Shlomo Benartzi

May 27, 2013

Piece originally published in Pensions and Investments.

Back in 1981, when employers started to adopt 401(k) programs, another exciting product full of promise debuted: the DeLorean sports car. With those gull-wing doors and brushed stainless-steel body, the DeLorean was first-rate automotive eye candy. But it was seriously underpowered, not to mention tough to get into - much like 401(k)s of the day.

DeLorean Motor Co. filed for bankruptcy the next year, but 401(k)s continue to this day and, much as with DeLoreans, they remain frozen in time. Yes, the model 401(k) offers some nice options not available originally. Automatic enrollment, for example, dramatically increases the number of plan participants, and more than half of large plans offer it. But 401(k) balances remain too low to provide a big boost to retirement income. Financial Engines Inc.'s survey last year of 401(k) plans found almost three-quarters of Americans are not on track to reach retirement income goals by the time they reach 65. And, only 14% of Americans say they are very confident they'll have enough money to live comfortably in retirement, according to a recent survey by the Employee Benefit Research Institute.

And that's just the savings side. When it comes to how those savings will be converted into retirement income, few 401(k) plans have good options. On this count, they're a metaphoric Model T.

So 401(k) design remains marooned in the past. Meanwhile, in the present, these savings plans have become crucial to retirement security as pension plans evaporate and more workers must rely on their own savings for a decent retirement.

The reason these plans haven't evolved is due to a more modern chicken-and-egg conundrum. Companies are understandably chicken when it comes to innovating because new ideas risk running afoul of state and federal regulations. And the government doesn't want bad eggs laid in plan design, so it won't approve new twists until they've been demonstrated grade A.

When innovation happens, it's painfully slow. For example, savings escalator programs were ready for testing in the mid-1990s. But it took two years before a brave company would try such a plan, and three years after that until a second company stepped up. The dramatic, positive results finally got a government recommendation in the Pension Protection Act of 2006, and now more than 50% of large plans have such programs, and they increase the savings of millions of Americans. However, it took more than 15 years for the change to happen.

What's needed is an incubator where plan design ideas can be safely hatched and tested in workplaces more quickly. A little bit of such testing can go a long way in many domains. In the U.K., Richard H. Thaler, professor of behavioral science and economics of the University of Chicago, is an adviser to the Behavioral Insights Team, a small group formed by Britain's Prime Minister David Cameron. The team tests ideas to improve everything from tax collection to increasing the number of homes with better-insulated attics. One success story: When delinquent taxpayers were told most everyone else paid on time, many felt pressured by that social norm to pay back taxes. When an idea is proven effective on a small scale, it's rolled out nationally.

We need such smart innovation in the 401(k) arena, and we need it now. We propose setting up a "safe haven for innovation." It can work this way: Academics, businesses and/or government officials devise inventive ways to improve 401(k) plans; the government makes sure they're good policy; and university human-subject committees review them to ensure subjects are treated properly, standard procedure for academic research. The best ideas then get blessed by the government, and a few employers are invited to try the innovations. The government can provide legal protection to those employers, or at least recognition or awards for their role as innovators.

In return, employers will share their testing results with academics to scientifically measure, for example, what motivates lower-income employees to save more. We envision the innovations to be along the lines of nudges and tweaks, with little or no downside risk. But as we've seen, a little nudge, such as automatic enrollment, can turbo-charge retirement savings.

The federal government is on board. J. Mark Iwry, senior adviser to Jacob J. Lew, secretary of the Treasury, said, "Policymakers need to step up our efforts to identify and test promising new ideas. We can help speed the pace of innovation by encouraging research and working more closely with academia."

Testing ways to increase savings is one priority, but another big problem begs for solutions: leakage. More than $80 billion leaves 401(k) plans prematurely every year, and $74 billion of that evaporates when people change jobs, according to a 2009 study by the Government Accountability Office. (The rest is lost mainly to hardship withdrawals for housing, health and other emergencies.) The temptation to cash and spend that fat 401(k) check, as opposed to rolling it over into another retirement plan, is just too strong for many.

Fixing leakage is arguably both a top priority and low-hanging fruit. Think of the $4 trillion of retirement savings as sitting in a leaky bucket. Plugging leaks in that bucket could be a relatively quick and easy way to increase aggregate retirement savings - over the next decade, we could add another $1 trillion just by using behavioral finance techniques to solve that one problem.

Given the money that leaks out of the bucket exceeds the employer matching money poured into the bucket (about $60 billion annually), employers certainly have motivation to plug the leaks. And employees, with a little help, would understand that rolling over retirement savings vs. cashing it out makes a huge difference in retirement security.

Simple solutions should be tested. But currently few plans attempt to even educate employees about the harmful effects cashing out has to their retirement security; and few plans make rollovers more likely through counseling.

Behavioral economists have shown they have effective tools to dramatically help people trade short-term gratification for long-term gains. However, under the current system, plan sponsors are reluctant to experiment because of the daunting complexity of rules and regulations.

Leakage is just one example of how our current system pushes companies into a safe haven of compliance instead of toward a safe haven for research. Three decades after the first 401(k)s, we need more innovation to build a retirement savings vehicle as efficient as a Prius, as powerful as a Ferrari and as safe and easy to pilot as the Google driverless car.

Shlomo Benartzi is professor and co-chair of the Behavioral Decision-making Group at the Anderson School of Management, University of California, Los Angeles. John Payne is at Duke University, Durham, N.C., where he is the Joseph J. Ruvane Jr. Professor of Business Administration at the Fuqua School of Business, as well as professor of psychology, professor of law, and research professor in statistics and decision sciences. Mr. Benartzi is chief-behavioral economics and Mr. Payne, academic adviser, to the Allianz Global Investors Center for Behavioral Finance, which seeks to turn academic ideas into practical tools for plan sponsors.

Media Relations