Automatically enrolling workers in employer-sponsored retirement savings plans—a practice that dramatically increases participation rates, particularly among younger workers—has drawn support from across the political spectrum. Over the past two decades, Congress has taken steps to incentivize automatic enrollment, and President Joe Biden has proposed further expanding its scope.
Bolstering automatic enrollment may seem like a no-brainer—how could increasing retirement saving possibly be bad? Unfortunately, putting money into a retirement plan may be the exact wrong choice for many young, college-educated workers who anticipate higher earnings in the future. More generally, one-size-fits-all policies that incentivize workers—regardless of age or circumstances—to save the same fraction of their earnings for retirement should be reconsidered.
Automatic enrollment is sometimes referred to as a “nudge”—a gentle form of paternalism. It’s gentle because someone who has made a thoughtful, rational decision not to save for retirement can easily opt out. However, nudges alter the behavior of those who aren’t perfectly rational, have trouble with complex decisions like retirement planning, and fall victim to inertia—in other words, most of us.
But automatic enrollment can have unintended consequences. In a recent study, we explored how a rational economic actor’s retirement saving choices compare with the choices that automatic enrollment incentivizes. We used a standard economic model in which the goal is to keep consumption spending relatively steady—avoiding large changes in standard of living—despite fluctuations or trends in their earnings. A key result is that for workers who expect higher labor earnings in the future—such as college graduates—saving for retirement during their 20s does not make sense. To keep consumption spending as steady as possible, it makes more sense to save nothing for retirement now and ramp up saving later, when earnings are higher.
In particular, we find that a college-educated worker with a typical earnings pattern would not start saving for retirement until their late 30s or early 40s. Standard levels of employer matching are not enough to change this behavior; in fact, it takes match rates above 1,000 percent to induce rational workers in our model to start saving in their 20s. Today’s persistent low interest rates—which make spending now more attractive relative to saving for the future—only strengthen these results.
There are caveats to our analysis, of course. Young workers who anticipate switching jobs—giving them an opportunity to cash out of their 401(k)—may find it worthwhile to participate and take advantage of the match. Our model also does not rule out saving for other goals, such as buying a home, taking a vacation, or insuring against future fluctuations in labor earnings. Finally, starting to save for retirement early may help young people make a habit of setting money aside.
But our findings raise serious questions—and should prompt deeper discussion—about the design of automatic enrollment. If nudges are supposed to bring behavior in line with that of a rational economic actor who is forward-looking and carefully weighs costs and benefits—and if standard economic models represent that behavior—then automatic enrollment seems to nudge younger workers in the wrong direction. Other researchers have similarly suggested that automatic enrollment may not be in the interest of workers with large amounts of high-interest rate credit card debt, who would benefit from paying off that debt before beginning to save for retirement. Of course, any workers who don’t wish to save can opt out of automatic enrollment—but automatic enrollment is premised on the fact that many workers find it difficult to overcome inertia and deviate from whatever option they’re defaulted into.
Automatic enrollment has become widely accepted among economists and policy makers, who generally feel that helping less-than-perfectly rational individuals save optimally for retirement is desirable. And we agree with that underlying goal. However, our work suggests that further discussion is necessary about the one-size-fits-all design of automatic enrollment programs, which uniformly encourage retirement plan participation at all ages and in all circumstances.
Jason Scott is the managing director of JS Consulting, John B. Shoven is the Charles R. Schwab Professor Emeritus of Economics at Stanford University, Sita Nataraj Slavov is a professor of public policy at George Mason University’s Schar School of Policy and Government and a visiting scholar at the American Enterprise Institute, and John G. Watson is a lecturer in finance at Stanford University’s Graduate School of Business.