Previously only the surviving spouse's interest in community property was excluded from the estate. The Act of 1948 introduced the deduction to one-half of the estate, motivated by a desire to equalize the tax treatment of transfers to spouses in common law states with those in community property law states (cite). The motivation for the unlimited marital deduction in the ERTA of 1981 was that husband and wife ought to be treated as one unit. In late 2010, a temporary provision was introduced to allow surviving spouses to carryover the unused portion of the estate tax exemption and apply it to their estates. Legislation enacted in 2013 made this portability of the exemption permanent. Once again, the justification was that husband and wife should be treated a one taxpaying unit, even when separated by death and time.

Repealing the marital deduction yields an estimated $20 billion in tax revenues using 2014 tax return information.¹⁷ This change would take us to the pre-1948 law. If we were to go back to the pre-1982 law, when spousal bequests were generally deductible up to about one half of the estate, then the revenue gain would decline to about $8 billion. Unlike repealing the deduction for charitable bequests, behavioral effects that have tax revenue consequences are likely to be smaller. As the experience prior to 1982 suggests, individuals may opt to make taxable transfers directly to the heirs, and set spousal bequest at the deduction limit.¹⁸

It is obvious that the views on the marital deduction have evolved over time. While these evolving views make it difficult to classify the deduction as a tax expenditure, the portion applied to spousal bequest in the form of QTIPs might fall more squarely in the category of a tax expenditure. As noted above, about one half the marital deduction reported on tax returns filed in 2014 was applied to transfers through QTIPs which are classified as spousal bequests even when the surviving spouse may have limited access to the proceeds.¹⁹ While the legislative intent of the marital deduction is to treat couples as one taxpaying unit, the surviving spouse may have little control over the principal in a QTIP. Thus, the transfer into the trust might be seen less as a transfer to the spouse and more as a transfer to the ultimate trust beneficiaries

Beginning in 2011, the benefit of the marital deduction was enhanced by expanding the availability of the unified credit. More specifically, the estates of married decedents were allowed to carry forward the unused portion of the $5 million (indexed) estate exemption, by virtue of the unified credit. This portability provision, allowed the surviving spouse's estate to claim the unused portion of the exemption regardless of the size of the estate of the first to die.²⁰ About $1 million in Deceased Spouse Unused Exclusion (DSUE) was reported on estate tax returns filed in 2014,²¹ which represents about $0.4 billion in additional benefits to married decedents. This benefit, however, is likely to grow over time as the stock of unclaimed exemptions by smaller estates grows over time.

g. Exclusion of certain life insurance proceeds:

The treatment of life insurance proceeds presents another challenge because the proceeds from policies owned by the deceased are taxable whereas those owned by others on the life of the deceased are excluded. As an example, proceeds from policies owned by the beneficiaries or held in trusts for their benefit are not included in the gross estate of the insured decedent and escape taxation. At times, business partners may acquire life insurance policies on the life of one another in order to insure against their mortality risk. This can be viewed as a legitimate business use, and raises the question of whether these partners should be treated differently than direct holders of life insurance policies. Repealing the exemption raises $1.8 billion in revenues using 2014 data, as shown in Table 3; about $6 billion in proceeds from life insurance policies were excluded (Table 2). Because they are subjected to tax, individuals may no longer employ life insurance policies as a way to minimize their estate tax liabilities, and are likely to seek alternative vehicles to reduce the burden of the estate tax.

h. Tax exclusive nature of gift tax:

Estate and gift taxes are unified. They share the same tax rate schedule and unified tax credit which applies to cumulative transfers. Yet because the gift tax applies on a tax exclusive basis, lifetime transfers receive a preferential treatment when compared to bequests. At a statutory estate tax rate of 40 percent, for instance, the effective gift tax rate is about 28.6 percent or 0.4/(1+0.4).²² About $2 billion in gift taxes were paid in fiscal year 2014, and if gifts were to be taxed on an inclusive basis, an additional $0.7 billion in tax revenues could have been raised, as reported in Table 3.

i. Credit for previously paid federal estate taxes:

The estate tax provides a tax credit for previously paid estate taxes. This credit is phased out over ten years, in two year intervals, from the date of death. This credit was introduced in 1954, but it has its roots in the Act of 1918 which allowed a deduction for taxes paid on property inherited within five years from the transferor's date of death. About $50 million in tax credits were claimed by 2014 estate tax filers. Its small size should not come as a surprise given the distance between the deaths of donors and recipients. This figure can become larger if the unified credit is dramatically reduced or the marital deduction scaled back. But its relative size will continue to be small given the distance in deaths.²³

While the credit reduces the cumulative tax burden on initial wealth transfers, and notwithstanding its size, it is unlikely to be a factor in shaping bequests by the initial donor. Otherwise, the donor would have had to anticipate the short life expectancy of the recipients, and more likely have altered her bequest division. Furthermore, it is not clear how to justify the credit given that the transferor, and not successive heirs of the initial bequest, is defined as the taxpaying unit. Other questions that arise include whether there should be a deduction instead of a credit, or whether the credit should be limited by some family ties, and equally important is the duration, e.g. more or less than 10 years, where the credit would apply.

j. The unified credit

Tax provisions available to all taxpayers are typically not treated as tax expenditures. In the case of the individual income tax, for instance, the personal exemption is not treated as a tax preference item. And because it is available to all taxpayers, it is treated as part of the tax rate structure. The unified credit under the estate tax which effectively provides an exemption of $5 million (indexed as of 2011) is similarly treated as part of the tax base. But the large size of the exemption raises the question of whether the tax relief it provides, that is likely well beyond the needs for tax administration and simplification, should be treated as a tax expenditure. If the size of exempted estate were rolled back to $1 million, as under the 2001 law (in effect in 2006), then the revenue from the estate tax can potentially quadruple.

k. Other tax provisions:

Some tax provisions do not fit the tax expenditure framework even though by their description may qualify as such. Consider the gift tax treatment for health and education related gifts. Payments to health and educational institutions for services provided to beneficiaries are exempt from the gift tax and are not reported on any gift tax return; they are taxable when transferred directly to the beneficiary to pay for such services. As an example, if a parent were to pay a college for her children's tuition, then the entire amount is exempt from the gift tax and is not reported on gift tax returns. In contrast, if the parent were to give her children the funds to pay the tuition, then the transfer becomes subject to the rules of the gift tax. Should these transfers be considered part of intergenerational transfers that should be subject to transfer taxes or should they be treated as part of the normal consumption patterns within the households? Little information is available on such gifts as they are not required to be reported on tax returns. The amount of money at issue, however, is likely to be quite large. For example, some 20.4 million students were enrolled in universities in 2013, at an average cost of $21,000 in tuition, fees, room, and board.²⁴

A similar question arises in the case of gifts to 527 political organizations. Such transfers became exempt from the gift tax in 1974 and are not required to be reported on gift tax returns. How should this exemption be treated? Should these gifts be taxable or are they simply a manifestation of free speech and beyond the reach of the tax code? No information is available its size as such gifts are not required to be reported.

5. Tax expenditures and tax reform

The tax system is often viewed as a legislative compromise featuring a narrow tax base and high tax rates. This is the reverse of what is typically seen as an economically efficient tax, which would generally have a broad base and a low rate. In such a tax system, exceptions from the broad base would only be granted to rectify clearly defined (and suitably large) cases of market failure or to pursue other well defined social and economic goals, such as an equitable distribution of the tax burden. Because they represent deviations from the broad base/low rate mantra, tax expenditures are often targeted as the revenue sources to pay for other proposed tax changes, such as a lower tax rate.

In its most basic form, the estate tax can be characterized as a tax with zero rate on the first $5 million in bequests to heirs, and at a 40 percent rate on transfers above this threshold. For those close to this threshold, the incentive effects can be sizeable. For those much wealthier, the infra marginal zero tax rate most likely has no incentive effects. But the 40 percent tax rate may be deemed too high by some so as to distort their behavior. This leads to the critical question of whether the tax base can be sufficiently expanded by curtailing preferences in exchange for lower tax rates.

Within the context of the estate tax, charitable bequests represent by far the largest tax expenditure. The static revenue effects of the repeal of their deductibility may increase estate tax revenues by about $7 billion, or some 40 percent of the tax liability observed on estate tax returns filed in 2014. This change is large enough to allow for a substantial reduction in estate tax rates. But if taxpayer behavior were to be considered, then much of the revenue gain from repeal can dissipate. As an example, donors may accelerate their contributions and make them during life instead.

The second largest tax expenditure is the deduction for state estate and gift taxes. Repeal of this deduction may raise about $1 billion in revenues, or about 6 percent of the estate tax liability. The revenue raised would only pay for a small relief in tax rates. The remaining tax expenditure provisions are too small to matter. Accordingly, repealing all provisions that may commonly be viewed as tax expenditures given the preferences they embody, will contribute little to lowering estate tax rates.

But if we were to move to the second tier provisions, those with ambiguities in whether they should be classified as tax expenditures, then the potential for base broadening in return for lower rates increases dramatically. Consider the unlimited marital deduction which is very large; $56 billion were claimed on returns filed in 2014. If deduction were to be repealed, as under pre-1948 law, it can easily double the revenue from the estate tax. Alternatively stated, it can easily pave the way for cutting the estate tax rate in half. But as alluded to earlier, the views on the marital deduction have evolved over the years, and its repeal is unlikely to be a viable option.

When compared to the repeal of the marital deduction, a stronger case can be made for repealing the marital deduction for bequests in the form of QTIPs only in exchange for lower tax rates. QTIP trusts do not give the spouse control over the assets, and so may seem more like bequests to the ultimate beneficiaries of the trusts, rather than bequests to the surviving spouse; the primary purpose of QTIPs is to ensure that certain beneficiaries, children from a previous marriage as an example, receive the residual of assets in these trusts. With the revenue raised, the tax rate may be reduced by about one third, depending on the induced behavioral effects in allocating spousal bequests.

With the exception of the deduction for spousal and charitable bequests in the case of the very wealthy, the size of exempted estate represents the primary source for the narrow estate tax base. With an exemption of $5.3 million in taxable transfers, only 16,764 estate tax returns were filed in 2014 out a decedent population of 2.5 million.²⁵ And of these filed returns, only about 5,112 paid estate taxes.²⁶ Lowering the exemption threshold would expand the tax base to pay for a much lower tax rate. While a smaller exemption can raise sizeable revenues, extending the scope of the estate tax to smaller estates also shifts the tax burden to them. Such change also does not address the appropriate size of the exempted estate. Should the latter be determined by simplicity and tax administration considerations which can be achieved at lower exemption level? Or should it be targeted at the wealthiest of individuals which calls for a high exemption level, but which sets in place strong incentives for tax avoidance as individuals jockey to get close to this exemption threshold?

A final consideration is whether tax expenditures under the estate and income taxes can be integrated. As an example, all preferences under the estate tax can be repealed, and be replaced with a deduction for basis of assets in the estate. This has the effect of converting the estate tax to a capital gains tax at death, thereby eliminating the benefit of step up in basis. But again, it is still limited to estates over $5.43 million. The static revenue pickup from the new tax regime is likely to be small.²⁷ But the dynamic effects from relaxing the lock-in effect are perhaps large enough to allow for a reduction in estate tax rates. This would capture one the often proposed role of the estate tax as a backstop to the income tax. But it would not reflect on the role of the estate tax as an instrument in enhancing the progressivity of the tax system and reducing wealth concentration.

6. Conclusion

This paper summarized the tax expenditure treatment of provisions under the estate tax. Provisions considered as tax expenditures in previously published budgets yield little revenue when behavioral effects are considered. But when the definition of tax expenditures is expanded, then the revenue potential increases dramatically and with it the potential for sizeable reductions in tax rates.

Some view the estate tax as a penalty, another layer of taxes. Under this view, preferences under the estate tax provide relief from tax penalty. A similar view is held for the corporation income tax even though tax expenditure estimates for preferences under the corporate income tax are published in the budget. Notwithstanding this asymmetry in its treatment, the estate tax may not necessarily add another layer of taxes to the accumulated assets in an estate. As an example the estate tax would capture some of the gains stepped up at death under the income tax. Another example would reflect gains from taxing previous income tax avoidance or tax evasion. But even if it were another layer of taxation, this could merely represent a higher tax rate applying to a particular activity than that suggested by examining one tax system at a time. It could be an appropriate tax on wealth accumulation per se, desirable as a revenue source and as a way to limit dynastic accumulations of economic power.

Regardless of how the estate tax is viewed, the current system is characterized by high tax rates and a narrow base. It continues to create incentives for tax avoidance and aggressive tax planning. Broadening the base in exchange for lower tax rates can make the tax more efficient. In broadening the base, however, policy makers will have to make difficult choices in selecting which provisions to scale back if not eliminate.

References

Bradford, David F., 1989. "Tax Expenditures and the Problem of Accounting for Government." In Bruce, Neil (ed.), Tax Expenditures and Government Policy , 427-434. John Deutsch Institute for the Study of Economic Policy, Kingston, Ontario.

Budget of the United States Government, Fiscal Year 1994.

Budget of the United States Government, Fiscal Year 2003. . . . https://www.gpo.gov/fdsys/pkg/BUDGET-2003-PER/pdf/BUDGET-2003-PER-5-3.pdf

Budget of the United States Government, Fiscal Year 2017.

Carroll, Robert, David Joulfaian, and James Mackie "Income versus Consumption Tax Baselines for Tax Expenditures," National Tax Journal 64:2, Part 2, June 2011, pp. 491-510.

Dodge, Joseph M. Feminist Perspective on the QTIP Trust and the Unlimited Marital Deduction, 76 N.C. L. Rev. 1729 (1998). Available at: http://scholarship.law.unc.edu/nclr/vol76/iss5/11

Gerzog, Wendy C. The Marital Deduction QTIP Provisions: Illogical and Degrading to Women, 5 UCLA Women's Law Journal 301, 305 n.11 (1995). http://escholarship.org/uc/item/68f6s9fz

Johnson, Craig and David Joulfaian. "A Dynamic Analysis of Estate Tax Repeal," Proceedings of the National Tax Association Annual Conference, Columbus, Ohio, 2007.

Joulfaian, David. The Federal Estate and Gift Tax: Description, Profile of Taxpayers, and Economic Consequences, U.S. Department of the Treasury, OTA Paper 80 December 1998.

Joulfaian, David. "The Behavioral Response of Wealth Accumulation to Estate Taxation: Time Series Evidence," National Tax Journal 59:2, June 2006, 253-268.

Surrey, Stanley S. 1973. Pathways to Tax Reform: The Concept of Tax Expenditures. Cambridge, MA: Harvard University Press.

 

Figure 1. Estate revenues

 

($billions and as percent of total receipts)

 

 

 

² The latest publications are available at https://www.whitehouse.gov/sites/default/files/omb/budget/fy2017/assets/ap_14_expenditures.pdf and at https://www.jct.gov/publications.html?func=startdown&id=4858.

³ Another consideration at the time was whether a tax provision can be replicated with a government spending program.

⁴ This is the current treatment in virtually all published budgets by state governments.

⁵ The identification of tax expenditures under the estate as well as the excise taxes were documented in 1988. See Seymour Fiekowsky xxx.

⁶ More specifically the Budget text reads "Tax expenditure estimates under the unified transfer (i.e., estate and gift) tax have been eliminated from the presentation because there is no generally accepted normal baseline for transfer taxes. . ." See https://www.gpo.gov/fdsys/pkg/BUDGET-2003-PER/pdf/BUDGET-2003-PER-5-3.pdf

⁷ The estimated revenue loss from this exclusion is estimated at over $200 billion per year, as reported in the Budget of the United States Government, Fiscal Year 2017. See https://www.whitehouse.gov/sites/default/files/omb/budget/fy2017/assets/ap_14_expenditures.pdf.

⁸ See Carroll, Joulfaian, Mackie (2011) for a discussion on this and other provisions with ambiguities in tax expenditure classifications.

⁹ See Shultz, pp. 156. Under a pure inheritance tax, the tax is determined by the relationship of the heir and the amount of transfers from the estate to each beneficiary. Under an estate tax, the tax liability is determined by the size of the estate.

¹⁰ See West (1908), Fisher (1916), and Graetz (1983).

¹¹ See US House of Representatives, 1932, pp. 8. Also see Seventy-Second Congress, Senate Article 105, 1932, pp. 40. For a brief period, a gift tax was introduced in 1924 only to be repealed in 1926.

¹² Note that proceeds from life insurance policies owned by others are not part of the gross estate. Only proceeds from policies owned by the insured are included in the estate and form part of the gross estate.

¹³ For an illustration of the importance of this treatment, assume a parent wishing to make a cash transfer of $10M faces a hypothetical gift or estate tax rate of 50 percent. Under the estate tax, the estate pays a tax of $5M, and the heirs receive $5M. Under the gift tax, the tax is $3.33M and the beneficiaries receive $6.66M, for a savings of $1.66M. The tax advantages of making non-cash gifts, however, are in part offset by the income tax treatment of capital gains. Assets are accorded a step-up in basis in the case of bequests and a partial basis carry-over in the case of gifts; the basis of the donee in the case of a gift is the donor's basis increased by the gift tax multiplied by the share of appreciation in the property transferred.

¹⁴ Alternatively, because of the estate tax's close conceptual relationship to an income tax, a set of estate tax expenditure estimates can be motivated as deviations from an income based norm. For example, the estate tax is similar to an income tax (because a tax on the accumulated increase in value of an asset is equivalent to a tax on rate of accumulation) and has often been rationalized as a backstop for the income tax, a measure to correct for leakages from the income tax as well as to increase its progressivity.

¹⁵ By reducing the value of the underlying property, conservation easements also have the effect of reducing the size of the gross estate. Conservation easements made during life also have the effect of reducing the size of the estate.

¹⁶ More specifically, one half the Adjusted Gross Estate, defined as the gross estate less funeral expenses, estate administrative expenses, and debts.

¹⁷ Because spousal bequests may, in part, show up in the second to die spouse's estate, one may argue that in the steady state, the deduction should not be costly to Treasury. Using estate tax returns data (Joulfaian, 1998, table 19), reports a median life expectancy of a surviving spouse of 9 years, and the latter can range well over 40 years; this figures may reflect remarriages. It would take a long time for spousal bequests to become fully subject to estate taxation, it at all. But with some real growth in asset values and number of estate tax returns filed, the nominal estimated revenue loss is unlikely to change over time.

¹⁸ With a marital deduction limited to 50 percent of the estate, very few made spousal bequests in excess of this limit.

¹⁹ For critiques of QTIP, see Dodge (1998) and Gerzog (1995).

²⁰ As an example, if an individual dies with an estate of $3 million, the surviving spouse's estate would eb able to claim an exemption of up to $7 million. Similarly, an individual with an estate of $10 million who chooses to claim an exemption of $3 million only, would enable the surviving spouse to claim an exemption of up to $7 million.

²¹ See year 2014 at https://www.irs.gov/pub/irs-soi/14es01fy.xls.

²² In the case of noncash transfers, the capital gains tax treatment narrows the gap in the tax rates, as basis carry-over applies to gifts and step up in basis in case of bequests.

²³ The credit claimed in 1976 was $84 million, or about 1.7 percent of the estate tax liability of $5 billion. At the time, the estate exemption was $60,000 and the marital deduction limited to one half the estate. See Table 1, Estate Tax Returns, Statistics of Income 1976.

²⁴ See http://nces.ed.gov/programs/digest/d14/ch_3.asp

²⁵ A number of estate tax returns below the filing threshold were also filed, but these are not reflected in the reported figures.

²⁶ In contrast, figures from estates of decedents in 2001 show about 110,000 estate tax returns filed, with half taxable. See https://www.irs.gov/uac/soi-tax-stats-estate-tax-statistics-year-of-death-table-1.

²⁷ Basis information from Form 8939 basis-carryover returns filed for decedents in 2010 was used in imputing the accrued gains.