It’s long been noted that inertia is a powerful force regarding behavioral finance and automatic enrollment — but it may have limits, according to a new study.
Coverage of the report — titled “Smaller than We Thought? The Effect of Automatic Savings Policies” — focused on how job change undermines retirement savings — both because of vesting, as well as the effectiveness of automatic enrollment, and more specifically auto-escalation, since those mechanisms tend to reset with the change in employers (and payroll).
Don’t get me wrong. The report states quite clearly that these automatic mechanisms provide a positive result — the authors comment only that it’s perhaps not quite as positive as most think. Their solution — give people less access to these monies before retirement, and require savings, rather than permitting an opt-out.From a pure mathematical stance, there’s little argument there — making people save and prohibiting pre-retirement access to those funds certainly benefits retirement savings, though it also exacts a financial toll in the here and now.
There’s little to be done about job change — which, as I’ve noted before, isn’t really all that different today than it was several decades back. And it should come as no surprise that folks that have been accustomed to automatically being defaulted[i] into saving at employer #1 will readily come to rely on that convenience at employer No. 2, if the option is available, and even if it resets their rate of savings.
Indeed, we’ve long embraced a working assumption that automatic enrollment takes participation rates of 65%-70% and turns them into 90%. Said another way, only about 1 in 10 take the time/energy to opt-out of automatic enrollment.
Opt-Out Observations
That said, what caught my eye here was what turns out to be an extraordinarily high rate of opt-out when it comes to automatic escalation. Among the plans/participants studied[ii], more than half - 57% - opted OUT of automatic escalation the very first time it came up. And it gets worse as time progresses; while on average, the acceptance rate of the auto escalation default is 43% on the first escalation date[iii], it slips to 36% on the second date, and 29% on the third date.
Admittedly that’s higher than previous research suggests — in fact, the researchers acknowledge that in the Vanguard “universe”, the acceptance rate of an auto-escalation default is 63%, 63%, and 60% after one, two and three years of tenure — though even that was significantly below the 85% found by in a 2013 study by Benartzi, Peleg, and Thaler.
Those differences can be attributed to different employers, different employee populations, even different periods of time during which the assessments are made. Just as significantly, we don’t know anything about their financial situations, the rate of deferral they were defaulted in at, or why they opted out.
That said, the opt-out rates struck me as higher than most might be led to expect — suggesting that while inertia can be a powerful force, it’s not without its limits.
- Nevin E. Adams, JD
[i] The 67th Annual Survey of Profit-Sharing and 401(k) Plans by the Plan Sponsor Council of America found that 64% of surveyed plans use an automatic enrollment feature (74.3% among larger plans). That same survey found that more than three-quarters of the plans that use automatic enrollment also employ contribution acceleration.
[ii] The study focuses on nine firms that, sometime between 2005 and 2011, introduced either (1) automatic enrollment on its own, (2) default auto-escalation in a 401(k) plan where automatic enrollment was already present, or (3) automatic enrollment and default auto escalation simultaneously. The automatic policies applied only to employees hired from a certain date onward, so they identified their effect by comparing 62,430 employees hired in the year after the policy introductions to 55,937 employees hired in the year before the policy introductions.
[iii] Opt-out rates that, as it turns out, aren’t very different from that reported from most of the state-run IRAs.