How Design of Pension Plans Influences Employee InvestmentsPublished: January 30, 2002 in Knowledge@Wharton
With the collapse of Enron, lawmakers, regulators, financial advisors and working people are taking a hard look at 401(k) plans, which employees in many U.S. companies use to save for retirement. Enron employees are said to have lost some $2 billion in retirement accounts packed with Enron stock, and many wonder whether the participants of other retirement funds are at similar risk. In the U.S. about $2 trillion is held in some 325,000 401(k)s owned by more than 40 million participants.
Among the issues raised: Should companies be prohibited from requiring employees to hold employer stock in their plans, as Enron did? To what extent should employers be accountable if they offer investment options that do poorly? Is it appropriate for employers to attempt to influence employees’ investment choices?
Many corporate executives believe that once a plan is set up, employees should be left on their own.
But it’s not that simple, says Wharton finance professor Andrew Metrick. Even if the employer wants to sit on the sidelines, his choices in designing the plan will strongly influence the level of employee participation, the percentage of income employees contribute and the investment choices they make.
Employees, said Metrick, typically take "the path of least resistance" in managing their 401(k) accounts. Employee choices, then, are largely determined by the plan’s "defaults" – the options automatically chosen if the employee does not take an active role in decision-making. "For better or for worse, plan administrators can manipulate the path of least resistance to powerfully influence the savings and investment choices of their employees," Metrick and three co-authors conclude in a recent study of 401(k) plans entitled Defined Contribution Pensions: Plan Rules, Participant Choices, and the Path of Least Resistance.
The co-authors are James J. Choi and David Laibson of the Department of Economics at Harvard, and Brigitte C. Madrian of the Graduate School of Business at the University of Chicago.
The paper will be one of several discussed at an upcoming March 22, 2002, Conference on Household Portfolio-Choice and Financial Decision-Making. The conference is sponsored by The Rodney L. White Center for Financial Research.
The research done by Metrick and his co-authors looked at the plans of seven large companies employing nearly 200,000 people in the food, office equipment, insurance, utility and consumer goods industries. Each of the companies, none of which was identified, had made a significant change in its plan, allowing a comparison of employee behavior before and after. The study used data from the plans as well as surveys completed by employees.
In one part of the study, the researchers found that two-thirds of the employees surveyed believed that they were not contributing enough to their plans, and that a third of this group intended to raise savings rates over the next two months. Plan records showed, however, that almost none of them did, even though their savings rates were indeed low.
"This finding introduces a theme that we return to throughout the paper," the authors wrote. "Specifically, at any point in time employees are likely to do whatever requires the least current effort: Employees often follow the ‘path of least resistance.’ Almost always, the easiest thing to do is nothing whatsoever, a phenomenon that we call ‘passive decision.’"
While the typical 401(k) requires an employee to sign up in order to participate, a growing number of companies enroll employees automatically. Employees can opt out, but, to do so, must make a decision and act. Changing the "default" to automatic participation at three of the companies studied caused a sizable increase in enrollment, even though the employees had the right not to participate that they had before the change.
Before the automatic enrollment began, employees who had worked for six months at the three companies participated in the 401(k) plans at rates of 26% to 43%. But among employees hired after enrollment became automatic, participation after six month’s tenure ranged from 86 to 96 – even though these employees could have withdrawn.
Even with more time, employees who had been automatically enrolled continued to participate at substantially higher levels than those who had been employed the same amount of time but had not been automatically enrolled. At 36 months tenure, the automatic enrollees participated at rates of 31 to 34 percentage points higher than those of employees who were not automatically enrolled.
Automatic enrollment improved participation of all demographic groups but had the greatest effect on those who traditionally participate at the lowest levels – younger employees, lower-paid employees, African Americans and Hispanics.
Similarly, the research found that employees who are automatically enrolled tend to adopt the default contribution rates and investment choices made by the plans. In practice, that meant the automatic enrollees tended to contribute less, and to invest more conservatively, than the participants who had made active decisions to participate. This was because the default choices involved such small contributions and such conservative investments. At those levels, the employees probably would not save enough for retirement.
"Employers who seek to facilitate the retirement savings of their employees need to respond to the tendency of employees to ‘stick with the default,’’ the researchers said. "Employers should choose defaults that foster successful retirement saving when the defaults are passively accepted in their entirety."
One of the plans studied attempted to capitalize on employees’ passivity by asking them to commit in advance to automatically increase their contributions as they received future raises. Using an idea developed by Richard Thaler of the University of Chicago and Shlomo Benartzi of UCLA, the "Save More Tomorrow" plan "alleviates problems of self-control and procrastination," the authors wrote. Employees find it easier to fund contribution increases out of raises because they never face a cut in take-home pay.
Of the 207 employees who were invited to join this "SMarT" plan – all people who had been reluctant to raise their savings levels - 162 enrolled and 129 stayed with the plan through three pay raises. During that time, their contribution rates rose from an average of 3.5% of income to 11.6% – well above the typical 401(k) contribution level of 7%.
Meanwhile, a group of 79 employees who were not invited to join the plan - because they had already indicated an eagerness to save more – increased their average contribution rate to a less impressive 8.7%.
"Since it is reasonable to assume that this latter group of workers represents a more highly motivated group of savers than the SMarT plan participants, the increases by the SMarT plan participants are very striking," the authors concluded.
At two other companies, the researchers looked at the effect of changes in employers’ matching contributions. At the first, the company match of 50 cents on the dollar was made against the first 5% of income contributed by the employees and the first 6% of income contributed by management personnel. That was then changed to a match on the first 7% for employees and 8% for management. The match on the additional 2% of income was made in the form of the company’s stock, while the initial match continued to be in cash to be invested at the employee’s discretion.
While the change did not cause more people to participate in the plan, it did cause participants to increase contributions in order to benefit from the larger employer match. There was little change in behavior, however, among participants who had already been contributing above the matching threshold – a group that would not benefit from the raised thresholds.
The second company went from no match to a match of 25% on contributions up to 4% of income. The researchers concluded that introducing this relatively modest match had increased participation in the plan by more than 40% and also increased contribution levels, though they caution the data is not definitive.
"In sum, our limited evidence suggests that employer matching does have a significant impact on both 401(k) participation and contribution rates," they write.
As policymakers debate whether to change 401(k) rules in the wake of the Enron collapse, they should realize that the people who design plans are largely responsible for employee’s investment results, Metrick said. "It’s almost a breach of fiduciary responsibility not to recognize that."
Employer stock should not be a default investment, since good investment strategy requires diversification, Metrick said. At Enron, employer matches were made in Enron stock, and employees could put their own contributions in to the stock as well. As a result, more than 60% of employee’s 401(k) assets were in Enron shares.
"A very simple way to avoid a debacle like Enron, and one that I would be in favor or, is to limit the amount someone can hold in company stock," he said, noting that employees who are eager to own their company’s stock can buy it as an ordinary investment.
Does the research point the way to an ideal 401(k) plan?
That depends, said Metrick, on the goals of the employer and participants. The research does suggest that in practice there is no such thing as a passive plan; any plan, by its design, will encourage certain behavior. "The goal," he said, "would be to get people to go beyond the passive approach of using defaults." That, he said, would probably require setting up a system for giving employees individualized guidance in setting a long-term financial strategy.
In addition to Metrick’s paper, the conference on March 22 will discuss other papers that address the following topics: How diversified are retail investors’ brokerage accounts, and what is the implication for the volatility of their portfolios? What type of mortgage should different people choose – fixed or variable rate – depending on the various risks they face (interest rate risk, inflation risk, income risk, etc.)? Finally, how sensitive has spending on luxury goods been to stock returns (the so-called wealth effect), and can this sensitivity explain the historically large return to stocks?
Experts from both the academic and business worlds will discuss each paper. There will also be a panel session examining the role of company stock in retirement accounts, a "particularly salient topic in light of recent events at Lucent and Enron," according to Wharton finance professor Nicholas Souleles, a conference organizer.