Employer Sponsored Retirement Plans as Source of Income
Analyses of retirement income adequacy and retirement income planning often focus on people age 65 and over because the overwhelming majority of workers have retired by this age. Thus it is possible to look at the sources and level of income available to people over 65 to get some sense of the standards of living that are achievable in retirement.
An elderly unit is a family in which at least one person is 65 years of age or older. 44 percent of the elderly units were receiving some pension income. The receipt of private pensions plans over the period grew considerably more than the receipt of public pensions. To a certain degree this was the result of the participation and vesting standards that had become mandatory with the passage of ERISA in 1974. It was also the result of the maturing of the pension system that had grown rapidly during the 1950s and 1960s but did not provide retirement benefits to many retirees until the plans in the system had been in operation for several years.
Pension coverage at middle and lower income levels and the potential benefits generated by contemporary retirement plans might lead one intuitively to conclude that employer-sponsored retirement plans are generating greater retirement income than the CPS respondents are reporting. Indeed, the National Income and Product Accounts (NIPA) developed by the Bureau of Economic Analysis on the basis of administrative and disclosure data suggest that pensions are paying benefits equivalent to those being paid by Social Security. NIPA estimates indicate that the benefits paid by federal retiree programs in 1990 totaled $53.9 billion; those paid by state and local government retiree programs were $40.6 billion; and those paid by private pension plan and profit-sharing plans were $148.8 billion. In aggregate, employer-sponsored retirement programs paid out $243.3 billion during 1990. According to the NIPA data, the Old-Age, Survivors, and Disability Insurance program paid out $244.1 billion in benefits during 1990. In other words, during 1990 employer-sponsored retirement plans paid out benefits that were almost precisely equal to Social Security benefits. Yet the CPS suggests that retirees over the age of 65 are only getting half as much from these plans as from Social Security. There are some potential explanations for the discrepancy.
Some people participating in employer-sponsored retirement plans do not roll their accumulated retirement assets paid in a lump sum into new retirement plans when they change jobs throughout their careers. When an individual takes a lump-sum benefit from a retirement plan, the plan administrator issues an Internal Revenue Service Form 1099-R to the beneficiary and to the IRS stating the total amount of the distribution and the reason for the distribution. These lump sums could be rolled over into an IRA and drawn down periodically by the recipient or used to purchase an fixed rate annuity.
It is unlikely that the lump sums being reported on the Form 1099-R are also being fully reported as benefit payments. If they were, more than half of all pension benefits paid during 1990 would have been in the form of lump-sum payments. According to U.S. Department of Labor tabulations of Form 5500 disclosure data, slightly more than half of all the benefits paid by private retirement plans in 1990 were paid by defined benefit pension plans. These would have been paid predominantly in the form of annuities. In addition, some of the defined contribution benefit would have also been paid as annuities.
Among public plans, the overwhelming majority of the benefits would have been paid in the form of annuities because public plans have historically been predominantly defined benefit pension plans in which the only benefit form is an annuity. Public employers’ experience with supplemental defined contribution benefit is relatively recent and the benefits being paid out of these plans in 1990 would have been dwarfed by the annuities paid out of their defined benefit plans.
When individuals take lump-sum payments and put them in self-managed IRA accounts with banks or mutual fund companies, periodic distributions from such accounts might not be reported as coming from employer-sponsored retirement program.
The problem in reconciling the discrepancy between the NIPA and CPS pension estimates is that there is no large source of administrative data that captures total income and its components for the whole population. The closest thing to it is the Internal Revenue Service’s Annual Tax Files, which are part of the Statistics of Income (SOI) files that the IRS develops for research purposes. These files, which have been produced since 1960 and which are referred to here as the IRS Tax Files, are an annual sample of tax filer records that have had all personal identification information deleted. One problem with the IRS Tax Files is hardly any information in them describes the characteristics of tax filers other than the financial information that relates to the filing of federal income taxes.
Despite their drawbacks, the IRS Tax Files may be a better indicator of the level of pension payments from employer-sponsored plans than any other potential source of such information. If pensions contribute to retirement income security, people receiving a pension in retirement should, on average, have higher levels of income than people who do not. Of course, higher levels of income trigger the responsibility to file a tax return. This responsibility is enforced by the threat of potential fines and even imprisonment for the failure to file and to do so accurately. In addition to information on annuity income, the IRS Tax Files provide information on a host of other income categories.
Pension, supplemental retirement annuity, and IRA income for elderly tax filers as computed from the IRS 1990 Tax File. It shows the percentage of filing units reporting to have received such income and the average amount reported by those who received the income. The population is split into 10 groups of equal size (i.e., deciles) based on their 1990 income. It shows the combined total income provided from the various sources. We include the IRA distributions in the analysis because IRA accumulations receive comparable tax treatment to employer-sponsored tax-qualified plans and because many workers roll employer-sponsored retirement accumulations into IRAs when they terminate employment with an employer and receive a lump-sum distribution.