Claim that 401(k)s Beat Defined Benefit Plans Stirs Controversy
A new study reignites the defined contribution versus defined benefit debate, and raises as many questions as it answers.
A basic — and increasingly nostalgic — assumption in the private pension world has long been that the financial payouts of defined benefit plans are much better than those of defined contribution plans, and it’s too bad that defined benefit plans seem to be heading for extinction.
Now an Employee Benefit Research Institute study has cast doubt on the conventional wisdom.
The study compared the tens of thousands of 401(k) plans in EBRI’s long-term database with two current defined benefit plan models — a standard three-year, final-average-pay pension and a cash balance plan. Institute researchers found that the 401(k) benefits were better for almost every age and income cohort. The median differences ranged from about even to 44 percent, with higher earners and those with longer tenure generally doing best with defined contribution plans.
How could that be?
“Conventional wisdom always made defined benefit plans more advantageous than they really are,” says Jack VanDerhei, EBRIs research director and the author of the study.
Not so fast, retort critics, arguing that the study makes a lot of assumptions that are biased toward 401(k)s. “In its day, a defined benefit pension was a better way of delivering benefits,” says Alicia Munnell, director of the Center for Retirement Research at Boston College. “What the paper is doing is looking at the effect of an institution in a world that no longer exists.”
One problem is that the report includes only voluntary 401(k)s, ignoring the automatic enrollment variety — even though the latter constitute more than half of all 401(k)s and are steadily gaining ground, according to the consulting firm Aon Hewitt. The study omitted them because researchers say there was not enough data available on participants who opt out of certain types.
In some ways, that methodology inherently tilts the conclusion in favor of defined contributions. That’s because the average contribution rate in auto-enrolled plans — 6.6 percent of salary, according to Aon Hewitt — is lower than the average rate of 7.9 percent in the voluntary type. (The most common auto-default rate — 3 percent — is even worse.) Thus, in the real world, most employees actually have significantly less money in their 401(k)s than the samples in this survey, reducing any comparative advantage against traditional pensions.
But VanDerhei argues that if the numbers are analyzed in terms of what really matters — how much money people are accumulating for retirement — the inclusion of auto-enrollment plans would actually boost defined contribution’s results versus defined benefits because automatic plans cover more people. “Auto-enrollment increases low-income participation and thus increases their benefits,” he says.
Munnell also points out that the study uses “extraordinarily high assumed rates of return for equities” in 401(k) plans — 8.6 percent. VanDerhei says this is the average historical return for U.S. stocks. However, rare is the amateur 401(k) investor who consistently matches the market’s historical returns. The paper also admits that, if the assumed return rate is lowered by 200 basis points and the purchase price for annuities is raised “to reflect today’s bond rates, results show that in many cases the [voluntary enrollment] 401(k) plans lose their comparative advantage” for lower-paid employees.
Even if all of the financial assumptions were accurate, the report largely ignores the concept of opportunity costs. Private pensions are essentially free money for employees, paid directly by the employer. By contrast, employees have to fork up most if not all of the contributions that go to build their 401(k) benefits. If they had a defined benefit, even a small one, “they could have used that 401(k) money for something else,” says Alan Glickstein, a senior consultant at Towers Watson.
The paper concedes this point as a “caveat,” promising that “a more nuanced approach to dealing with this difference will be dealt with in a future EBRI publication.”
Ultimately, the biggest controversy revolves around whether any defined benefit–defined contribution comparison even makes sense. The paper assumes — realistically, in today’s world — that people don’t stay with the same employer for their entire career. That is not a problem for a portable defined contribution plan. However, a traditional pension doesn’t work for frequent job-hoppers. Employees would have to restart the benefit accrual from zero at each new job — assuming they could keep finding jobs with defined benefit plans — which would diminish the benefit because the payout is based in part on the employee’s longevity at that company.
This might seem to bring the argument back to the conventional wisdom: Pensions are nice, but they are gone with the wind. However, VanDerhei says his calculations serve a valuable purpose in countering part of that nostalgia — the theory that, due to the disappearance of defined benefits, “retirement income has gone down so incredibly much that GenXers are going to be worse off than Boomers.”
Karen Ferguson, director of the Pension Rights Center in Washington, D.C., says the paper’s chief value could be to add to the ongoing debate over designing a better retirement system that might include the best features of both defined benefit and defined contribution plans. One solution implied in the paper is cash balance plans with higher contribution rates. That may already be happening. Since the passage of the Pension Protection Act in 2006, says Towers Watson’s Glickstein, “many of the ones we work with have more complicated formulas now,” including market-based rates.
VanDerhei says he agrees that cash balance plans with a higher employer contribution could be a good solution. “You don’t have to worry about the number of employees participating,” he points out. “You don’t have to worry about the investment risk” falling on employees. But the key question, he says, is “how high would the contribution have to be?”