Lapas attēli

Appendix III

Current Tax Policies for Funding Retiree
Health Benefits

Voluntary Employee
Benefit Associations

Advance funding of retiree health liabilities may be the only way that many U.S. companies will be able to pay promised benefits in the future. Companies have given various reasons for not building up reserves in the past to pay such costs. Some prefer the pay-as-you-go method because costs are small; others do not want to commit themselves to providing lifetime health benefits to retirees as such a prefunding posture would imply. In addition, even though two vehicles are available for taxdeductible contributions, tax incentives may not be sufficiant to induce companies to advance-fund.

Few companies now fund their retiree health payments in advance, even though they can receive tax deductions for doing so. The Internal Revenue Code provides that employers may either (1) establish a voluntary employee benefits association (VEBA) fund under section 501(c)(9) or (2) set aside funds in a qualified pension plan under section 401(h). These tax-advantaged funding options were intended to encourage companies to set aside money to pay retiree health benefits. It seems neither method has provided substantial encouragement for them to do so.

Until the Congress passed the Deficit Reduction Act (DEFRA) in 1984, some tax practitioners recommended that employers make their contributions for retiree health benefits to VEBA trusts. Before DEFRA, funds contributed to a VEBA trust received tax treatment similar to that given employers' contributions to qualified pension plans, but with fewer restrictions.

The Congress restricted the tax subsidies available through VEBAS for several reasons.' It was concerned that (1) small employers were on the verge of expanding the use of VEBAS as a tax shelter, (2) VEBA rules allowed employers to take deductions before making payments, thus allowing an excessive tax-free accumulation of funds, (3) tax revenue lost to the Treasury as a result of prefunding could be unacceptable, and (4) VEBAS were not subject to nondiscrimination rules, which ensure that a broad cross section of workers in a sponsoring company receive benefits.

Employers still may receive tax deductions, within certain limits, for contributions to VEBAS to fund their retiree health liabilities. The effect

'Joint Committee on Taxation, Staff. General Explanation of the Revenue Provisions of the Deficit Reduction Act of 1984, Dec. 31, 1984, pp. 775-84.

Current Tax Policies for Funding Retiree
Health Benefits

[ocr errors][ocr errors]

of DEFRA rules, however, was to discourage VEBA use by requiring the following:

Contributions to VEBA trusts for retiree health benefits are limited by a requirement that the cost of health benefits for future retirees used in the calculations be the same as the cost of health benefits provided to current retirees. But adjustments for future medical inflation are not permitted by law. Thus, companies could not fund their retiree health liabilities if they limited contributions to the amounts allowed by available tax rules.

Investment earnings on reserves held in VEBAS for postretirement medical benefits may be subject to the tax on unrelated business income. Because such a tax would lower the funds' income, companies would have to make higher contributions to advance fund their retiree health liabilities.

Funding Retiree
Health Benefits
Through Employers'
Pension Plans

The Internal Revenue Code also allows companies to make taxdeductible contributions to their pension plans to fund their retiree health obligations. Under section 401(h), a defined benefit pension plan may provide for the payment of health expenses of retired employees and their dependents through contributions to a separate account maintained under the plan. Medical benefits, together with life insurance benefits, under this section, must be subordinate to the plan's retirement benefits. Specifically, nonpension contributions cannot exceed 25 percent of the aggregate contributions made to the pension plan.

This method of advance funding has been used rarely. The 25-percent limit may not permit contributions to existing pension plans as large as needed to fully fund accrued health liabilities. This is the case for many companies with a high ratio of retirees to workers. Also, because many companies' pension plans are overfunded, allowable pension contributions are very low, even zero, thereby effectively preventing tax-deductible contributions for retiree health.2

2However, the Internal Revenue Service issued a private letter ruling, permitting such contributions in at least one instance. See Bureau of National Affairs, Daily Tax Report, Apr. 3, 1989, pp. G-3 to G


Appendix IV

Data Used to Estimate U.S. Companies' Retiree Health Liabilities

To develop our baseline estimate of private sector employers' total and accrued liabilities for retiree health benefits, we

1. estimated the number of workers expected to receive retiree health benefits and companies' average expected cost of providing early and normal retiree health benefits;

2. made assumptions about worker and retiree mortality, the percentages of workers retiring at ages 60 and 65, and future medical inflation rates; and

3. selected a discount rate to make present-value calculations.

To arrive at values for our baseline estimate of companies' liabilities, we
reviewed and combined information from a number of sources to esti-
mate a model of retiree health liabilities developed by the Department of
Labor. Our model of companies' liabilities estimated the numbers of
active and retired workers covered by company-sponsored retiree health
plans and the projected average costs of these benefits. The stream of
estimated future benefit payments was discounted to obtain the present
value of these future liabilities. The model's output-potential retiree
health liabilities and funding costs-depend on a variety of demo-
graphic, economic, and actuarial assumptions. For example, we needed
to estimate (a) the ages and numbers of covered workers and retirees,
(b) employers' past and projected average costs of providing retiree
health benefits, and (c) future mortality rates and medical inflation.
Because of the inherent uncertainty in these parameters, we performed
sensitivity analyses of retiree health liabilities and funding costs to dif-
ferent parameter values. We then combined these values to create high
and low boundaries around the baseline estimate.

Retirees Covered by
Health Benefit Plans

About 7 million retired workers received company-sponsored retiree health benefits in 1988, according to our estimates. We calculated this number as follows:

1. To arrive at the number of covered retirees in 1983 who survived to 1988, we (a) accepted Labor's estimate of the number of retirees receiving pensions in 1983, (b) estimated how many of these had company

1U.S. Department of Labor, Pension and Welfare Benefits Administration, Employer-Sponsored Retiree Health Insurance, May, 1986, app. D, pp. 73-79.

Data Used to Estimate U.S. Companies'
Retiree Health Liabilities

sponsored health coverage, and (c) applied mortality rates from a commonly used actuarial table (UP842 ).

2. To determine the number of active workers who would retire between 1983 and 1988 and receive health benefits, we estimated (a) the number of active workers who would retire during that period, (b) the number of these workers whose companies provided retiree health benefits, and (c) the number of covered workers who would survive to 1988 to receive these benefits.

As a starting point for approximating the number of retirees receiving health benefits in 1988, we used data on retirees receiving pensions in 1983. According to the Census Bureau's Current Population Survey (CPS),3 about 7 million retired workers received pension benefits from private-sector employers in 1983. About 60 percent of active workers whose employers offered group health insurance would have this coverage continued after retirement, a 1984 survey by the Health Insurance Association of America (HIAA) showed. Because most firms with retiree health plans also sponsor pension plans, we multiplied the total number of pension recipients (7.1 million) by 60 percent and estimated that 4.2 million retirees received employer-sponsored health benefits in 1983.

These estimates have both a downward and upward inherent bias. On one hand, to the extent that some workers without pensions were receiving retiree health benefits, our estimates are understated. On the other hand, retiree health coverage expanded rapidly during the 1960s and 1970s. To the extent that older retirees in 1983 therefore were less likely to be receiving retiree health benefits, our use of HIAA's 60-percent coverage figure may overstate coverage in the already retired population.

In updating the 1983 information on retirees to 1988, we applied commonly used mortality rates and estimated that about 3.5 million of the covered retirees survived. Using CPS data on active workers in 1983, HIAA'S coverage rates, and our retirement rate assumption (described below), we estimated that an additional 3.1 million workers retired with

2William W. Fellers and Paul H. Jackson, The UP-84 Table, The Wyatt Company, May 22, 1975, table 10, p. 18.

3U.S. Department of Commerce, Bureau of the Census, Money Income of Households, Families, and Persons in the United States, 1983, Series P-60, table 50, p. 172.

*Health Insurance Association of America, A Profile of Group Major Medical Expense Insurance in the United States, p. 13.

Data Used to Estimate U.S. Companies'
Retiree Health Liabilities

Figure IV.1: Estimated Number of
Workers and Retirees Covered by
Companies' Retiree Health Plans by
5-Year Age Cohorts (1988)

health benefits after 1983 and survived to 1988. In total, about 6.6 mil-
lion retirees received employer-sponsored retiree health benefits in
1988, according to our estimates. The estimated numbers of retiree bene-
ficiaries in 1988 by age group are shown in fig. IV.1.

[merged small][graphic][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

Retirement Rate and
Age Assumptions

Most workers retire between 60 and 65 at an average retirement age of about 62. We assumed that 60 percent of workers would retire at age 60 (early retirement) and the remaining 40 percent at age 65 (normal retirement). This assumption allowed us to estimate the retiree health liabilities for workers who retired between 1983 and 1988 and those who will retire in future years. Common in actuarial valuations, this type of simplified assumption results in an average retirement age of 62 years.

« iepriekšējāTurpināt »