Statement of Wilbur A. Steger, Ph.D. and Frederick H. Rueter, Ph.D., CONSAD Research Corporation, Pittsburgh, Pennsylvania

May 2002

DYNAMIC SCORING EXPERIENCE: TAX BILL OF 2001

1.0 Introduction

The purpose of this report is to describe an analytical modeling effort performed by CONSAD Research Corporation, to design, implement, and analyze policy analysis models, the results of which were to be utilized by parties involved and interested in Hearings before the Senate Finance Committee’s Subcommittee on Taxation and IRS Oversight on “Preserving and Protecting Family Business Legacies”. This report draws heavily on Dr. Wilbur A. Steger’s “Testimony before the Senate Finance Committee Subcommittee on Taxation and IRS Oversight on ‘Preserving and Protecting Family Business Legacies’” (March 13, 2001), as well as an earlier Final Report, “The Federal Estate Tax: An Analysis of Three Prominent Issues” (February 7, 2001), prepared by CONSAD for the Food Marketing Institute. Dr. Steger’s colleague and Vice President of CONSAD, Dr. Frederick H. Rueter, participated on an equal basis in all these efforts; Mr. Scott Kinross, a CONSAD analyst, participated throughout, also.

Many analytic/modeling tools were used in addressing issues. In Section 2.0, below, we review the results of macroeconomic (more precisely, regional econometric) modeling estimating outcomes -- changes in regional and national product and employment, and in Treasury revenue -- resulting from reductions and, then, elimination of the Federal estate tax. Relative to such outcomes, other analytic investigations have examined changes in liquidity-related vulnerability of family-owned businesses due to changes in the estate tax; and the prevalence of family-owned businesses in taxable estates. Section 3.0 analyzes other effects of reducing and/or estimating the estate tax, particularly that relating to the growth of unrealized capital gains in the hands of heirs, and the related revenue effects.

The above analyses, and those described below, would be described, by some, as “dynamic scoring”, i.e., reflecting prospective behavioral (micro) and aggregative (macro) effects (see statement on “Dynamic Scoring”, Kevin Hassett, American Enterprise Institute, May 2, 2002 for current thinking about and attitudes toward the inclusion of such effects on the policy-making process). Section 4.0 discusses the implications of the experience described in this paper for future dynamic scoring efforts.

2.0 Summary of Macroeconomic and Related Effects of Reducing or Eliminating the Estate Tax

There is a lengthy and complex history to deliberations regarding the estate tax and capital gains. The March, 2001, Senate Subcommittee on Taxation and IRS Oversight was unique, however, since they focused on the economic effects on family business and workers of reducing or eliminating the estate tax, and of the direct and side effects of freeing locked-in capital markets.

First, CONSAD addressed the issue of the magnitude of the problem and explored, explicitly, who is impacted. We were able to find a more accurate measure for defining the financial attributes of an estate that includes a family-owned business. The summary data that the Internal Revenue Service (IRS) has compiled from estate tax returns indicate that the assets of family-owned businesses are sizable portions of the estates reported on a substantial percentage of taxable estate tax returns. Rather than being less than 500 in a typical year, we estimated the total number of taxable estates that consist largely of family-owned businesses likely exceeds 10,000 annually.

Based on macroeconomic modeling, we found that important economic benefits would result from the reduction or elimination of the estate tax and, in the context of repeal, changing the basis for taxing capital gains. These effects include the following:

• Aggregate economic effects are positive. Currently, many small business owners, and estates with non-liquid assets, must break up their business or holdings in order to raise money to pay their estate tax debts. All sides of the debate agree that this has a considerable disruptive effect on many family businesses, including farmers. Proposals to reduce or eliminate the estate tax would make it much easier for these businesses to continue to operate without undue disruption. The research CONSAD has conducted estimates the macroeconomic consequence of the elimination or substantial reduction of the estate tax: i.e., the extent to which these would beneficially affect employment, national income, and economic output. While we did not consider (in that report) the revenue and economic effects of the carryover of basis, as called for in many legislative proposals, we continue to believe that the investment and liquidity-enhancing effects of the elimination or reduction of the estate tax will increase the survivability of family business and their positive effects on local and regional economies. Our research also confirms the benefits of speeding these effects, e.g., through immediate reduction or elimination, particularly if and as economic conditions worsen. Here, CONSAD used an economic forecasting model developed by Regional Economic Models, Inc. to estimate the effects of the resulting changes in those peoples’ consumption and investment spending on the aggregate economy: this is one of the few macroeconomic models that take account of differences in regional productivity, energy sources and uses, and industrial structure. The estimates indicate that eliminating the tax would result in an initial surge in gross domestic product (GDP), the aggregate value added for all firms in all industries throughout the economy. Then, by the fourth year after the repeal, the benefits from reallocating resources toward investment, tempered by monetary policy aimed at averting inflationary pressure, would stabilize and yield steady growth in value added.
• Revenue losses will be lower than were initially anticipated. Experts differ on the estimates of the precise revenue consequences of both eliminating the estate tax and changing the tax treatment of capital gains. Our ongoing research appears to indicate that the revenue gain from the correlate change to the carryover basis will yield annual revenue gains beginning at $5 billion and gradually rising to more than $15 billion yearly. The change in basis at death will lead to more revenue gains than are currently contemplated (see below, Section 3.0). Similarly, while we did not estimate the revenue consequences for the Treasury of the macroeconomic effects described immediately above, they would be positive and substantial.
• Preserving family businesses. Currently, families and estate executors face a complicated set of overlapping tax rules that include the estate tax, capital gains tax, and the gift tax. Many Americans devote considerable time and resources on estate planning to arrange their personal and business affairs in an attempt to minimize their total taxes at death. Unfortunately, without such planning, some estates face an unnecessarily high tax burden that hurts families and small businesses. In the ideal economic model, the simplification of the tax code that would flow from the elimination of the estate tax would result in a clearer picture of expected tax burdens at death, and free up resources now spent on navigating the maze of the tax code.

3.0 Elimination of the Estate Tax and Unlocking Unrealized Capital Gains

Since the middle of the last century, this subject has enjoyed an active history. Not surprisingly, during the early years of the Clinton Administration, the President’s economic think-tank called for an end to the (income) tax exemption for unrealized capital gains held when a person dies. This proposal cited an ultimate revenue yield of $5 billion per year as well as enhanced equity as justifications (Shapiro, 1992). This marked the approximately fiftieth anniversary of the pathbreaking article on this subject -- with a similar objective to President Clinton’s -- by the celebrated income tax specialist and reformer, Stanley S. Surrey (Surrey, 1941).

3.1 Background and History

Professor Surrey was destined to bring this important notion, and an affirmative assessment of its constitutional validity, to the attention of Presidents Kennedy and Johnson while serving as their Assistant Secretary of the Treasury for Tax Policy during the 1960s. Under President Johnson, a Treasury Department study recommended taxing gains as income on a decedent’s final tax return. Then House Ways and Means Chairman Wilbur Mills, working with Surrey and the principal author of this report (Steger, 1957, 1961) during this period, held committee hearings on this and closely related income and estate tax subjects (Steger, 1959; Heller, 1955). Also during this period, leading public finance economists of the day (F.M. Bator, R. Blough, J.K. Butters, R.F. Gemmill, J.K. Lintner, L.H. Seltzer, H.M. Somers, L.E. Thompson, and others) provided excellent insights into prospective economic and equity effects of taxing capital gains as though realized at death and/or disallowing the stepped-up basis.

Most recently (2001), tax expenditure estimates by the Treasury Department’s Office of Tax Analysis, based on a retrospective analysis, were indeed quite high. Conversely, the CBO estimate of revenue gain appears to have been lower, as explained below. Such analyses were performed using different, but reconcilable, assumptions. The estimate in Mandate for Change (Shapiro, 1992), for example, assumed the continuation of the current exemption for capital gains on assets willed to a spouse or donated to a charity, as well as gains in a small business or a farm, and provides additional exemptions (up to $125,000) for gains from the sale of a residence.

• Aside from its uncertain but clearly substantial revenue consequences, a variety of economic and equity reasons were advanced for reform of the tax treatment of assets at death (Steger, 1957, 1959, 1961; Surrey, 1941; CBO, 1992; Butters, 1953):
• Reducing the disparity between those who save through an appreciating asset and those whose income is entirely taxable (i.e., the Haig-Simons-Vickery economic concept of taxable income)
• Reducing the incentive for investors to hold assets until death to avoid capital gains taxes (the “lock-in” effect), thus diminishing (or preventing) the blocking of otherwise economically efficient investment decisions
• Assessing a tax on income at death involves adverse consequence for economic incentives and efficiency during lifetime, both for the decedent and their heirs.

The Bush Administration (2001) appears to have supported the tax treatment at death for unrealized gains described in the Kyl-Breaux Estate Act of Tax Elimination Act of 2001. (There were similar arrangements in other bills.) The proposal allowed every individual to continue to step-up the tax basis of assets in his or her estate to the fair market value at the date of death, subject to an overall limitation on untaxed capital gains of $2.8 million per individual (or $5.6 million per married couple). The per-person exemption was to be indexed for inflation. The limited step-up in basis would protect small estates from any new capital-gains tax liability and reporting requirements. Such liability and reporting requirements was to apply only to estates with unrealized gains in excess of $2.8 million (or $5.6 million in the case of a married couple). Other bills took different approaches, also using the decedent’s tax basis in one way or another.

Questions have been raised about these unrealized capital gains -- considered together with the degree to which the estate tax is curtailed or eliminated:

1. What is the current magnitude of these unrealized capital gains and their distribution among asset classes?
2. What would be the revenue effects of various treatments (e.g., degree and method of carryover, phasing, grandfathering, etc.)? How would each variation affect the current estimates of the decrease in tax revenues that would result from repealing the estate tax?
3. What effect on the economy would result from alternative treatment, in terms of employment and output in specific industrial sectors by state and region. How might these economic effects alter estimates of impacts on tax revenues?
4. What would be the effect on different demographic groups (e.g., income, age, family type) from each treatment variation?

CONSAD conducted a preliminary analysis using a regional econometric model and associated analytic software and interpretation of tax research results to estimate the revenue, economic, and demographic consequences of a set of “what if” realization patterns of these capital gains. This research is ongoing.

3.2 Estimating Consequences

Consider, for illustrative purposes only, that $15 trillion for capital gains (in current dollars) are created and accrued over a 25 to 30 year “generation” of taxpaying earners. This rough estimate draws upon research findings made by Steger (1957, 1959) and, thirty years later, by Gravelle and Lindsey (1988) that: (a) on average, only 3.1 percent of the stock of accrued gains are realized in any given year, over a 25-year period; and (b) that realized capital gains in each year average only 24 percent of the total capital gains accruing to the household sector in that year, specifically:

• Approximately fifteen trillion dollars (more or less, as of late 2001) of unrealized capital gains will become more free and fluid to serve the interests of American businesses and their workers. We have come to know, through research and judgment [Steger, 1957; Gravelle and Lindsey, 1988; Burman, et al., 1997; Auten and Joulfaian, 2001 (forthcoming)] that there is an immense pool of accrued but yet unrealized capital gains. As estimated above, these currently amount to as much as $15 trillion, and are growing. Proposals to transition from the stepped-up tax basis for capital gains to the carryover basis will result in increased revenues, partially offsetting the loss in estate tax revenues. The stepped-up basis will, by and large, diminish in importance with the elimination of the estate tax.

Many economists believe that the majority of capital gains, under the current system (with a stepped-up basis), are never realized, but, instead, are passed on to heirs with a step-up in basis or given away in a tax-free transaction. It would seem that, were unrealized gains taxed at current capital gains tax rates, either at death or to heirs over their lifetimes, a yearly equivalent of many billions of dollars in additional taxable gains might result. How would these complement the current revenue of approximately $90 billion for realized capital gains?

During the spring of 2001, CONSAD conducted a new study of the economic and revenue consequences of then-current alternative proposals, using analytic software and matching databases addressing the following issues related to the reduction or elimination of the estate tax and its capital gain correlates:

• Federal government revenue changes (from both the income and the estate tax),
• Changing patterns of capital gains realization,
• Changing acquisition and disposition patterns of capital assets.

The possible economic and fiscal impacts range from relatively minor to significant. The purpose of this research was to narrow the range of prospective outcomes, such that they would provide information helpful in distinguishing among alternative policy options.

In addition, through the study of the positive aggregate economic effects of the elimination of the estate tax (Section 2.0, above, employing the most widely utilized regional econometric model), we discovered that reducing or repealing the estate tax would free up substantial resources for alternative purposes. The heirs of people who die would inherit additional funds that otherwise would have been collected as taxes. Also, the resources that people now expend (i.e., planning costs) to mitigate the consequences of the estate tax would be released for other uses. We also discovered that the aggregate gains in value added in the majority of U.S. industry substantially exceeded the decreases that would occur in the few industries that would experience decreases in demand for their services due directly or indirectly to the reduction or repeal of the estate tax. This research also established the additional benefit, particularly in tight economic times, of making the reduction or elimination take place as quickly as possible, including immediately. Our ongoing research has altered these estimates only slightly while, at the same time, realizing increased revenues to the Treasury.

3.3 Interim Results: Incorporation of Behavioral Consequences

The combination of the estate tax and the stepped-up basis at death determine the total tax paid by estates and their heirs. So, alternatively, would a system with no tax (at death) on estates and a carryover (primarily) of basis. However, just as it took time for the current system to settle into a relatively predictable pattern, it will take years for any new system to settle into its routine.

This section summarizes the results that have been produced by the model that CONSAD has developed for estimating the changes in government revenues that would occur under a number of proposals that change provisions of the estate tax, the gift tax, and the generation-skipping transfer tax. [This draws heavily on research performed by Douglas Holtz-Eakin (see References, below).]

The elimination or phasing-out of these transfer taxes would naturally lead to a decrease in government revenues. Our analysis indicates that there are several behavioral responses to a change in tax structure that would offset at least a portion of the revenues forgone.

In particular, there would be a positive effect on the rate of capital gains realizations by older people who currently experience a "lock-in effect" as they age and plan for their demise. If the existing step-up in the basis for measuring taxable capital gains is replaced with a carry-over in basis, these people will likely realize gains at rates similar to those observed for somewhat younger people. These extra realizations by prospective decedents would be taxed at the capital gains tax rate, and would yield additional government revenues.

The second behavioral change relates to realizations of accrued capital gains bequeathed to the heirs of large estates. These realizations would also be taxed at the capital gains tax rate throughout the lives of the heirs, adding revenues to the government throughout the period after initiation of the carry-over in basis.

CONSAD has developed a computer model that estimates the effects of these behavioral changes and revisions of tax structure on government revenues. [A description of the general procedures (translated to a computer model) for estimating incremental capital gains realization and associated tax revenues under specified realization and carryover assumptions is available on request.] The methods and evidence used to model each of these effects are discussed briefly below. (The associated databases and explanation of the calculations are available on request.) Then, the estimates developed for several specified tax reform scenarios are summarized briefly.

Realizations by prospective decedents -- The empirical research literature suggests that under the current estate tax system the wealthy begin to experience a "lock-in effect" after age seventy-five. With the elimination or phasing-out of the tax at death, people will lose the incentive to retain the accrued capital gains in investments in anticipation of death. Instead, they will be free to exercise the opportunity to realize and re-invest accrued capital gains at rates similar to those that they were applying earlier in life. Such behavior will generate additional capital gains tax revenues for the government.

Realizations by heirs -- People who inherit wealth can be expected to use that wealth in ways that are similar to the use observed for others in similar financial circumstances. In addition, the economics literature suggests that inheritances provide windfalls that fundamentally change some people's economic behavior, such as their participation in the labor force and their willingness to become entrepreneurs. The elimination or phasing-out of the estate tax also naturally increases the amounts bequeathed to heirs. Initiation of a carry-over in basis for measuring taxable capital gains will cause these actions and events to yield additional revenues for the government.

Phase-out of tax provisions -- Phasing out of the estate, gift, and generation-skipping transfer taxes over several years can be accomplished in many ways. All of them will involve, at least, reductions in the tax rates applied to the values of assets contained in the estates, and increases in the amount of assets that can be bequeathed without incurring tax liability (the Unified Credit). Changes in those provisions of the transfer tax structure will produce changes in the average tax rates that effectively are imposed on estates of different sizes. In comparison to the current situation, estate tax revenues will decline; however, in comparison to immediate repeal of the taxes, additional government revenues will be collected.

Results derived for specified scenarios -- In the initial application of the computer model, CONSAD has analyzed three scenarios for potential changes in transfer tax policy. Two scenarios consider the immediate repeal of the estate tax. In Scenario One, repeal is combined with a phased-in limitation of the step-up in basis (i.e., a phased-in provision of carry-over in basis) over five years. In Scenario Two, repeal is accompanied by immediate establishment of a limited step-up in basis at the high level proposed in the Kyl Proposal (S. 275). In the third scenario, the transfer taxes are phased out over a four-year period, after which the taxes are repealed and a limited step-up in basis is established at the maximum level specified in Scenario One. The government revenues estimated for the three effects throughout the ten-year period from 2002 through 2011 for those scenarios are summarized in the following table:

The amounts reported in the table are offsets against the transfer tax revenues that would be foregone if the transfer taxes were immediately repealed. In response to suggestions by model users, we are still refining the estimators that we are using for some calculations. Thus, some of the values in the table may change slightly when the final calibration of the model is completed.

4.0 Concluding Remarks

Those who have looked at the policy/budgetary process and the history and future of dynamic effects when scoring spending and tax bills (e.g., “Dynamic Scoring”, op. cit.) have not said “put these aside”. Rather, they have focused on better ways for Congress (JCT), the Federal Reserve, CBO, et al., to use their revenue estimating staffs and what “information revelation” practices would accompany such changes.

The technical approaches discussed in this report (Sections 2.0 and 3.0) found their ways into the above organizations (as well as the Treasury, CEA, and NEC) -- during the estate tax debates and discussions of the 2001 Tax Bill primarily -- and, indeed, appeared to have the intended effect: to raise questions about “official” estimates produced and used during the final discussion process. Where the behavioral effects -- of prospective decedents and/or unrealized gains-heavy heirs -- had been reported in reputable research journals, where the modeling software utilizing these finding was transparent and available for validation and sensitivity-testing purposes, and where these models could be turned over (for use) by parties involved in the many (often closed-door) discussions -- we believe that such dynamic effects had telling consequences, in opening up minds and discussion.

We believe there is a role -- semi-academic, semi-policy analytic -- for such “entrepreneurial”, informed intrusions into the process. We predict these “intrusions” will become more welcome with time and with increasingly successful instances of behavioral effects/scoring. There are two arenas where we believe these instances will occur more often:

1. The two way and interactive causality of the personal income tax systems with respect to the size (e.g., wage bill) of the business yielding that income. How income from a business (corporation or subchapter C or S) is taxed leads to more or less savings/income to invest in a business as well as its liquidity position (research by Henry Rosen, James Poterba, and Douglas Holtz-Eakin) and, therefore, the ensuing and affected elasticity of the wage bill (i.e., "jobs").
2. The personal tax system's incentive effects (vis-a-vis capital gains and losses) on entrepreneurship opportunities and self-employment investment (Poterba).

Figures 1 and 2 represent an attempt to describe a general framework for subjects where behavioral research could be put to unified policy analytic studies; and how CONSAD’s work, to date, does its technical behavioral scoring work.

5.0 References

A complete list of references drawn upon for writing this paper is available upon request of the longer version of the paper.