Download PDF Chairman Harkin, Ranking Member Enzi and Members of the Committee:
Thank you for the opportunity to testify at the roundtable with regard to Pension Modernization for a 21st Century Workforce. I have structured my written testimony as answers to the questions posed by the Chairman and Ranking Member.
Question 1: Defined benefit pensions have provided a secure retirement for millions of middle class Americans, but it is clear that the traditional pension system is in decline and that existing defined benefit pension models may not be well-suited for some of our 21st Century workforces. What should our pension system look like to meet the challenges of the global economy and the need to provide retirement security for working Americans?
A defined benefit (DB) pension plan offers certain important advantages: simplicity and predictability of benefits, protection against market risk, and insurance against outliving your retirement assets, all of which are highly valuable to employees. Of course, these benefits to employees inevitably come at a cost to employers, particularly smaller businesses, who may be ill-suited to managing market and longevity risks.
Moreover, DB pensions carry certain disadvantages in a modern economy. In particular, DB pensions lack portability and are discriminatory against short-term employees, who often are the young, mobile professionals that high-tech businesses seek to attract. There is evidence that a significant part of the shift from DB to defined contribution (DC) pensions was driven, not by businesses, but by the employees businesses seek to attract.1 In addition, certain individuals may desire the greater liquidity that DC pensions allow for relative to DB plans.
A pension system for the 21st Century will attempt to capture the advantages of DB pensions while avoiding their disadvantages. It will be:
- Streamlined: Because complexity discourages participation and encourages mistakes.
- Low-cost: Because higher administrative costs eat into employee savings. I understand the desire to provide ordinary savers with “professional money management,” but I fear that the extra costs are rarely justified by higher returns and, in the case of more exotic investment strategies, carry risks that often are ill-understood.
- Convertible to a lifetime income: Most economists hold that annuities are extremely valuable relative to lump sum pension payouts, because they efficiently manage against the problem of not knowing how long one’s retirement savings must last. It remains a puzzle why so few Americans choose to purchase annuities; the expected causes, such as administrative costs or the presence of Social Security, which already pays benefits as an annuity, do not provide a full explanation.
- Financially transparent: One method by which governments implicitly encourage pensions, either for state/local government employees or through the Pension Benefit Guaranty Corporation (PBGC), is by allowing them to operate under accounting precepts that differ from how private financial markets would value liabilities and risk. One example is allowing pensions to use higher discount rates to value their liabilities. These efforts, however well-intentioned, are dangerous both to the plans and to taxpayers and should not be tolerated. If the government wishes to subsidize pensions it should do so explicitly through appropriations, notby facilitating accounting arbitrage that encourages underfunding and excessive risk-taking. Policy efforts to encourage retirement income security should be very careful not to go down this road.
Question 2: What would make it easier and attractive for businesses-especially small businesses-to provide their employees with a traditional pension benefit? Would reducing the employers' risk and plan complexity help?
Pensions, both DB and DC, can carry significant regulatory and administrative burdens. While large firms can bear these costs, for smaller employers they may be prohibitive.2 Reducing the cost and complexity of pension provision may encourage more widespread adoption.
Reducing employer risk in provision of DB pensions or similar types of benefits would presumably encourage employers to offer such plans. However, we should be wary for two reasons. First, because DB pensions have shrunk despite the fact that most economists believe that the pension insurance provided by the PBGC is already significantly underpriced. In other words, we have a natural experiment in the degree to which subsidies can increase DB pension coverage. Clearly these were not sufficient to maintain DB plans even in larger employers, much less expand coverage among small employers.
And second, the government is capable only of transferring risk, not of eliminating it. If employers are relieved of risk with regard to financing DB pension plans, that risk is simply shifted onto taxpayers. Simply because federal budget rules do not adequately account for the cost of risk does not mean it does not exist. In the wake of Fannie Mae and Freddie Mac, of TARP and of the General Motors bailout, the potential cost of contingent liabilities should be foremost in the minds of policymakers. They should consider costs to the taxpayer of a policy not based on what is expected to happen, but on cases in which things go wrong.
Question 3: What do employees need from a pension plan to ensure that they will have a secure retirement?
A pension plan should have a variety of characteristics. First, adequate levels of saving, although these may differ significantly from person-to-person. The amount a person should save for retirement may differ based upon their income, their marital status, the number of children they have, and other factors. Unfortunately, there is no simple rule that will dictate a correct saving rate for each person and we should be wary of policy solutions that do not allow flexibility in this regard.
Second, simplicity of design. I have shown in other work that many Americans, even those on the verge of retirement, have very little idea what they will receive from Social Security despite receiving annual benefit statements.3 This is due, I believe, to excessive complexity of the Social Security benefit formula. Traditional DB pensions are simpler than Social Security and this simplicity should be retained.
Third, lifetime income. Management of retirement income cash flows is difficult in the face of uncertain life spans. A DB plan provides annuitization automatically, while a DC plan may do so through the purchase of annuities with account balances at retirement. Alternatively, individuals might purchase deferred annuities on an annual basis, thereby building up a pseudo-DB benefit within a DC structure. In whatever context, however, greater attention to managing the decumulation phase of retirement saving may be helpful to middle income households for whom Social Security is not the major source of retirement income.
1 Aaronson, Stephanie, and Julia Coronado. 2005. "Are firms or workers behind the shift away from DB pension plans?" Federal Reserve Board Finance and Economics Discussion Series Working Paper No. 2005-17. Washington, DC: Federal Reserve Board of Governors.
2 See Perun, Pamela J. and Steuerle, C. Eugene. "Why Not a 'Super Simple' Saving Plan for the United States?" The Urban Institute. May 2008.
3 Biggs, Andrew G. “Improving the Social Security Statement.” (October 2010). Financial Literacy Center Working Paper No. WR-794-SSA.