Testimony Submitted to the Working Group on Defined Benefit and Defined Contribution Plans

Although pension reform measures in last year’s Small Business Jobs Protection Act made great strides toward simplification, there is much work left to be done. To get at real reform, policymakers should forget trying to fine-tune a number of the remaining burdensome requirements by clarifying their language and providing "safe harbors" from certain provisions. It should discard them altogether.


When Congress began chipping away at tax "loopholes" in the 1980s, tax preferences for pension plans were targeted because well-paid workers were much more likely to be covered by a pension than low-paid workers. The goal of 1980s-era pension law changes was to increase federal revenue while spreading pension coverage more equally across the workforce. The result was layer upon layer of new nondiscrimination rules which tied the contributions and benefits awarded to highly-paid employees to the level of those received by rank-and-file workers in the same firm.

Today, employers must perform complicated mathematical tests to ensure that the difference between contributions to, and benefits from, pension plans for both groups falls within an acceptable range. The regulations are so arcane that, in many cases, even pension professionals and tax attorneys have difficulty determining whether or not a particular pension contract passes muster. Furthermore, these mandates restrict the ability of employers to design plans which best suit their needs and their workers’ demands. They cannot offer significantly higher pension compensation to more valuable workers.

The consequences are well known to members of the Advisory Council. After World War II, pensions had become an increasingly important part of worker compensation, and pension coverage rose steadily for more than three decades. By 1979, 50 percent of all civilian non-agricultural workers participated in an employer-sponsored pension plan.1 However, the increased over-regulation of private pensions since that time has brought their growth to a screeching halt. Within a brief period of time after enactment of the "top heavy" rules in the Tax Equity and Fiscal Responsibility Act of 1982, the most egregiously lopsided pension plans were terminated. But the laws kept coming. The Tax Reform Act of 1986 and the Omnibus Budget

Reconciliation Act of 1987 dragged many more plans into the domain of noncompliance. Despite the creation of thousands of new plans during the past decade, the worker participation rate had fallen to 47 percent in 1993,2 and it appears to be stuck at about half the workforce. Worse, still, all of this had no noticeable affect on pension participation among low-income workers.

Changing labor force demographics, evolving worker preferences, and escalating global competition account for part of this stagnation. Although these forces are largely beyond our control, another major factor—government over-regulation—is thoroughly within our power to change. Despite great promise, though, the last round of pension reforms failed to reduce the real regulatory burdens significantly.

The Small Business Jobs Protection Act

The Small Business Jobs Protection Act (SBJPA) was heralded as the pension industry’s savior. It simplified the definition of highly compensated employees (HCE), repealed the family aggregation requirement, exempted defined contribution plans from the section 401(a)(26) additional participation requirements, simplified several distribution rules, provided targeted simplification for small employers, adjusted the definition of compensation for calculating certain limits, repealed the section 415(e) combined annual limits for contributions to defined contribution plans and benefits earned under defined benefit plans, and made a host of "miscellaneous" simplifications. All of these changes were needed and are welcome, but they generally were not considered to be the focal point of the legislation.

Most industry and public policy analysts agreed that nondiscrimination reform was the key to any real revival of private pensions in the United States. Much has been said about the expected positive effect of the discrimination provisions in the SBJPA. In truth, however, those provisions were modest at best. The Act introduced Savings Incentive Match Plans for Employees of Small Employers (SIMPLE), a new "model" defined contribution plan for businesses with 100 or fewer employees, and it created a safe harbor for all 401(k) plans. These provisions eliminate the need for much nondiscrimination testing due to designed-in participation and contribution requirements.

The 401(k) safe harbor and SIMPLE plan provisions are very much alike. In general, under the safe harbor, 401(k) plans are deemed to have satisfied the nondiscrimination requirements if the employer matches 100 percent of each non-highly compensated employee’s (NCE) elective contributions up to three percent of compensation and 50 percent of elective contributions between three percent and five percent of compensation, or the employer makes a nonelective contribution of at least three percent of compensation for all NCEs, whether or not they make elective contributions. The safe harbor allows HCEs to make the maximum allowable deferrals, even if the average of NCE deferrals otherwise would prevent them from doing so.

SIMPLE plans can take the form of a 401(k) or an IRA established for each employee, but the nondiscrimination provisions are essentially the same for each. The employer must match 100 percent of NCE contributions up to three percent of compensation or make nonelective contributions of two percent of compensation. The maximum allowable SIMPLE deferral in any year is only $6,000, however, whereas the limit is $9,500 in standard 401(k) plans.

Unfortunately, the SIMPLE plans and the 401(k) safe harbor only provide for alternative methods for meeting the nondiscrimination rules; they do not eliminate them. The provisions essentially increase the nondiscriminatory coverage requirement to 100 percent to compensate for the elimination of admittedly complicated and often costly testing. Employers must accept the new burden of making expensive contributions or potentially expensive matching contributions on behalf of all NCEs in exchange for being exempt from some of the test requirements.

Moreover, workers in the new SIMPLE plans can defer only as much as $6,000 each year, and the employer can match no more than three percent of each worker’s compensation. In the words of Forbes magazine reporter Janet Novack, "deciding whether you are better or worse off with a new option is not simple at all."3

Small Businesses and Pensions

SIMPLE plans are aimed at increasing the relatively low rate of pension coverage among employees of small businesses. In 1993, 83 percent of firms with 100 or more workers sponsored a pension plan, while only 19 percent of firms with 25 or fewer employees sponsored a plan.4 Unfortunately, there are a number of reasons why small firm sponsorship is so low, and SIMPLE doesn’t address any of them.

SIMPLE can’t remedy the low revenue or disadvantageous economies of scale suffered by small businesses. It can’t change workers’ preferences for cash wages or other benefits. Most importantly, it does less to ease the nondiscrimination and paperwork barriers than Simplified Employee Pensions (SEPs) or Salary Reduction SEPs (SaR-SEPs) do already. The SIMPLE legislation is not the holy grail of pension law for which many critics had been searching. The ultimate effect is to add more, and only marginally useful, twists and turns to a tax code that ought to be less, not more, complicated.

Why can’t an employer simply contribute three percent of salary for each NCE and be done with it? Some employers will certainly be able to do so, because employers do not pay for pensions, workers do—in the form of reduced cash compensation. Many employers, however, will feel pressure to pay NCEs higher cash wages. In a competitive labor market, firms are bound on one side by the necessity to pay employees no more than their marginal product, and on the other by an NCE’s desire for a minimum level of cash compensation.

Low-wage employees often need the higher cash compensation just to make ends meet. This fact caused pension expert Kathleen Utgoff to note, in a September 1991 National Tax Journal article, that even those NCEs who do receive pensions due to their employers' attempt to comply with the nondiscrimination laws are not necessarily better off. The nondiscrimination rules may cause some firms to offer pensions to low-income workers, but "[t]hese additional pensions do not mean, however, that even these low-income workers are more financially secure as a result."5

The real problem is that revenue in small businesses is limited. Employees of small businesses enjoy few workplace benefits of any kind. These workers tend to be younger than average, more transient than average, and, when given a choice, tend to prefer such other benefits as health care insurance to pensions.6 Some employees, especially young workers, would rather choose to postpone saving for a few years in expectation that their compensation will increase. So long as young workers fully understand the consequences of their waiting, this decision is not irrational.

Sometimes, higher paid employees see an advantage in subsidizing NCE saving in order to take advantage of the tax preference for qualified plans. This will occur only as long as the subsidy does not exceed their expected gain from pension tax advantages. However, in most small firms that do not now offer pensions, there are few (perhaps only two or three) highly compensated employees who would have to subsidize the saving of ten, twenty, or more non-highly compensated workers at a rate of several hundred or several thousand dollars each, as well as the administrative costs of simply providing the plan. In these cases, the HCEs may decide that the benefit is not worth its cost.

Unfortunately, too much attention is paid to very small firms, because most large firms already offer some or all of their workers a pension. Any substantive growth in pension participation is likely to come from firms of the "medium" size—those that are small but growing. These successful small and mid-sized firms are at the margins of pension coverage. Given enough leeway, they will gladly establish plans. For these firms, the newly created 401(k) safe harbor is promising, since it is not limited to firms of a given size. Unfortunately, the expected initial burst of 401(k) creation is unlikely to be sustained if pension law continues to tie the benefits of highly-compensated to those of low-paid workers.

Congress traditionally has blamed the business world for its disparate treatment of low-paid and highly-paid workers. The real question is, should government policy act as though all NCEs desire exactly the same treatment and tie the hands of middle- and upper-income workers? The goal should be to realize that all workers are different and provide an array of options for retirement saving.

Defined Contribution or Defined Benefit?

While the Small Business Jobs Protection Act attempted to reduce the barriers to defined contribution pension saving, it largely neglected defined benefit plans, which face much harsher regulatory burdens. Defined contribution plans have many commendable elements. Vesting is often completed in a much shorter period of time than with defined benefit plans, and vesting of elective contributions is immediate. Workers covered by defined contribution plans commonly earn greater benefits earlier in their working careers. Perhaps most importantly, defined contribution plans are essentially portable. If the worker’s new employer offers a similar defined contribution plan, the accumulated funds from the previous employer often can be added to the new account. Where this option is not available, accumulated funds can be deposited in an Individual Retirement Account to continue growing tax-free until retirement.

That said, the advantages of defined benefit plans should not be overlooked. Where participation in defined contribution plans is voluntary, many young workers prefer income for current consumption over retirement saving. Workers who expect that they will be able to make up for their deferral later in life may fail to understand the real benefit of compounding investment returns. Even many of those who do participate are unaware of the amount of money they must accumulate to receive the benefit level they wish to achieve in retirement—even when this information is provided to them.7 Moreover, this planned "backloading" becomes more and more difficult as workers age due to problems both in plan design and federal government limitations on annual contribution amounts. Fearing financial risk, many employees (especially women) are too conservative with their self-directed investing.8 But in defined benefit plans, employers bear the investment risk. Finally, many job-changers raid their defined contribution funds immediately upon mid-career employment separations.9 Although larger distribution amounts are more likely to be re-invested than smaller ones, even the diversion of small amounts early in one’s working life can significantly reduce total lifetime accumulation.

The decision to sponsor a defined benefit or defined contribution plan—and the specific details of plan design—is best made by the managers and employees of each particular firm. It is important that government be neutral regarding plan type, so that economic decisions are not distorted by political pressures. This year, two groups of mostly Republican senators tried to remedy the discrepancy in current treatment by introducing legislation (S. 883 and S. 889) that would authorize a simplified defined benefit plan type for small businesses.*

SAFE Annuities and Trusts, which are available to businesses with 100 or fewer employees, would allow an employer to purchase individual retirement annuities for each employee or establish a pension trust with a separate account for each employee who receives at least $5,000 in annual compensation. The employer makes annual contributions to the annuity or trust in the amount of three percent of compensation for every employee, but may reduce that percentage, if business income requires it, in two years within any five-year period.

In return, SAFE plans are treated as meeting the requirements of a raft of non-discrimination laws including the section 401(a)(4) contributions and benefits rules, the section 401(a)(26) minimum participation rules, the section 410 minimum participation standards, and the section 416 top-heavy rules. SAFE Trusts are further excluded from coverage under ERISA’s Title IV plan termination insurance and are thus exempt from paying Pension Benefit Guaranty Corporation premiums.

Like the 401(k) safe harbor, this provision is promising, inasmuch as it significantly reduces the paperwork burdens on defined benefit-like plans. It also allows for "catch-up" contributions for prior years. It offers a reasonable, low administrative-cost alternative to defined contribution plans. Unfortunately, it is only available to businesses with 100 or fewer employees. This may prove to limit the popularity of the SAFE plans for the same reasons that SIMPLE defined contribution plans will likely be limited—most small businesses cannot afford to provide pension coverage for all of their lower paid employees. Furthermore, SAFE plan sponsors cannot sponsor a second plan, and the SAFE plans are still subject to the section 401(a)(17) considered compensation limit ($160,000 in 1997). Because the annual accrued benefit can be no more than three percent of compensation (up to the considered compensation limit), and because employee contributions are not permitted, these restrictions limit annual contributions for each employee to a maximum of $4,800—lower than even SIMPLE plans.

Perhaps the best thing that can be said about the recent pension proposals is that they are a step in the right direction. For the first time since ERISA, Congress has begun to reverse the tide of federal pension over-regulation. Many of the measures included in the Small Business Jobs Protection Act and other bills under consideration in the current Congress will indeed be helpful. We should not, however, pretend that the work on pension deregulation is finished.

What Comes Next?

Despite a fifteen-year experiment with nondiscrimination, the gap in pension participation between low-paid workers and high-paid workers is still quite large. In 1993, the most recent year for which data are available, only 28 percent of those making $20,000 per year or less participated in a pension plan, while 83 percent of those making $50,000 per year or more participated in a plan.10 Congress must pursue a more transcendent goal for pension law—one that looks beyond concerns of tax revenue loss to those of boosting retirement saving generally. We must move beyond marginal attempts to reduce the cost and complexity of administering a pension plan. Real simplification entails more than just alleviating part of the paperwork burden and making the language of the regulations easier to understand. Language simplification alone will not solve the problem, because the intent of many of the rules per se is too onerous. Employers must be allowed to offer higher pension compensation to those employees who desire it most. A more fundamental overhaul of pension rules is needed.

The key reforms should concentrate on the nondiscrimination rules, beginning with the repeal of the nondiscrimination contributions and benefits requirements in Code sections 401(a)(4) and 410(b)(2), and the top heavy rules in Code section 416. Congress should amend the section 410(b)(1) requirement that qualified plans must "benefit" 70 percent of NCEs (or 70 percent of the percentage of HCEs benefited) to require instead only that plans cover, or offer enrollment to, their ERISA-defined workforce.** Along with the additional participation requirements in section 401(a)(26), this change would still require firms to offer pension coverage to most rank-and-file workers, but it would permit a much broader range of benefit levels. It also would reduce significantly the cost and complexity of pension administration by eliminating the need for the testing of contributions or benefits.

Nondiscrimination rules have been a fundamental part of pension law since the 1920s, and they have played an increasingly larger role since the enactment of ERISA. Their goal is to encourage employers to cover low-paid workers or to deny tax preferences to those who do not. Despite wreaking havoc on the pension industry, nondiscrimination has had no substantial effect on pension coverage among low-paid workers. On the other hand, denying tax qualified status for highly compensated employees does not just prevent the "rich" from taking advantage of yet another tax loophole. After all, the amounts that even highly-compensated workers can save each year in qualified plans is already limited. The most pernicious effect of the non-discrimination rules is to preclude from pension coverage millions of middle-income Americans working for marginal firms.

What can Congress do to encourage pension creation among small businesses? Unfortunately, not very much. Pension sponsorship has always been dominated by large firms. Employees of small firms have never been likely to be covered by a pension for a number of reasons, including their preference for cash wages or other benefits (such as health insurance) first. Although we should not overlook the value of employer-sponsored pensions, we cannot expect marginal tinkering with the tax code to help in any significant way. Individuals have to want to save. It is the rare small business that can afford to subsidize the saving of low-paid workers. A better solution would focus on helping businesses create more and better jobs and making saving easier and more fruitful.

One way to expand individual opportunities to save, and to make up for any individual losses in pension coverage, would be to enact a general equalization of the annual contribution limits for IRAs to make them more closely equivalent to those available in employer-sponsored defined contribution plans. All those not covered by an employer-sponsored pension should be permitted to make tax-deferred contributions to IRAs equal to the lesser of the $9,500 deferral limit (indexed) for 401(k) plans or 25 percent of compensation (in this case, 25 percent of adjusted gross income). Workers covered under an employer-sponsored plan, but whose annual deferrals to them, or benefits accrued under them, are lower than the lesser of $9,500 or 25 percent of income should be allowed to supplement their retirement saving with tax-deferred contributions to an IRA with the combined contributions not exceeding the limit.

Another easy way to expand saving opportunities would be to make businesses eligible under section 403(b) "tax deferred annuity" provisions now available only to certain nonprofit organizations and public schools. These plans must be offered to all ERISA-defined employees, which inherently makes them nondiscriminatory in coverage, but individual workers decide how much, if any, to save. The plans are at least as easy as SEPs (and probably easier) to administer. They are inexpensive for employers, because administrative costs are paid by the workers from investment earnings. They also provide the disciplined, tax-advantaged retirement saving typical of an employer-sponsored plan. Even under current rules, employers face no real nondiscrimination requirements, so long as there is no match component.

Pensions and the National Savings Rate

Despite small rises in the rate of personal saving in recent years, the U.S. savings rate, hovering between 5 and 5½ percent,11 is very low compared with other industrialized countries and compared with historical domestic rates. Members of the ERISA Advisory Council already understand the value of retirement saving. Saving, however, is not just important for financing consumption during retirement. It is necessary for economic growth, which, in turn, generates growth in jobs and wages and generates a higher standard of living.

The tax advantages for pensions are a departure from the general treatment of saving under the federal tax code, where money invested is usually taxed twice. Under normal circumstances, income earned by capital is taxed first at the corporate tax rate, and again as income to individuals when it is paid in dividends. Pensions escape this second tax, because contributions are invested before taxes and investment returns accrue tax-free within the pension trust. Workers pay income tax only on the distributions during retirement. Much of these benefit payments are themselves shielded from taxation by personal exemptions and standard deductions.

Because pensions and the tax preferences for pensions continue to go mostly to upper-income earners, some critics have questioned the entire concept of providing tax preferences for pensions.12 Supporters of the tax preferences argue in return that (1) people should be encouraged to save in order to support themselves in retirement, (2) we need tax incentives to boost net national saving in the interest of economic growth, and (3) by not taxing fund income, firms are able to devote less money to pension saving—the rest then either can be reinvested to foster job growth, or it will go to current incomes that would be taxed anyway. These assertions are true, but do not seem to satisfy those who disparage the disparate treatment of pension saving. From a revenue perspective, Congress seems to be more than willing to deny tax preferences for saving that does not achieve its social goals.

The real injustice, however, is not that qualified pension saving escapes multiple taxation, but that it is one of the few avenues for doing so. Eliminating the multiple taxation of other forms of saving would boost personal saving, help boost net national saving, and would contribute to economic growth. It also would help workers whose employers do not sponsor pension plans, or those who wish to supplement an employer-sponsored plan with individual saving, to prepare adequately for their retirement. In addition, tax neutrality would further deteriorate justification for much of the demonstrably harmful government regulation of pensions.

This is not intended to overlook the fact that employer-sponsored pensions are especially valuable forms of retirement saving. Apart from their tax advantages, pensions offer participants an organized and more disciplined plan for accumulating retirement income. However, extending to all saving the main tax preference now enjoyed by certain forms of retirement saving would vastly simplify the tax code and remove existing disincentives for saving. The large increase in saving and investment likely to occur would contribute to economic growth, boost living standards, and help finance the growing burden of federal entitlement spending.

In the absence of fundamental tax reform, however, there are ample reasons for encouraging pension saving. As long as we continue to penalize saving generally, we should continue to encourage retirement saving through favorable tax policy that provides greater opportunities for individuals to save.


Pension benefits are deferred compensation. As such, benefits are awarded in much the same manner as cash compensation: in a manner intended to attract and to keep the best qualified and most productive workers. Total compensation is awarded in an amount reflecting each worker’s marginal productivity. Workers pay for pensions with reduced cash compensation. No amount of legislation can change these fundamental facts. While mandates may help lift some low-wage workers into the realm of pension coverage, its ultimate effect is to erect barriers keeping out many middle-class workers.

Since minimum vesting standards have reduced the utility of pensions for employers, the main reason they maintain pensions is to meet the preferences of workers. Where workers continue to demand pension coverage and are willing to take the tradeoffs, coverage will continue. Where they do not, coverage will lapse. This is true with or without the non-discrimination laws. Nondiscrimination stands as a barrier to marginal firms and marginal workers who are unwilling to bear the costs of compliance.

Today, 68 percent of America's pre-retirees and baby boomers name Social Security or an employer's pension as their primary source of retirement income.13 But Americans can no longer depend on Social Security to pay for much of their retirement. As the United States begins to confront a looming entitlement crisis, it has become apparent that private forms of saving will become increasingly important to retirees in the next century. Unfortunately, federal policy regarding saving, and pension saving in particular, inhibits adequate accumulation.

Regulatory attempts to expand pension coverage and limit the loss of tax revenue have increased compliance costs and made it far more difficult to establish and maintain a private pension plan. It is imperative that Congress undertake a serious examination of the nation’s pension laws and embark on a process of real deregulation. An elimination of the nondiscrimination rules is not only fair, it is an essential first step in the restoration of strong pension saving in the United States. To increase retirement income security, Congress must also expand existing saving programs and allow for innovative new ones. This cannot be done through further government mandates. Congress should forget trying to fine-tune these burdensome requirements, and discard them altogether.


1 Employment-Based Retirement Income Benefits: Analysis of the April 1993 Current Population Survey, Employee Benefit Research Institute Issue Brief Number 153, (Washington, DC: EBRI, September 1994), p. 21.

2 Ibid.

3 Janet Novack, "You call this simple?" Forbes, October 21, 1996, p. 308.

4 EBRI Issue Brief Number 153, p. 28.

5 Kathleen P. Utgoff, "Towards a More Rational Pension Tax Policy: Equal Treatment for Small Business," National Tax Journal, Vol. 44, September, 1991, p. 386.

6 Testimony of John J. Motley, III, Vice President for Federal Government Affairs, National Federation of Independent Business, before the Senate Committee on Finance, Hearing on Expanding the Use of Retirement Accounts, February 9, 1995.

7 Participant Education: Actions and Outcomes, Employee Benefit Research Institute Issue Brief Number 169, (Washington, DC: EBRI, January 1996).

8 Worker Investment Decisions: An Analysis of Large 401(k) Plan Data, Employee Benefit Research Institute Issue Brief Number 176, (Washington, DC: EBRI, August 1996).

9 James Poterba, Steven F. Venti, and David Wise, Lump-Sum Distributions from Retirement Saving Plans: Receipt and Utilization, National Bureau of Economic Research Working Paper Number 5298, (Cambridge, MA: NBER, October 1995).

10 EBRI Issue Brief Number 153, p. 24.

11 Congressional Budget Office, The Economic and Budget Outlook: Fiscal Years 1998-2007, (Washington, DC: CBO, January 1997), pp. 4-5.

12 See for instance, Alicia H. Munnell, "Current Taxation of Qualified Pension Plans: Has the Time Come?" New England Economic Review, March/April 1992, pp. 12-25; Munnell, "It’s Time to Tax Employee Benefits," New England Economic Review, July/August 1989, pp. 49-63; Munnell, "Employee Benefits and the Tax Base," New England Economic Review, January/February 1984, pp. 39-55; and Michael J. Graetz, "The Troubled Marriage of Retirement Security and Tax Policies," University of Pennsylvania Law Review, Vol. 135, No. 4, April 1987, pp. 851-908.

13 Merrill Lynch, A Rude Awakening from the American Dream, (Princeton, NJ: Merrill Lynch, Pierce, Fenner & Smith, Inc., June, 1992), p. 9.