When we have leftover holiday cookies at home, I’m tempted to eat those little sugary treats. Rather than have them sit around calling my name, I drop them off in the coffee room at work. They vanish within hours, and I avoid the guilt (and pounds!) that come with giving in to temptation.

A similar insight set Richard H. Thaler down the path to winning the Nobel Prize in Economic Sciences in 2017, which was presented in Stockholm on December 10.

When Thaler invited some fellow economists over for dinner in the 1970s, they proceeded to fill up on cashews. Before they could ruin their appetites, he took the bowl of nuts back into the kitchen. His friends thanked him for removing the temptation.

“The idea that it could make you better off to reduce your choices—that idea was alien to economics,” Thaler observes in The Undoing Project, an outstanding history of the birth of behavioral economics by Michael Lewis.¹

According to traditional economic theory, we’re rational people who make optimal decisions. Our rational side knows we shouldn’t feast on cookies, candy, or cashews. But as anyone who steps on a scale after the holidays can attest, the rational self often isn’t in charge of our daily lives.

Help with doing the right thing

A related mistake people frequently make is overvaluing the present and discounting the future. Left to their own devices, for instance, young employees often waited five to ten years to join their retirement plan. Today’s needs seem so insistent, so vital—and retirement feels as remote as Mars.

Thaler and a former student, Shlomo Benartzi, launched an experiment to overcome this bias. What if you asked employees to join their retirement plan not today, but next year? And what if they were asked to increase their savings not now, but in the future, when they expect to receive a raise?

When Thaler and Benartzi tested their plan, which they called Save More Tomorrow (SMarT), at a factory in the late 1990s, the results were tremendous. The 50% of employees who chose to sign up for automatic annual increases saw their average savings rate rocket from 3.5% to 11.5% in a little more than two years.²

Then it went nationwide

This breakthrough didn’t go unnoticed, and over the next decade, the 401(k) plan was redesigned to nudge us into making better decisions. Now the majority of employees in plans administered by Vanguard have automatic enrollment, with contributions defaulted into target-date investments. Most also have automatic savings increases, usually by one percentage point a year. My colleague, Steve Utkus, wrote about how Thaler and Benartzi’s work impacted Vanguard and retirement savers in a November blog.

The rise of autopilot plans has led to clear success in two areas, but there’s still work to be done in a third.

First the successes: Plans that automatically enroll eligible employees have a 90% participation rate, compared with 63% in plans where enrollment is voluntary. And 90% of plan participants now hold balanced portfolios of stocks and bonds. Ten years ago, three times as many participants held an extreme portfolio of either no stocks or all stocks.³

Now the area that needs work: Participants have remained stuck in one area—their savings rate. The average savings rate was 6.2% in 2016, down from 6.9% in 2015. The drop is attributable to the increased adoption of automatic enrollment plans. Many of these plans set an initial savings rate of 3%, which tends to draw down the average savings rate among participants.

Time for courageous plan design

Vanguard research suggests that most plan participants wouldn’t object to a much higher initial savings rate. Between 2014 and 2016, the opt-out rate among plans that set their initial savings rate at 6% was identical to the opt-out rate among plans that set the initial savings rate at 3%. In both cases, 7% of those automatically enrolled elected to opt out of the plan.

In an article published in The Wall Street Journal in December, Thaler and Benartzi recommend setting more aggressive targets in today’s autoenroll plans. They wrote there would be “little blowback and a significant boost to worker savings” if the initial savings rate were set at 7% or more. They also recommend setting automatic savings increases at two percentage points annually, rather than the typical setting of one.4 This approach ties back to my move-the-cookies method—basically, make it easier to do right.

A 6% initial savings rate followed by annual increases of two percentage points would lift a plan participant into Vanguard’s suggested 12% to 15% retirement savings range by the beginning of their fifth year of employment, even without any company contributions. By contrast, an initial savings rate of 3% followed by one percentage-point increases wouldn’t get a participant into range even after seven years, the median tenure of a worker in a Vanguard plan in 2015.

Source: Vanguard, 2017.

Thaler and Benartzi say the lesson is clear: Plan design does work. But if the starting savings rate or annual increase is too low, plans could be unwittingly encouraging participants to save too little or cause them to get to an ideal savings rate years later than they might otherwise.

¹ Michael Lewis, The Undoing Project: A Friendship That Changed Our Minds, 2016.

Lewis previously wrote the bestsellers Moneyball, The Blind Side, and The Big Short, all of which were made into successful Hollywood movies.

² Richard H. Thaler and Shlomo Benartzi, “Save More Tomorrow: Using Behavioral Economics to Increase Employee Saving,” Journal of Political Economy, February 2004.

³ How America Saves 2017, Vanguard 2016 defined contribution plan data.

Shlomo Benartzi and Richard H. Thaler, “The Secret to Getting Workers to Save More for Retirement,” The Wall Street Journal, December 11, 2017.