Still a Better Bang for the Buck: An Update on the Economic Efficiencies of Defined Benefit Pensions
By William B. (Flick) Fornia and Nari Rhee
National Institute on Retirement Security (NIRS)
Editor’s note: Benefit generosity is a separate question from the economic efficiency of a retirement plan. While either a defined benefit or a defined contribution plan can offer more or less generous benefits, DB plans have a clear cost advantage for any given level of retirement benefit. The following is an excerpt from this paper, published by the National Institute on Retirement Security (NIRS). It may be found in its entirety here.
Over the past three decades, private employers have shifted away from defined benefit (DB) pensions that provide employees with a steady retirement income stream, towards defined contribution (DC) retirement accounts—such as 401(k) plans—in which individual workers manage their own investments. By and large, public employers have faced growing pressure since the 2008 financial crisis to make a similar change. Contrary to popular belief, however, DC retirement accounts are not inherently less costly than a pension, and switching from a DB to a DC system saves money only if it involves substantial benefit cuts.
In fact, DB pensions feature critical efficiencies that make them significantly less expensive to provide a given level of retirement benefit compared to DC plans. This was documented by the National Institute on Retirement Security (NIRS) in its 2008 study, “A Better Bang for the Buck: The Economic Efficiencies of Defined Benefit Pensions.”1 The study found that a typical large DB pension plan provides a given level of retirement benefit at about half the cost of a 401(k) style plan, because of three factors:
- The pooling of longevity risk in DB pensions enables them to fund benefits based on average life expectancy, and yet pay each worker monthly income no matter how long they live. In contrast, DC plans must receive excess contributions to enable each worker to self-insure against the possibility of living longer than average.
- DB pensions realize higher net investment returns due to professional management and lower fees from economies of scale.
- DB pensions are able to maintain portfolio diversification over time, while DC participants must shift to lower-risk, lower-return investments as they age. This means that over a lifetime, DB pensions earn higher gross investment returns than do DC accounts.
In summary, when it comes to providing retirement income, DB pensions are more efficient because they pool risks across a large number of individuals, invest over a longer time horizon, and have lower expenses and higher returns.
While these facts have not fundamentally changed since 2008, this study updates the comparison of retirement benefit funding costs based on an enhanced methodology that takes into account key changes in the DB and DC plan landscapes with regard to investment strategies and fees. We compare a typical large public sector DB pension to two kinds of DC plans—an individually directed DC plan with industry average fees and reduced investment returns based on typical investor behavior, and an “ideal” DC plan with fees well below industry average and asset class investment performance as strong as that achieved by professionals. Both DC plans are modeled with a target date fund (TDF) asset allocation pattern.
All three plans—the typical DB plan, the individually directed DC plan, and the ideal DC plan—are modeled with the same underlying demographic and economic assumptions regarding employee wage growth, retirement age, life expectancy, target monthly retirement income, inflation, and projected rates of return for each asset class. We also assume that all plans receive consistent, adequate contributions required to fund target benefits. In addition, we study the cost impact of annuitizing the account balances in the DC plans.
Even with updated assumptions and methodology, we still find that DB pensions offer substantial cost advantage over DC plans (see Figure 1 below).
- A typical DB plan, with advantages based on longevity risk pooling, asset allocation, low fees, and professional management, has a 48 percent cost advantage compared to a typical individually directed DC plan.
- A DB pension costs 29 percent less than an “ideal” DC plan with below-average fees and no individual investor deficiencies.
- Annuitizing DC account balances—that is, converting the account balance at retirement into an insurance contract for lifetime income—does not erase the DB pension cost advantage. This is because insurance companies use a more conservative asset allocation and charge much higher fees than a DB pension. Annuities purchased at historical average interest rates only modestly decrease DC benefit costs, while annuities purchased at 2014 rates would increase benefit costs.
In other words, a typical DB plan provides equivalent retirement benefits at about half the cost of a typical DC plan, and 29 percent lower cost than an ideal DC plan modeled with very generous assumptions.
Conversely, it would be 91 percent and 41 percent more expensive for a typical DC plan and an ideal DC plan, respectively, to deliver the same level of retirement income as a typical DB plan. Thus DB pensions continue to offer a significant cost advantage. While shifting from a DB pension to a DC plan offers a way to reduce the investment risk borne by employers and taxpayers, this comes with an unavoidable tradeoff—either increased benefit costs or, more likely, significant retirement benefit cuts that are larger than the savings realized by the employer.
Defined Benefit, Defined Contribution and Hybrid Plans
Employers who offer retirement benefits can consider two basic approaches: a traditional defined benefit (DB) pension plan and a defined contribution (DC) retirement savings plan. The DB plan is designed to provide predictable retirement income throughout a worker’s retirement years. Assets are pooled, and investments are managed by professionals who are responsible for acting in the best interest of participants. The DC plan, in contrast, is focused on accumulating retirement wealth expressed as a lump sum, with individual participants ultimately responsible for garnering adequate investment returns and managing their own accumulated wealth throughout their retirement years. This would entail estimating how much they can safely withdraw each year of retirement without running out of money, attempting to evaluate the best annuitization alternative in the open market, or some combination of the two.
Each type of plan has certain distinguishing characteristics that influence its cost to employers and employees.
How DB Plans Work
While employers have a large degree of flexibility in designing the features of a DB plan, there are some features all DB plans share. DB plans are designed to provide employees with a predictable monthly benefit in retirement. The amount of the monthly pension is typically a function of the number of years an employee devotes to the job and the worker’s pay—usually at the end of their career.2 For example, the plan might provide a benefit in the amount of 1.5 percent of final average pay for each year worked. Thus, a worker whose final average salary was $50,000, and who had devoted 30 years to the job, would earn a monthly benefit of $1,875 ($22,500 per year), a sum that would “replace” 45 percent of her final average salary after she stops working. This plan design is attractive to employees because of the security it provides. Employees know in advance of making the decision to retire that they will have a steady, predictable income that will enable them to maintain a fairly stable and predictable portion of their pre-retirement standard of living.3
Benefits in DB plans are pre-funded. That is, employers (and, in the public sector, most employees) make contributions to a common pension trust fund over the course of a worker’s career. These funds are invested by professional asset managers whose activities are overseen by trustees and other fiduciaries. A typical DB pension fund’s asset allocation policy—i.e., the share of holdings allotted to different asset classes such as stock, bonds, and treasuries—is based on a careful analysis of plan demographics and liabilities as well as short- and long-term financial market projections.4 The earnings that build up in the fund, along with the dollars initially contributed, pay for the lifetime benefits a worker receives when she retires.
How DC Plans Work
DC plans function very differently than do DB plans. First, there is no implicit or explicit promise of retirement income in a DC plan. Rather, the level of retirement income that an account will provide depends on a number of factors, such as the level of employer and employee contributions to the plan, the investment returns earned on assets, whether loans are taken or funds are withdrawn prior to retirement, and the individual’s lifespan.
While DC plan assets are also held in a trust, that trust is comprised of a large number of individual accounts. DC plans are typically “participant directed,” meaning that each individual employee can decide how much to save, how to invest the funds in the account, how to modify these investments over time, and how to withdraw the funds during retirement.
Retirement experts typically advise individuals in DC plans to change their investment patterns over their lifecycle. In other words, at younger ages, because retirement is a long way off, workers should allocate more funds to stocks, which have higher expected returns but also higher risks. As one gets closer to retirement, experts suggest moving money away from stocks and into safer but lower return assets like bonds. This is to guard against a large drop in retirement savings on the eve of retirement, or in one’s retirement years.
The high degree of participant direction makes DC plans very flexible in accommodating individuals’ desires, decisions, and control. Unfortunately, a substantial body of empirical and experimental research indicates that this flexibility tends to lead to adverse outcomes. First, too many workers fail to contribute sufficient amounts to the plans.5 Second, individuals’ lack of expertise in making investment decisions can subject individual accounts to extremely unbalanced portfolios with too little or too much invested in one particular asset, such as stocks, bonds, or cash.6 One team of researchers thus concluded, “The likelihood of investment success increases as the participant’s involvement in investment decisions decreases.”7
Another important difference between DB and DC plans becomes apparent at retirement. Unlike in DB plans, where workers receive regular monthly pension payments, in DC plans it is typically left to the retiree to decide how to spend down their retirement savings. Research suggests that many individuals struggle with this task, either drawing down funds too quickly and running out of money, or holding on to funds too tightly and enjoying a lower standard of living as a result.8 In theory, employers that offer DC plans could provide annuity payout options, but in practice they rarely do.9
Hybrid Retirement Benefits
There is growing interest in “hybrid” retirement benefits that combine some of the features of DB and DC plans, and ostensibly offload some risks onto employees while maintaining some of the retirement security offered by traditional DB pensions.
There are two main types. One type is a “side by side” or “stacked” hybrid, in which the core retirement benefit consists of a combination of a DB pension (typically with less generous benefits) and a DC plan. The other is a “blend” between DB and DC such as a cash balance (CB) plan. Under a CB plan, each employee has a notional account balance, as the employer credits each employee with a set percentage of her annual pay plus an interest rate that is either predetermined or tied to an index. A CB plan is legally a DB plan—benefits are guaranteed, albeit as a lump sum, and assets are pooled in a trust and managed professionally. However, CB plan benefits typically are less generous than a traditional DB pension, and generally participants do not obtain longevity protection.
Importantly, the relative costs of hybrid plans depend largely on benefit structure. To the extent that hybrid benefits emphasize DB-like characteristics, they can be more cost efficient. To the extent that they off-load risks onto individual workers, they will be less cost efficient.
DB Plans Reduce Costs by Nearly Half
Our analysis clearly demonstrates that DB plans are far more cost-effective than DC plans. We find that to achieve roughly the same target retirement benefit that will replace 53 percent of final salary, the DB plan will require contributions equal to 16.3 percent of payroll, whereas the individually directed DC plan will require contributions to be almost twice as high as the DB plan—31.3 percent of payroll. Even the “ideal” DC plan, generously modeled with the same fees and investor skill as the DB plan—provides benefits at a substantially higher cost of 23.0 percent of payroll.
We find that due to the effects of longevity risk pooling, maintenance of portfolio diversification, and greater investment returns over the lifecycle, a DB plan can provide the same level of retirement benefits at about 29 percent lower cost than an ideal DC plan and about 48 percent lower cost than an individually directed DC plan.
Table 1 breaks down the cost savings realized by the DB plan relative to the individually directed DC plan.
First, the longevity risk pooling that occurs in the DB plan accounts for 10 percent cost savings. Second, DB plans’ ability to maintain a more diversified portfolio drives another 11 percent cost savings. Third, superior net investments returns across the lifecycle generate an additional 27 percent reduction in cost compared to an individually directed DC plan—bringing the total cost savings to 48 percent.
Our results also indicate that DB plans can do more with less. That is, they can ensure that all individuals in the plan (even those with very long lives) are able to enjoy an adequate retirement benefit that lasts a lifetime, at the same time that they require less money to be contributed to a retirement plan and fewer assets to accumulate in the plan. We calculated the amount of money that would be required to be set aside for each retiree in each type of plan, to provide a modest retirement benefit of about $2,700 per month. As shown in Figure 7, at retirement age, the DB plan requires only about $500,000 to be set aside for each individual, whereas the ideal DC plan requires about $700,000 and individually directed DC plan requires about $800,000. The difference—about $200,000 and $300,000 for each and every employee under ideal DC plan and individually directed DC plan, respectively —illustrates that the efficiencies embedded in DB plans can yield large dollar savings for employers, employees and taxpayers.
Despite notable changes in the retirement benefit landscape since 2008, including some improvement in DC performance and fees, DB pensions retain their cost advantage as a means of providing retirement benefits to workers. In this study, we compared the cost of providing equivalent benefits through a typical large public sector DB plan, an ideal DC plan, and an individually directed DC plan. Even compared to the ideal DC plan with no disadvantage in terms of fees and investor skill, the DB plan reduces costs based on longevity risk pooling and the maintenance of portfolio diversification. And when we examine the individually directed DC plan with more realistic assumptions regarding fees and investor skill, the DB plan realizes a hefty additional cost advantage due to its low expenses and professional management of assets.
The sources of cost savings in DB plans reflect, at a very basic level, the differences in how DB and DC plans operate. Group-based DB plans provide lifetime benefits and feature pooled, cost-efficient, professionally managed assets. These features drive significant cost savings that benefit employers, employees, and taxpayers. While well-designed DC plans can theoretically mimic some of these advantages—for instance, employers may select low-fee TDFs as a default investment option for their workers—DB plans would still retain their advantages of longevity risk pooling and long-term portfolio diversification. Using private annuities to convert DC account balances at retirement into a lifetime income stream does not close this gap because such annuities are expensive, especially when they include the kind of inflation protection offered by public DB plans.
When considering our results, it is important to keep in mind that in our effort to construct an “apples to apples” comparison, we made a number of simplifying assumptions that actually reflected more favorably on DC plans. For instance, we did not model any asset leakage from either the ideal or individually directed DC plan before retirement through loans or early withdrawals. We also assumed that individuals followed a sensible “Goldilocks-like” withdrawal pattern in retirement— not too fast, not too slow, but just right. We used conservative estimates of the difference in actual investment returns between DB and DC plans. And, we used 80th percentile life expectancy to project required accumulations in the DC plans, rather than “full” life expectancies.
Thus, if anything, our analysis underestimates the cost of providing benefits in a DC plan and thereby understates the cost advantages of DB plans.
Due to the built-in economic efficiencies of DB plans, employers and policymakers should continue to carefully evaluate claims that “DC plans will save money.” As discussed, benefit generosity is a separate question from the economic efficiency of a retirement plan. While either type of plan can offer more or less generous benefits, DB plans have a clear cost advantage for any given level of retirement benefit. Consequently, shifting from a DB plan to a DC plan and maintaining the same contribution rate will generate significant cuts in retirement income. Considering the magnitude of the DB cost advantage, the consequences of a decision to switch to a DC plan could be dramatic for employees, employers, and taxpayers.
Finally, policymakers should consider proposals that can strengthen existing DB plans and promote the adoption of new ones. When viewed against the backdrop of workers’ increasing insecurities about their retirement prospects and the economic and fiscal challenges facing employers and taxpayers, now more than ever, policymakers ought to focus their attention and energy on this important goal. The very features that make DB plans attractive to employees drive cost savings for employers and taxpayers. In this way, DB plans represent a rare “win-win” approach to achieving economic security in retirement that should be recognized and replicated.
1 B. Almeida and W.B. Fornia, 2008, “A Better Bang for the Buck: The Economic Efficiencies of Defined Benefit Pension Plans,” National Institute on Retirement Security, Washington, DC.
2 The benefit factor could also be a function of a worker’s earnings over their entire career (a so-called “career average plan.”) Or, the factor could be a flat dollar amount: for example, the plan will pay a monthly benefit equal to $50 per year of service, so that a 30 year employee would have a benefit of $1,500 per month. “Flat dollar” plans are primarily seen among blue-collar workers in the private sector.
3 Inflation protection varies among DB pensions. Private DB pensions typically do not offer Cost of Living Adjustments, while public DB pensions usually offer some level of inflation protection.
4 R. Jung and N. Rhee, 2013, “How Do Public Pensions Invest? A Primer,” National Institute on Retirement Security, Washington, DC.
5 While not incorporated into our model, the lack of sufficient contributions can be a problem for DB plans and is a widespread problem for voluntary DC accounts. The median 401(k) contribution rate among participating workers peaked at 5.2 percent in the early 2000s, and stood at just 5 percent in 2010. (See Table 6, p. 44 in B.A. Butrica and K.E. Smith, 2012, “401(k) Participant Behavior in a Volatile Economy,” CRR WP 12-24, Center for Retirement Research at Boston College, Chestnut Hill, MA.) In addition, 29 percent of private sector wage and salary employees who have access to a 401(k) type plan do not participate (U.S. Bureau of Labor Statistics (BLS), 2013 National Compensation Survey Employee Benefit Survey, BLS, Washington, DC, http://www.bls.gov/ncs/ebs/benefits/2013/ ownership/private/table02a.htm.)
6 There is a wealth of research on behavioral biases in retirement saving and investing. See for instance S. Benartzi and R. Thaler, 2007, “Heuristics and Biases in Retirement Savings Behavior,” Journal of Economic Perspectives, v21n3: 81-104. For an accessible overview of research in this field, see S. Benartzi, 2007, “Implications of Participant Behavior for Plan Design,” Alliance Berstein.
7 J.C. Chang, S.W. Simon, and G.K. Allen, 2005, “A Step Beyond Erisa Section 404(c): Improving on the Participant Directed 401(k) Investment Model,” Journal of Pension Benefits, v12n4.
8 C. Copeland, 2007, “How Are New Retirees Doing Financially in Retirement?,” EBRI Issue Brief. No. 302, Employee Benefit Research Institute, Washington, DC; D. Love, P.A. Smith and L. McNair, 2007, “Do Households Have Enough Wealth for Retirement?” Finance and Economics Discussion Series. 2007-17, Federal Reserve Board, Washington, DC.
9 P. Perun, 2007, “Putting Annuities Back into Savings Plans,” In T. Ghilarducci and C. Weller, eds., Employee Pensions: Policies, Problems, and Possibilities, Labor and Employment Relations Association, Champaign, IL.
This paper was published by the National Institute on Retirement Security, a non-profit research institute established to contribute to informed policy making by fostering a deep understanding of the value of retirement security to employees, employers, and the economy as a whole. NIRA works to full this mission through research, education, and outreach programs that are national in scope. This excerpt is used with permission.
About the Authors
William B. (Flick) Fornia is President of Pension Trustee Advisors, Inc., specializing in public sector retirement plans. He has 35 years of actuarial and consulting experience, primarily in the areas of retiree pension and healthcare benefits. Mr. Fornia is an author and frequent speaker on all aspects of retirement programs including financing, design and litigation. Mr. Fornia earned a Bachelor of Arts in Mathematics at Whitman College. He is a Fellow of the Society of Actuaries, Enrolled Actuary, Member of the American Academy of Actuaries, and Fellow of the Conference of Consulting Actuaries. He currently serves on the Faculty of the Society of Actuaries Fellowship Admissions Course, and the Conference of Consulting Actuaries Public Pensions Community Steering Committee.
Nari Rhee is Policy Specialist at the Institute for Research on Labor and Employment/Center for Labor Research and Education at the University of California at Berkeley, focused on research and policy development to improve the retirement security of low-wage workers. She served as Manager of Research for the National Institute on Retirement Security from September 2012 to November 2014, conducting research on issues ranging from state level public pension reform to the private sector retirement savings crisis. She holds a PhD from the University of California at Berkeley and an MA from the University of California at Los Angeles. She is a member of the National Academy of Social Insurance.
We are grateful to the members of the Advisory Committee which provided valuable data, insight and advice on the design of this study as well as comments on an earlier draft of this report: Mike Heale, Rocky Joyner, David Kausch, Rebecca Merrill, Josh Shapiro, Karl Paulson, and Ron Peyton. We also thank Ron Baker and Leslie Oliver for providing comments on the draft report. However, the views and any errors in this report are those of the authors alone.
Archive – CE available with all 2019 and 2018 issues
- 2019 2nd Qtr Issue
- 2019 1st Qtr Issue
- 2018 4th Qtr Issue
- 2018 3rd Qtr Issue
- 2018 2nd Qtr Issue
- 2018 1st Qtr Issue
- 2017 4th Qtr Issue
- 2017 3rd Qtr Issue
- 2017 2nd Qtr Issue
- 2017 1st Qtr Issue
- 2016 4th Qtr Issue
- 2016 3rd Qtr Issue
- 2016 2nd Qtr Issue
- 2016 1st Qtr Issue
- 2015 4th Qtr Issue
- 2015 3rd Qtr Issue
- 2015 2nd Qtr Issue
- 2015 1st Qtr Issue
Other past issues (earn CE for 2019 and 2018 issues)