BC SealBoston College Magazine Spring 2004
In This Issue
Linden Lane
Works and Days
Letters to the Editor
BCM Home
Contact BCM
Coming Events
. Features

Retirement blues


Coconut Creek, Florida, 2001. By Ed Kashi/Corbis

Coconut Creek, Florida, 2001. By Ed Kashi/Corbis

Twenty-five years after the rise of the 401(K), the first do-it-yourself generation of investors is about to retire. Will the baby boomers—and their younger co-workers to follow—go bust?

By Alicia Munnell

Even before the collapse of Enron highlighted the misuse of employee retirement accounts, and before the current protracted bear market reduced the nest eggs of many workers, questions about how to provide Americans with adequate retirement income were high on the national policy agenda—and with good reason. The number of Americans over age 65 will double by 2030. With a life expectancy at age 65 of roughly 20 years, today's workers will spend more time as retirees than any generation before.

So far, concerns have centered on Social Security, the basic tier in the U.S. retirement system. Social Security today provides benefits equal to 41.3 percent of pre-retirement earnings for the average worker retiring at age 65, or 38.5 percent after the deduction for the Medicare premium. By 2030, however, the figure will likely be only 26.3 percent, as Medicare premiums rise, taxation of benefits increases, and benefit cuts are phased in, along with other measures designed to eliminate the program's structural deficit.

Receiving considerably less attention, though, has been the second tier of retirement security: employer-sponsored benefit plans, including most notably the 401(k). These programs are a crucial source of retirement income for middle-class families, and a close examination raises grave questions about how well they will provide for the future, if the current habits of Americans—whether young adult, middle-aged, or in retirement—continue. Put frankly, the nation appears headed toward a financial crisis among its elderly that will change the face of growing old.

RETIREMENT PLANS aren't what they were 20 years ago. In the 1970s, most people with coverage had a traditional, defined-benefit plan. It typically provided lifelong monthly retirement checks, based on years of service and final salary. The annuity might be a percentage of final salary for each year of service, say 1.5 percent, so workers with 20 years of service would receive 30 percent (20 years at 1.5 percent) of final salary for as long as they lived. The employer financed the benefit by making pretax contributions into a pension fund; the employee typically did not contribute. The employer held the assets in trust, directed the investments, and bore the risk.

We live now in a 401(k) world. The 401(k) plan is essentially a tax-deferred savings account, into which the employee—often joined by the employer—contributes a specified percentage of earnings. The money is invested, usually at the direction of the employee, mostly in mutual funds of stocks and bonds. When the worker retires, the balance in the account determines the retirement benefit, almost always paid in a lump sum.

It is easy to understand the popularity of the 401(k). The account is portable, which means that today's mobile workers can take their balances with them from job to job. Many plans permit borrowing, which adds an element of liquidity. And as the 401(k) allows employees to choose investments that match their personal tolerance for risk, it fosters a sense of control. Rapidly rising account balances during the stock market boom of the 1990s greatly enhanced the 401(k)'s popularity. With the sharp market decline in 2000, enthusiasm may have waned somewhat.

Employers also like 401(k)s. When the stock market plummets, it is the worker, not the employer, who loses money. When interest rates fall, rather than the employer having to dig down to make up the annuity, the retiree simply realizes a lower income.

The 401(k) and similar defined-contribution plans are now the dominant form of private pension in the United States. Between 1981 and 2001, the percentage of people with pension coverage who relied solely on such plans rose from 19 percent to 57 percent. At the same time, the share of comparable workers who depended solely on employer-controlled, defined-benefit plans dropped from about 58 percent to 13 percent. Although the 401(k) and similar plans emerged only in 1981, they will determine the economic security of a significant portion of the baby boom generation. And by and large, they are being mishandled.

ON PAPER, the formula is promising. Simulations suggest that workers can accumulate substantial savings for retirement using a 401(k) plan. For example, a worker participating steadily from the age of 30, who starts out with a salary of $17,000 and ends up with a salary of $52,560 at age 62, will accumulate pension wealth equal to $353,408—or 6.7 times final earnings—an amount that, combined with Social Security income, will provide an adequate replacement of pre-retirement earnings.

But actual accumulations in 401(k) accounts are significantly lower. The median combined balance of a 401(k) and an individual retirement account (IRA) for household heads in their late forties and early fifties is $37,000. Even employees in their sixties have assets equal to only 2.9 times earnings, less than half the 6.7 factor needed.

Worse still, as recently as 2001, some 26 percent of workers who had the chance to sign on to a 401(k) plan failed to do so. The reasons vary. In one survey, many explained that they "couldn't spare the money." This statement suggests that low-income workers may be less likely to have a 401(k), and the evidence bears this out. Among eligible workers, participation is around 50 percent for those earning less than $20,000, but approaches 90 percent for those earning more than $80,000. The voluntary nature of 401(k) plans appears to put low-income workers at greater risk of failing to accumulate enough retirement wealth.

Age also affects participation. The sooner workers start contributing, the better off they will be in retirement, yet workers age 20 through 29 are less likely to pay into a plan than older workers. Indeed, fewer than half of eligible workers age 20 to 40 with earnings below $20,000 choose to participate in a 401(k). And even for those young workers with salaries close to average ($20,000-$40,000), participation rates fall below 70 percent. It is not until these workers enter their fifties that the gap in participation that separates them from higher earners becomes negligible. Unfortunately, a worker who postpones participating in a 401(k) until age 50 will have only 26 percent of the retirement wealth of a similar worker who started paying in at age 30.

THE OPTION TO participate in a 401(k) is only the first of several key crossroads where workers can and do go wrong. The next crucial decision is how much to contribute. Though this can dramatically affect a person's wealth in retirement, few participants pay in as much as the law allows. For example, as the contribution limit rose from $10,500 in 2001 to $13,000 in 2004, fewer than 10 percent of workers raised their payments to keep pace. Not surprisingly, employees with higher earnings are far more likely to contribute the maximum amount. About 53 percent of Americans earning more than $100,000 choose to do so, compared with less than 1 percent of those earning between $20,000 and $40,000. Several studies show that employees' contributions tend to cluster at the level of their employer's matching rate. That is, if the employer matches up to 6 percent of earnings, employees are likely to contribute 6 percent to take full advantage of the match.

THE NEXT CRUCIAL decision workers face is how to invest their contribution. Modern portfolio theory provides for diversifying one's holdings over stocks, bonds, and fixed income securities, so as to balance risk and return. The portfolio should shift moderately away from stocks and toward bonds as a worker gets older, and it should be rebalanced from time to time to correct for the generally higher growth rate of stocks. Unfortunately, most 401(k) participants don't follow these strategies. Indeed, more than half have either most of their accounts invested in stock or nothing at all invested in stock. As a result, the majority risk either ending up with inadequate retirement income or being exposed to large swings in the value of their retirement assets.

In particular, low-wage participants are more likely to hold no equities and to invest their entire 401(k) accounts in interest-earning assets such as guaranteed investment contracts issued by insurance companies and money market funds. That means they are more likely to see a return of about 2.4 percent (adjusted for inflation), in contrast to the historical 7.2 percent return on stocks. This difference in return has an enormous impact on accumulations 30 years down the road. Investing $1,000 for 30 years produces $2,098 (in today's dollars) from a no-stock portfolio, $4,141 from a portfolio that is half stocks and half bonds, and $8,050 from an all-stock portfolio.

Of course, the higher returns associated with stocks also bring more volatility, and one might argue that low-wage workers cannot accept that additional uncertainty. That is, lower paid workers already face a lot of earnings risk, including greater odds of becoming unemployed, and they therefore need to invest in relatively safe assets. On the other hand, some economists contend, the uncertainty in earnings for most people is unrelated to the performance of the stock market, so it makes no sense for low-income workers to shun stocks and forgo the higher returns. Further, unless low-income workers make stock investments, at least when they are young, their savings are unlikely to provide adequate replacement income when they retire.

WHAT ABOUT the employees who do diversify? Do they allocate their 401(k) assets wisely? The evidence suggests that in general they do not.

Participants in 401(k) plans have a myriad of choices. Between 1995 and 2000, for example, Fidelity Investments, which represents 18 percent of the holdings of defined-contribution plans, more than doubled the number of its investment options. Plans for groups of at least 10,000 participants, which offered 14 options in 1995, now offer on average 38. These options include money market funds; guaranteed investment contracts; short-term, intermediate, and long-term bonds; high-yield bonds; large-cap and small-cap domestic equities; global equities; and equities from emerging markets. Granted, Fidelity's menu may be at the high end, but it nonetheless represents the trend: more choices.

So many options can overwhelm a person, and the psychology literature suggests that when individuals are overwhelmed, they often fall back on simple rules of thumb. Some researchers believe that 401(k) participants tend to opt for the 1/n solution. That is, they divide their contributions equally among all investment options. If their plan offers only a few choices, this strategy may result in a diversified portfolio. On the other hand, if the plan offers a disproportionate array of equity funds—as many plans do— these participants end up investing too large a share in stocks. Participants who hold mostly stock as they approach retirement run the risk that a market downturn will significantly reduce the size of their 401(k) account, as happened to many people during the bear market of 2000. They may have to postpone retirement, or face a hard drop in their standard of living; and they may still end up with lower retirement income than they expected.

Americans' investment allocations look even worse when their employer company's stock enters the picture. As the Enron debacle illustrates, it is important to avoid concentrating both one's income and one's investment security on a single bet. Nevertheless, U.S. employees remain enthusiastic actors in the drama. Eight million workers—approximately 20 percent of all 401(k) participants—hold more than 20 percent of their 401(k) assets in company stock, many for the simple reason that it is familiar. Among large corporations with the highest percentage of employees' assets lodged in company shares, Procter & Gamble ranks at the top, with 94.7 percent. Following closely are Coca-Cola (81.5 percent) and General Electric (77.4 percent). To put these concentrations in perspective, Enron's employees, at their peak, held a relatively modest 60 percent of their 401(k) assets in company stock.

New York City, 2003. By Eugene Richards/Magnum

New York City, 2003. By Eugene Richards/Magnum

UNDER THE CONTROL of their individual owners, 401(k)s also have an unfortunate potential for leakage. Even workers who maximize their contributions and invest their funds wisely can substantially reduce their retirement savings when they borrow from their 401(k) and fail to repay or—a greater concern—neglect to reinvest their account after a job change.

Whenever workers change jobs they have the option of taking a lump-sum distribution of their 401(k) instead of rolling over those assets into another plan. And the incidence rate of leakage when this occurs warns of serious problems. For example, tax data show that in 1995 lump-sum distributions from 401(k)s represented 9.1 percent of all 401(k) assets. Of these distributions only 77 percent were rolled over into an IRA or another 401(k) plan.

In particular, small accumulations held by young people tend to be held out and spent. When researchers examine the role that age, income, and dollar amount play in the decision not to roll over, they find that the single most important determinant is the size of the account. Less than 30 percent of participants who receive a distribution of $10,000 or less roll over their funds. (Laid-off workers are another group likely to cash in their distributions.)

Using the same simulation of a worker's path to retirement described above illustrates the potential harm. If our typical participant, scheduled to retire with $353,408, changes jobs at age 35 and uses his 401(k) distribution to buy a new car and go on vacation, his retirement wealth at age 62 will be reduced by 22 percent.

BUT LET'S SAY our typical worker does everything right. Throughout his career he's made the wise decisions that yield the full expected value of his 401(k), and the time has come for him to take his lump sum and retire. Now what? In truth, that is the toughest question he will face. Yet it is also where the research thins. Because 401(k)s are a relatively recent phenomenon and most participants are still in the workforce, few studies have been done on the withdrawal phase. Yet as more plans mature and millions of baby boomers begin to retire, how 401(k) money is directed in retirement will become an increasingly important issue.

One thing seems certain: Without the regular annuities provided by past employer-managed pension plans, and with Social Security's income replacement rates scheduled to shrink, members of the baby-boom generation are going to have a difficult time in retirement. They run two risks: of consuming their nest egg too quickly, thereby using it up, or of consuming it too slowly, thereby restricting unduly their standard of living.

Both these risks could be eliminated through the purchase of an annuity, but the annuity market in the United States remains tiny. One reason is that the average American views annuities as being expensive—as being a gamble with the insurance company that the company is favored to win. Viewed another way, however, the only true cost to the buyer is that the annuity payments stop at death; if retirees place no value on wealth after death, the cost of the annuity is zero.

But another reason that Americans are reluctant to annuitize is, in fact, their desire to leave a bequest. Asked by researchers "Do you (and your spouse) think it is important to leave an inheritance to surviving heirs?" 67 to 78 percent respond that it is at least somewhat important. Many Americans are also reluctant to annuitize out of concern that they may yet incur large unanticipated expenses, especially related to their health. And others fear a gradual dwindling of an annuity's value, since virtually all annuities in the United States share a common inadequacy—they are fixed in nominal terms and therefore do not protect against inflation.

Several approaches might encourage greater annuitization of retirement wealth. First, make annuities—indexed for inflation, of course—the default payout mechanism under all defined-contribution plans. For married couples, the default annuity should be what's called a joint-and-survivor plan. The lesson learned in the early years of traditional pensions is that employees will otherwise fail to provide for their spouses; as husbands generally selected the single-life annuity for its higher monthly benefits, wives, who typically outlive their husbands, lost all pension income when their husbands died.

Second, the insurance industry should design annuity products that better respond to people's concerns. If the major worry is future medical expenses, especially the cost of long-term care, combine a fairly priced annuity with extra benefits payable at the onset of disability. Similarly, if people do not want to lose the "bet" and leave their money to the insurance company, design an annuity pool so that they explicitly leave their money to people with whom they choose to share—fellow musicians, or teachers, or maybe just an extended family if large enough. If the purchaser dies early, he is assured that people he cares about directly or indirectly will have a continued stream of income thanks to his participation. (Of course, even in a regular annuity the money from those who die never goes entirely "back to the insurance agency," but is used to pay benefits for those who live a long time.)

And too, give government a role to play in specifying standards and serving as a clearinghouse directing consumers to the companies that meet them. The goal would be a public-private partnership.

AS WE HAVE SEEN, 401(k) plans have come up short. Although workers in theory can accumulate substantial wealth under the plans, in practice they do not. The problem is that the entire burden is on employees, and many make mistakes at every step along the way. If we don't find a way to solve this problem, we can expect more retirees to depend on the social safety net, and that means more stress on such government programs as Supplemental Security Income (SSI), which supports impoverished older people, and Medicaid, which pays long-term care costs for those who cannot afford them. Changes are clearly needed to make 401(k) plans a more certain vehicle for retirement income. One obvious solution would be to take advantage of participants' inertia.

Studies have shown that procrastination and inertia are important explanations for lack of participation in 401(k) plans; for the fact that those who do participate rarely change their contributions; and for the fact that few rebalance their portfolios to reflect either age or asset performance. Pension plans are complicated, and to many participants, the decisions appear overwhelming. Often it seems easiest just to do nothing.

Public policy could leverage this inertia by setting the defaults in 401(k)s to the desirable outcomes. That is, all eligible participants would be automatically enrolled; their contribution would be set at the level of the employer match; their portfolio, say, at age 30, would be 70 percent stocks and 30 percent bonds and automatically rebalanced as they aged; investment in company stock would be restricted. In addition, lump-sum distributions resulting from a job switch would automatically roll over to a new retirement account; and all plans would convert upon retirement to inflation-indexed, joint-and-survivor annuities.

Of course, individual employees could opt out at any stage. But research indicates that people tend to stay where they are put, and these defaults would put them in a much better place, and the country along with them.


Alicia Munnell is the Carroll School's Peter F. Drucker Professor of Management Sciences and director of the Center for Retirement Research at Boston College. She has served as senior vice president and director of research at the Federal Reserve Bank of Boston and, in the Clinton administration, as assistant secretary of the treasury for economic policy and a member of the President's Council of Economic Advisors. Her essay is drawn from Coming Up Short: The Challenge of 401(k) Plans by Alicia Munnell and Annika Sundén (2003), by permission of the Brookings Institution Press.


. . .
Alumni Home
BC Home