Pensions

For individuals, future pension benefits provided by employers substitute for current wages and personal saving. A person would be indifferent between pension benefits and personal saving for retirement if each provided the same retirement income at the same cost of forgone current consumption. Tax advantages, however, create a bias in favor of saving through organized pension plans administered by the employee’s firm and away from direct saving.

For a firm, pension plans serve two primary functions: first, pension benefits substitute for wages; second, pensions can provide firms with a source of financing because pension benefits need not require current cash payments. The current U.S. tax code provides additional advantages for using pension plans to finance operations.

Virtually all private plans satisfy federal requirements for favorable tax treatment. The tax advantages are three: (1) pension costs of a firm are, within limits, tax deductible; (2) investment income of a pension fund is tax exempt; and (3) pension benefits are taxed when paid to retirees, not when earned by workers.

To qualify for these tax advantages, pension plan benefits must not discriminate in favor of highly compensated employees, and plan obligations must be satisfied through an organized funding program (see McGill et al. 1996 for further details on these and other institutional features).

Benefits are calculated through formulas established in the pension plan. There are two primary plan types: defined contribution and defined benefit.

Defined contribution (DC) plans specify (define) a firm’s payments (contribution) to the pension fund. The funds are allocated to individual employees, much like a bank account or mutual fund. When the individual reaches retirement age, he or she usually can take the accumulated money as a lump-sum payment or use it to purchase a retirement annuity. Special cases of defined contribution plans include Internal Revenue Code (IRC) sections 401(k), 403(b), and 457 plans.

For example, suppose a DC pension plan specifies that 5 percent of a worker’s salary be contributed each year to a pension fund. Suppose the worker starts at age thirty-five, retires at age sixty-five, and earns $50,000 annually. Then the firm’s annual contribution would be $2,500 (5 percent of $50,000). If the fund earns 7 percent annually, the worker would have $244,277 in the pension fund at retirement, which could purchase a twenty-year annuity paying $22,727 annually.

A defined benefit (DB) plan specifies (defines) the monthly payment (benefit) a retiree receives instead of the annual contribution the employer makes. The benefit is typically specified in terms of years of service and percentage of salary. For example, a plan might specify that a worker will receive 1.5 percent of his or her average monthly salary in the last five years of service, times years of service. If the worker began at thirty-five, retired at sixty-five, and earned an average of $50,000 in the last five years, the annual retirement payment would be $22,500 (45 percent of $50,000). The worker’s firm must pay the promised benefit, either by taking money from the pension fund annually or by purchasing an annuity for the worker from an insurance company. For the pension expense to be tax deductible, the firm must establish an actuarial funding program designed to accumulate enough assets to provide promised benefits.

An actuarial funding program combines data on plan specifications, employee characteristics, and pension fund size with assumptions about future interest, salary, turnover, death, and disability rates. Given these assumptions and data, an actuary estimates the firm’s future pension obligations and an annual payment schedule to satisfy those obligations. Different interest rate and salary assumptions can have a substantial effect on annual contributions, given a plan’s characteristics and an actuarial funding method. A rule of thumb is that an increase of the valuation interest rate by one percentage point will lower pension liabilities by 15 percent, holding all else constant.

Similarly, different actuarial funding methods can substantially affect required and allowable contributions in any given year, even with the same plan characteristics and actuarial assumptions. For example, when Financial Accounting Standards Board Statement no. 87 specified a particular actuarial method for financial disclosure, this raised pension expenses by nine million dollars for Firestone but lowered expenses by forty million dollars for Goodyear when first applied in 1986 (Bleiberg 1988). Federal law, regulations, and accounting standards have reduced the latitude available in the choice of assumptions and in funding methods.

Defined benefit plans are more common among large firms, unionized labor forces, and public-sector employees. Defined contribution plans are smaller, on average, but more numerous, and they frequently supplement an existing defined benefit plan. Defined contribution plans have been growing more rapidly, possibly because government regulation has made defined benefit plans relatively more costly to operate, especially for small firms.

Hybrid plans combine elements of defined benefit and defined contribution plans. The most common type is known as a cash balance plan. Benefits are defined in a manner similar to that of a defined benefit plan. However, a cash balance is established for each employee, which is really just a bookkeeping device to track benefit accruals. An employee who leaves the firm before retirement is permitted to take the cash balance on leaving.

Pension plans can be terminated. With defined contribution plans, the firm merely passes the pension fund management to an insurance company and stops making future contributions. Terminating a defined benefit plan is more complicated and controversial since pension fund assets do not necessarily equal the present value of pension benefits. If assets exceed promised benefits, the excess assets may revert to the employer, subject to certain tax penalties. If a company fails and its pension assets fall short of obligations, deficiencies are partially insured by the Pension Benefit Guaranty Corporation (PBGC), a federal corporation established by the Employee Retirement Income Security Act of 1974 (ERISA). Recent large terminations of pension plans by LTV Steel, TWA, and Polaroid, coupled with the stock market decline, have moved the PBGC from a surplus to a deficit position.

A basic premise of the extensive economic literature on pension policy in the United States is that pension benefits are not free goods: they are provided to workers as substitutes for current wages. Economists have found that the higher a person’s marginal tax rate, the higher his pension is likely to be as a percentage of his wages. This makes sense because pensions are, in part, a means of tax avoidance. Also, the higher an individual’s income, the higher his pension benefit will be as a percentage of current income. So, if a person’s income doubles, the pension portion of his current total compensation might rise from, say, 8 to 12 percent.

How do firms choose how much to fund pension plans and what kinds of assets to invest in? For defined contribution plans, the first half of the question is simple: the employer has to make the promised contribution (e.g., 5 percent of salary or wages) each year and has no other funding decisions to make. Thus, the only ongoing issue for a defined contribution plan is how to invest the assets. Standard portfolio theory suggests that workers would be best off with some well-diversified combination of stocks, bonds, and treasury bills. The relative weights on the portfolios would depend on the worker’s tolerance for risk: the more risk the worker wants to take, the higher the proportion in stock. Because attitudes toward risk differ among individuals and for any one individual as he ages, defined contribution plans frequently allow participants to allocate their contributions among a handful of mutual funds.

For defined benefit plans the answers to both parts of the question are much more complex. In practice, most pension plans are roughly “fully funded” (meaning that assets equal the present value of benefits already earned by workers), and the pension fund is split equally between stocks and bonds. Economists are not sure why this is so but have come up with two possible explanations. The first assumes that the firm owns the pension fund. If so, the firm should choose the funding and portfolio strategy with the highest net present value to it. This leads to two polar-opposite solutions: underfund and buy risky assets, or overfund and buy high-grade bonds.

Why does the first strategy—underfunding and buying risky assets—make sense? Under federal law, a firm can terminate its pension plans without further cost only if the pension fund is greater than accrued benefits or if the firm is bankrupt. In the latter case, the PBGC absorbs the excess liability up to a certain maximum. The insurance premium the firm pays the PBGC bears no relation to the riskiness of its pension fund investments and only limited relation to its funding ratio—the ratio of pension fund assets to pension liabilities. (The basic premium is nineteen dollars per participant with an incremental nine dollars per thousand dollars of unfunded vested benefits.) Therefore, the government’s insurance plan gives the firm an incentive to fund the pension plan only to the minimum required, to substitute pension benefits for wages as much as workers will allow, and then to invest the fund in risky assets such as stock or junk bonds. If the investment works out, the firm gains. If the investment fails, the government and the firm’s workers lose. Note the similarity to federal deposit insurance: both create an incentive to invest in risky assets because the government (actually, the taxpayer) covers losses.

Now for the second strategy—overfunding and buying bonds. This makes sense as a way of reducing taxes for stockholders. Stockholders can reduce their taxes by shifting highly taxed assets out of their taxable personal portfolio and into the tax-exempt pension fund portfolio. This strategy works if stockholders implicitly own the pension fund and can claim those funds later by reducing funding. Furthermore, stockholders can increase their wealth by issuing debt on the corporate account (which is tax deductible at the corporate tax rate) and investing the proceeds in bonds owned by the pension fund. The overfunding strategy works best for ongoing plans because when plans terminate, the reversion of excess pension fund assets to the firm is subject to a stiff excise tax that ranges from 20 percent to 50 percent, depending on the circumstances.

Because one strategy leads to investment in stocks and junk bonds and the other leads to investment in low-risk bonds, firms might be following a mix of the two strategies. This could explain the roughly fifty-fifty split in pension plans’ investments. Alternatively, the fifty-fifty split may simply reflect the “risk-averse” behavior of plan trustees as fiduciaries attempting to satisfy ERISA’s “prudent person” investment standards.

Another explanation for this split is based on the idea that workers, as well as employers, implicitly own the pension fund by sharing in pension fund performance. How? While total pension benefits are allocated by a defined benefit formula, the size of total benefits implicitly depends on fund performance. In such a case, balanced investments serve workers’ aggregate interests because they share in good and bad pension fund performance.

Why would firms allow workers to own the pension fund? One reason, in the case of salaried employees, would be to signal to the employees that it is safe for them to make a long-term commitment to the firm. For union employees, the firm might be concerned that the union will hold out for high wages that, in the long run, will drive the firm out of business. The firm wants its workers to have a strong interest in its long-term survival. If their pension plan is underfunded, workers will have such an interest because they can collect their full benefits only if the firm survives. These incentives are weaker for overfunded defined benefit plans or for defined contribution plans (which must always be fully funded). This can explain why union plans have been almost exclusively of the defined benefit variety and, prior to ERISA, were 30 percent less funded than nonunion plans.