SSUE BRIEF. EMPLOYEEBENEFITRESEARCHINSTITUTE 1920 N Street,NW/Suite 520/Washington, DC (202) December 1985

Transcription

1 BRI SSUE BRIEF EMPLOYEEBENEFITRESEARCHINSTITUTE 1920 N Street,NW/Suite 520/Washington, DC (202) December 1985 #25 PENSION-RELATED TAX BENEFITS ABSTRACT In attempting to reduce federal deficits, congressional analysts are placing increasing emphasis on tax expenditures -- or the revenue losses that accrue to the federal Treasury as a result of special exceptions to the tax code. Criticisms have been made that private pensions cost the federal government considerable tax revenues and yet benefit primarily those with higher incomes. In fact, the distribution of pension-related tax benefits depends as much on the structure of the income tax system as on the structure of pension coverage. In a progressive tax system such as ours, people will always benefit to different degrees from tax-favored treatment of various sources and uses of income. Nevertheless, most pension-related tax benefits flow to the middle-class, i.e., those with incomes between $20,000 and $50,000 per year. EBRI's analysis shows that three-quarters of the real value of pension tax deferrals is repaid in retirement. Moreover, the cost in federal revenue yields significant benefits to society in terms of higher incomes for the elderly and personal saving among low-income workers who otherwise have no source of savings.!!

2 The distribution of retirement program tax expenditures or tax benefits among income groups is a recurring issue in public policy debates, even though the concept of "tax expenditures" is still relatively new and subject to varying definitions. The Congressional Budget Act of 1974 defines tax expenditures as provisions in the tax code that "allow a special exclusion, exemption, or deduction from gross income or... provide a special credit, a preferential rate of tax, or a deferral of tax liability." In general, tax expenditures are viewed as substitutes for federal spending programs to achieve certain national goals; and as substitutes for direct outlays, they are increasingly viewed by congressional budget analysts as a form of entitlement. When it comes down to actually deciding which tax code provisions are considered "tax expenditures," some differences have emerged recently because of varying approaches adopted by the Joint Committee on Taxation and the Congressional Budget Office, on the one hand, and the Administration, on the other. These varying applications of tax expenditure concepts, however, do not affect the discussion of private pensions, which are uniformly considered by the government to be a tax expenditure. Data on tax expenditures are published annually as part of the President's budget submission to Congress. Although the quality of these revenue loss estimates has been questioned, the numbers are given wide currency.!/ Critics of current law provisions, for example, frequently argue that a disproportionate share of pension-related tax expenditures benefits higher-income persons. The distribution of pension-related tax benefits among employees depends partly on the distribution of pension coverage and income and partly on the tax code structure. This issue brief discusses the effects of the tax structure, the distribution of income, and the distribution of pension coverage among income groups in relation to the distribution of pension-related tax benefits. The first section analyzes from a lifetime perspective the distribution of pension-related tax benefits among income groups drawing from an EBRI study currently in progress.2/ The second section juxtaposes the costs to the federal Treasury against the benefits to society from a private pension system. Pension-Related Tax Benefits: A Lifetime Perspective As Congress and the Administration wrestle with high and rising budget deficits, policymakers are turning their attention to federal tax expenditures, or benefits that are believed to accrue to certain taxpayers as a result of the favorable treatment accorded to certain sources or uses of income. Pension-related benefits are not the only way society chooses to support retirement income objectives. The Social Security program, for example, is designed to provide proportionately more benefits to participants at the lower end of the income scale, while these persons benefit the least from favorable tax provisions since they are in the lowest income tax bracket. This issue brief, however, focuses specifically on pension-related benefits in isolation from other elements of retirement income policy. Z 1/ See, for example, "Retirement Program Tax Expenditures: A Case of Unsubstantiated, Undocumented, Arbitrary Numbers," EBRI Issue Brief #17, April / The Federal Tax Treatment of Retirement Income Programs: Issues t Effects and Alternatives (forthcoming).

3 5 The historic growth in pension coverage, inflation-induced '_racket-creep," and recent changes in the way the estimates are made all helped make pension-related tax benefits the single largest tax-expenditure item in the Administration's fiscal year 1984 budget.3/ Pension-related tax expenditures as computed in the budget, however, are larger than actual long-term federal revenue losses. Taxes deferred by today's pension participants will be reflected in the tax payments due on their pension benefits in retirement; but budget measures of tax expenditures do not reflect this lifetime perspective. Instead, the budget measures pension-related tax expenditures in a cross-sectional context. That is, taxes deferred on pension contributions and earnings accruing to currently-employed participants are offset against taxes paid on benefits by current recipients. This cross-sectional analysis attempts to compare the experience of groups that are not strictly comparable; current workers have higher incomes and both the absolute number and proportion of current workers covered by tax-favored pensions is greater than the number and proportion of current retirees receiving pension benefits. Measuring the revenue costs of current tax policy toward pensions requires a measure of tax benefits that offsets tax deferrals by participants against their own future expected tax payments. To measure tax benefits to persons who are currently working, it is necessary to forecast future pension coverage and benefit levels. Partly under contract to EBRI, ICF, Inc. has developed the Pension Retirement Income Simulation Model (PRISM) to generate pension participation forecasts for a representative sample of labor force participants. Assessing pension-related tax deferrals and payments in a lifetime perspective allows us to take into account two key features of the pension system. First, the pension system can be expected to expand in the future. Projected expansion in coverage will mean that coverage will be much more evenly distributed among income groups. For example, all persons aged 25 to 54 earning $10,000 or less in 1979 accounted for 3S percent of the labor force and S.0 percent of those covered under employer pensions. These persons, however, account for 27.2 percent of those who will be covered, which is much closer to their share of the labor force. Second, projected coverage growth means that persons at all current income levels can expect to reap some benefits from pension-related tax provisions during their working careers. Younger workers earning $10,000 or less in 1979 can expect to receive 14 percent of total net tax benefits accruing to their age group during their work careers (Table 1).4/ By comparison, persons earning SS0,000 or more in 1979 can expect to receive 22 percent of their age group's tax benefits. Persons earning $20,000 to SS0,000 receive 57 percent of their age group's tax benefits. The largest share of net pension-related tax benefits thus accrues to the middle class. 3-TBudget of the U.S. Government_ Fiscal Year 1984: Special AnalTsis G. 4/ The PRISM sample was designed to be representative of the 2S-to-64 year old segment of the 1979 labor force. The calculations presented below are limited to the 25-to-34-year-old group because these person's whole work and retirement careers can be simulated. Results for only one segment of the sample should not therefore be generalized to the labor force as a whole. These percentages would be higher if the calculations only included what can be called the ERISA-relevant labor force, that is, persons at least ZS years old, who work 1000 hours or more per year, and have worked at least one year for their current employer.

4 The distribution of pension-related tax benefit shares among income groups differ from both the distribution of population and the distribution of total tax payments. For example, persons earning $50,000 or less in 1979 constituted 95 percent of workers between the ages of 25 and 34 but they pay about 69 percent of total federal income taxes paid by individuals and receive 78 percent of total pension-related tax benefits (Table 1). Only S percent of the persons in this age group earned over $50,000, but these persons pay 31 percent of total federal income taxes and receive 22 percent of pension-related tax benefits. Thus, the distribution of tax benefits among income groups closely resembles the distribution of tax liability. 4 TABLE 1 Net Lifetime Pension-Related Tax Benefit Shares Among Workers Age 25 to 34 Under Private Pensions As _ of Lifetime _ of Federal Income a/ Labor Force Benefit Shares Payments b/ $i0,000 or less 35_ 14_ 4_ $10,001 to $20, $20,001 to $30, $30,001 to $50, $50,001 and over S Source: EBRI calculations based on PRISM simulation results and U.S. Department of Commerce, Bureau of the Census, Statistical Abstract of the United Statesp , p a/ 1979 income in 1983 dollars. b/ Cross-sectional estimates. In a lifetime framework, the share of total taxes paid by higher-income persons could be higher since incomes grow over the work career pushing persons into higher tax brackets. The Costs and Benefits of Private Pension Systems Pension-related tax policies do cost the federal government some tax revenues in the long run. From a lifetime perspective total tax expenditures for private pensions are much lower than from the cross-sectional method used in the federal budget -- for the reasons discussed above. Preliminary results from the EBRI study cited earlier suggest that about three-quarters of the real, or inflation-adjusted value of tax deferrals is repaid in retirement. In contrast, pension-related tax payments total roughly 17 percent of deferred taxes computed on a cross-sectional basis.s/ S/ In 1982, tax deferrals on pension contributions and earnings computed using the Treasury method totaled approximately $54.4 billion, and taxes paid on pension benefits totaled $9.1 billion.

5 But what does society get in return for this tax expenditure? Without the benefits conferred by the favorable tax treatment of pensions, there would likely be no private pension system. This would result in lower incomes for the elderly and a much more unequal distribution of savings than currently exists. EBRI's published research 6/ has shown that the rates of receipt of pension benefits may be considerably higher in the future than they are today, which will have significant effects on the retirement incomes of the elderly. Table 2 presents the simulation estimates of rates of receipt of pension benefits and average benefit levels for each of four cohorts. TABLE 2 Age in 1979 and Estimated Percentages of Families to Receive Pension Benefits at Age 65, and Average Benefits in 1982 Dollars, by Marital Status All Families Married Couples Single Persons Cohort Percentage Average Percentage Average Percentage Average Age in to Receive Amount of to Receive Amount of to Receive Amount of 1979 Benefit Benefit Benefit Benefit Benefit Benefit $12, $14, $8, , , , , , , , , ,718 Source: Social Security: Perspectives on Preserving the System (Washington, D.C.: _BRI, 1982), p. 90. The benefit recipiency rates are projected to increase fairly rapidly between the oldest and the second oldest cohorts attaining age sixty-five. Beyond the retirement of the second cohort, these rates continue to grow steadily but slowly, especially between the retirement periods of the third and the youngest cohorts. Among workers aged twenty-five to thirty-four, between two-thirds and three-fourths o the family units are projected to be eligible for a pension by the time they reach age sixty-five. Because there is a fairly direct linkage between the growth in wages and individual pension entitlements, the average family pension benefits (in 1982 dollars) are estimated to rise steadily over time. For defined benefit plans that base benefits on years of service and final average salary, the effect of growing wages is direct. In unit-benefit plans, especially prevalent in union-negotiated pensions, increases in benefit levels are generally negotiated in accordance with negotiated wage levels. In defined contribution plans there is also a link between the annual 6/ See Social_Security: Perspectives on Preserving the System, pp

6 6 contribution and wages. In each instance, the assumed growth in future average wage levels portends substantial growth in pension benefits for successive cohorts of retirees, eventually delivering a projected average of $14,541 in annual pension benefits to some 75 percent of married couples in the twenty-five to thirty-four age cohort. It is also clear that private pensions constitute an important form of nondiscretionary savings for a significant portion of the labor force that would otherwise have no savings. Table 3 shows that more than 56 percent of the labor force has no savings, and that their average annual income of $7,672 was nearly half the annual income of the 44 percent of the labor force with savings. Some 27 million workers -- about half of the total labor force without savings -- are covered by private pensions. Other EBRI analysis also suggests that private pension coverage is by far more effective in reaching low-income workers than, for example, individual retirement accounts. TABLE 3 Pension Coverage and Savings Patterns of Civilian Labor Force, 1979 Not Covered Covered by Pension by Pension Average Annual Income (in 1979 (millions) (percent) (millions) (percent) dollars) (percent) No savings a/ $ 7, Some savings b/ $13, Total $11, Source: EBRI tabulations of May 1979 Current Population Survey. _/ Persons are classified as having some savings or no savings according to whether or not they reported any asset income in the survey. b/ Includes persons reporting negative asset income. Conclusion The impact of the tax structure and of the pension system on income distribution should be considered separately. A progressive income tax rate structure means that even tax deductions, exemptions, or deferrals that are distributed equally among income groups will result in a distribution of tax benefits that favors those whose tax brackets are higher.

7 7 This does not mean that pension-related tax provisions result in a regressive redistribution of income, however. The tax code makes important tax benefits to plan participants contingent on the breadth of a plan's coverage in the sponsor's labor force. A tax-qualified plan has to benefit a broad cross-section of a sponsor's employees, and cannot discriminate in favor of the prohibited group (i.e., officers, shareholders and highly paid employees). Therefore, pension plans distribute wealth and savings more broadly than would be the case if no tax incentives for employer plans were in force. Tax expenditures are an inadequate measure of the effects and the effectiveness of social and economic policies. Tax expenditures for employer pensions measure the current year revenue costs of pension-related tax deferrals. They do not measure the lower lifetime revenue cost and they cannot measure the benefits that accrue to pension participants in the form of higher saving rates during their work careers and higher incomes during retirement. Only by comparing social benefits with "tax expenditures" can the federal government determine whether particular incentives are warranted by public policy.

8 The Employee Benefit Research Institute's Education and Research Fund (EBRI) was established to contribute to the expansion of knowledge in the employee benefit field. The Education and Research Fund is tax exempt under section 501(c) (3) of the Internal Revenue Code and is not a private foundation. Individuals, corporations, companies, associations and foundations are eligible to support EBRI's work through tax-deductible contributions. EBRI's policy forums, research studies, issue briefs, pamphlets and other publications aid public and private sector decision makers, managers, the press and the general public in formulating and articulating positions on employee benefit issues. As health and retirement issues receive increasing attention, EBRI strives to make effective and responsible contributions to public policy. iili lilli i iiliilllilli I O lillilll glib I ii I il lilli i gii i il illli lii i ii il RETURN TO: EBR[, 1920 N Street, NW/Suite 520/Washington, DC (202) EBR! NOTES, a bimonthly newsletter, includes information about: EBRI's sponsors automatically receive an annual subscription to I ) newly released employee benefit statistics: 2) brief summaries EBRI Notes_Issue Briefs. Others can subscribe to these publications of EBRI's ongoing research studies and forum discussions; and 3) for $100 annually. reviews of outside research and publications. [] I would like information on becoming an EBRI sponsor. ISSUE BRIEFS, published monthly, offer summary information [] i would like to subscribe to EBRI Notes_Issue Briefs. My check and critical analysis of evolving employee benefit issues and trends, for $100 is enclosed. NAME TITLE TELEPHONE STREET ADDRESS CITY STATE ZIP CODE ORGANIZATION TYPE OF ORGANIZATION AREA OF INTEREST Articles published in Notes and Briefs do not necessari(v represent the views of EBRl's sponsors, and they are not intended to support or oppose any legislation before Congress. Reproduction is,eranted provided proper attribution is given. me iilluuiliii lilliilililllil l illlilllillllillilillill[[illlililliililliolilllllildililil EBRI EMPLOYEE BENEFIT RESEARCH INSTITUTE EDUCATION AND RESEARCH FUND 1920 N Street NW/Suite 520/Washington, DC Address Correction Requested Return Postage Guaranteed!