Yale Study: A Few Extra Words Can Make a Major Difference for 401k Investors

As shown in last week’s article, (“New Study Reveals Three 401k Strategies More Important than Asset Allocation,” August 14, 2012), one of the biggest factors in achieving retirement comfort is to invest early and often in a retirement plan. This suggests 401k plan sponsors should adopt policies and procedures which encourage such behavior on the part of their employees. A Yale study released earlier this year suggests nudging employees towards such self-serving actions might be as simple as adding a few words to their quarterly report.

In their National Bureau of Economic Research working paper “Small Cues Change Savings Choices,” (James J. Choi, Emily Haisley, Jennifer Kurkoski, Cade Massey, NBER Working Paper No. 17843, February 2012), the authors conclude adding just one or two sentences to employee 401k emails can increase their savings rate as high as nearly 3%. Compound that additional savings over the working life of the typical employee, and that can have a significant impact on the employee’s total 401k assets at the time of retirement.

The roots of this new study can be found in earlier research by Choi et al (“Defined Contribution Pensions: Plan Rules, Participant Decisions, and the Path of Least Resistance,” In James Poterba, ed. Tax Policy and the Economy 16, pp. 67-114, 2002) and Benartzi and Thaler (“Heuristics and biases in retirement savings behavior,” Journal of Economic Perspectives 21, pp. 81-104, 2007). Both of these predecessor studies point out people tend to simplify their 401k decisions. For example, they tie their contribution amounts to specific amounts related to the plan’s stated maximum possible contribution or the minimum necessary to attain the maximum company match.

This general decision heuristic falls under the behavioral trait of “anchoring.” Since anchoring works with any arbitrary number, the researchers sought to determine if choosing a specific number might change savings patterns. James Choi, Associate Professor of Finance at the Yale School of Management in New Haven, Connecticut, told FiduciaryNews.com, “A Yale alum at the company where we ran the experiments approached us because he was interested in finding ways to raise 401k contributions among the company’s employees. I’ve had a long-standing interest in how 401k savings outcomes are influenced by psychological levers, so I saw this as a great opportunity to study the effect of some extremely subtle nudges.”

The paper presented findings consistent with earlier research showing small subtle changes can influence individual savings rates. “I was surprised by how large and persistent the cues’ effects were,” says Choi. “A one or two sentence cue added to an email affected subsequent contribution rates much more than an expensive match would, and for as long as a year after the email.”

The results also mimic the conclusions of behavioral economics research in areas outside of retirement plan savings. This other research shows how anchoring can influence decision making with regards to everything from corporate merger activity to bidding strategies at art auctions. The same theory that drives these behaviors can show the way to help 401k investors achieve a better retirement.

The researchers tested three types of cues using a range of actual numbers. In all cases, these cues followed the statement “To take greater advantage of [Company]’s 20120 match, increase your contribution rate soon before the year is over.” Here are the actual words and numerical ranges used:

Anchor Cue Type:

“For example, you could increase your contribution rate by x% of your income and get more of the match money for which you’re eligible. (x% is just an example, and shouldn’t be interpreted as advice on what the right contribution increase is for you.)”

The x anchors tested were 1%, 3%, 10% and 20%

Savings Goal Cue Type:

The researchers tested these two sentences, the first representing a $7,000 savings goal and the second representing an $11,000 savings goal:

“For example, suppose you set a goal to contribute $7,000 for the year and you attained it. You would earn $500 more in matching money this year than you’re currently on pace for.”

“For example, suppose you set a goal to contribute $11,000 for the year and you attained it. You would earn $2,500 more in matching money this year than you’re currently on pace for.”

Savings Threshold Cue Type:

The researchers tested these three sentences, two testing a dollar threshold (one at $3,000 and the other at $16,500) and one testing a percentage threshold. The dollar threshold was written this way:

“The next $(x-y) of contributions you make between now and December 31 will be matched at a 100% rate.”

In the above x represents one of the savings goals (either $3,000 or $16,500) and y represents the recipient’s year-to-date match-eligible contributions.

The percentage threshold was written this way:

“You can contribute up to 60% of your income in any pay period.”

The Results of the Experiment:

The study did contain a control group for comparison purposes. It concludes, relative to the Anchor Cue Type test, “the lower two anchors we tested (1% and 3%) depress average contribution rates, but the higher two anchors (10% and 20%) do no increase average contribution rates relative to the no-anchor control group.” IN the Savings Goal Cue Type test, “the $11,000 goal group never contributes less than the control or $7,000 goal groups.” Researchers concluded the 60% Savings Threshold Cue Type test “has a larger effect on people contributing little at the time the email was sent.” This pertains to savings rates, not dollar amounts. There was weaker evidence to suggest lower dollar savers were similarly motivated by the Dollar Savings Threshold Cue Type test.

In the end, the paper concludes “small numerical cues can influence decisions as economically significant and familiar as retirement savings plan contributions. Low cue decrease contribution rates by up to 1.5% of income, and high cues increase contributions by up to 2.9% of income.” Moreover, the researchers write, “unintentional cues buried in mundane communications can also affect behavior. Thus, organizations and policymakers should take responsibility for the cues they disseminate and wield them mindfully.”

Said another way – and this is why the conclusion is of vital importance to 401k plan sponsors – even an innocent aside written with seemingly appropriate disclaimers can unduly influence 401k investors. And increase plan sponsor fiduciary liability.

Thus, Choi warns, “Plan sponsors should be mindful of the savings cues they include in their communications to participants. If they want to encourage high contribution rates, then their communications should only contain high contribution rate cues, not low contribution rate cues.”

But there may be an easier and less risky way to increase employee savings rate. We’ll leave you with this added tidbit from Choi: “One of our findings is that mentioning the very high match threshold at this company raised contribution rates. This is evidence that participants use the match threshold as a cue for their own contribution rate. Therefore, implementing a high match threshold, even if it means that the matching rate has to fall, will raise contribution rates.”