In the study, “The Effect of Providing Peer Information on Retirement Savings Decisions,” the authors found that employees who were not participating in their employer’s 401(k) and who were not eligible for automatic enrollment into the plan were less likely to enroll after they were presented with information about the percentage of their coworkers who were already participating. The negative response to coworker savings was concentrated among lower-income employees.
“Our findings suggest that exposing employees who are saving little or nothing for retirement to information about their higher saving coworkers is a discouraging reminder of their low economic status, and that makes them less likely to increase their savings,” says study coauthor James Choi, professor of finance at the Yale School of Management.
To study how coworkers affect retirement savings choices, the authors conducted a field experiment in partnership with a large manufacturing firm and its retirement plan savings administrator. Within the firm, employees fell into two groups: those who were automatically enrolled in the firm’s 401(k) at a contribution rate of 6% of their pay unless they opted out or elected a lower contribution rate, and those who were able to join the plan but who were not eligible for automatic enrollment. In the experiment, non-participating and low-saving employees received a simplified form that allowed them to either enroll in the plan or increase their contribution rate to 6%. The recipients were randomly assigned to receive information about the percentage of their coworkers in the same 5- or 10-year age bracket who were already enrolled, to receive the percentage who were already contributing at least 6% of their pay, or to receive no information about their coworkers’ savings.
Providing coworker savings rates decreased enrollment among non-participating employees who were not eligible for automatic enrollment by nearly four percentage points—6.3% enrolled compared to 9.9% of employees who received no information about their coworkers. Enrollment also decreased as the reported percentage of coworkers who were already saving increased. A one-percentage-point increase in the reported percentage of coworkers already participating in the plan decreased the probability of enrolling by 1.8 percentage points and reduced the average contribution rate change by 0.11% of income.
The negative response from this group of employees was a surprise to the authors, as many studies have shown that providing information about peer behavior causes individuals to conform to that behavior. These employees also should have been the most susceptible to being influenced by peer information, as many were likely not contributing due to inertia and may have weaker convictions about saving. The authors investigated the negative response and found that it was concentrated among employees with low salaries relative to other employees within the local geography, suggesting that the employees were discouraged by social comparisons.
Highlighting peer behavior can provide a helpful nudge toward typical choices when individuals may not know which choices are appropriate for them, but the research reveals an important drawback of highlighting peer behavior, says Choi. “Peer information will always contain an element of social comparison. If individuals with low status are likely to react negatively, then interventions need to be designed in such a way to minimize making their low status salient.”
“The Effect of Providing Peer Information on Retirement Savings Decisions” by John Beshears (Harvard University), James J. Choi (Yale School of Management), David L. Laibson (Harvard University), Brigitte C. Madrian (Harvard University), and Katherine L. Milkman (Wharton School) is forthcoming in the Journal of Finance.