Is Auto-Enrollment ‘Not Optimal’?

A recent industry trade article questions the efficacy of saving early for retirement—and notes that there “may even be such a thing as saving too much.” 

What launches that premise is a research paper titled “Is Automatic Enrollment Consistent with a Life Cycle Model?” That turns out to be a relatively fancy academic title for a simple concept: Does automatic enrollment make sense for younger adults? (The title isn’t that restrictive, but the researchers early on indicate that their focus is young adults.) 

To get to that result, they claim to have crafted “a plausible wage profile for college-educated workers” in which retirement saving does not begin until the late 30s or early 40s, even with standard employer matching. Indeed, they claim that “inducing workers in their mid 20s to participate in a retirement plan requires employer match rates of more than 1000 percent.”

They spend some time examining younger workers reluctance to participate in defined contribution plans—but rather than simply lay that off to lower incomes and a high propensity (and/or need) for spending on things other than retirement at that life stage, the researchers concoct a “life cycle” model that purports to explain this behavior… rationally. Specifically, they cite three factors: (1) a relatively steep earnings profile (a fancy way of saying their income will shoot up in the future), (2) near-zero real interest rates (which they say makes things cheaper now than in the future), and (3) the reality that Social Security’s high replacement rate (at least for low income workers) provides a rational “excuse” for not saving.   

But to produce the mathematical support for their conclusion regarding automatic enrollment, they create a formula which they claim proves that if one’s labor market earnings at age 25 are only 42% of peak earnings at age 45 or 50 (which they state is typical for college graduates), then “not saving for retirement in the early years of one’s career may be completely optimal and rational in a life-cycle model.” Moreover, they state that “if a life cycle model represents optimal behavior, then automatic enrollment that applies to workers of all ages could be nudging young people to make—rather than avoid—a mistake.”

Some of that is based on the trade-offs we all make in living—the choice to save versus paying off credit card debt, for instance—or to buy beyond our (current) means in anticipation of a wealthier future. And no small amount is tied up in the weighting of the value of current consumption compared with the value of wealth in the future. They do countenance the dilutive impact of leakage—which is not only more likely among younger workers (higher turnover, and—with smaller balances, more likely to be spent rather than rolled over), and they do assume a relatively generous (though not uncommon) employer match of 50% on the first 6% deferred. However, they also imbed an extraordinarily conservative 3% real return assumption. 

But at this point it seems rational to parse their language—“if” a life cycle model represents optimal behavior (it may not); and of course there’s their definition of “optimal” (which is purely economic, and based on assumptions with which you might disagree). Moreover, they state that not saving in the early years may be completely optimal and rational—which, of course, leaves open the possibility that it may be less than completely optimal and irrational. None of that is how their language is designed to be interpreted, of course—but it does provide a certain amount of “wiggle room” for what might be considered common sense. 

Indeed, the notion that automatic enrollment might be a “bad” thing is, to put it mildly, counter-intuitive. In fairness, the researchers here conclude not that automatic enrollment is “bad,” or irrational, but rather that it is “not optimal” for workers at the outset of their career to save for retirement—that they would be better off (economically) by postponing that action. 

But to align with that conclusion, it seems to me that you’d have to: (a) accept their model’s assumptions (including, not without controversy, the aforementioned conservative assumption of a 3% real rate of return); (b) believe that one can affix a precise generic economic value to individual decisions (such as having a family) that aren’t, in the real world, always bounded in by economic realities; and (c) believe that when making complex financial decisions and trade-offs, human beings conduct themselves… rationally.

Those of us who work with retirement plans—who know both that many (never) manage to save what seems to be “enough,” and that younger workers are (all too often) disinclined to set money aside for an obscure event (retirement) in the distant future—have long seen automatic enrollment as a marvelous solution to help plug that savings “gap” by getting workers in the practice (if not habit) of saving early, maximizing the time for that “magic of compounding” to do its thing. 

Indeed, it doesn’t take a complicated formula laden with odd Greek-lettered variables to know for a certainty that saving early and consistently greatly increases not only the amount of savings you will be able to accumulate… but the certainty that you’ll have savings sufficient to provide for a future that is anything but certain. 

To argue otherwise is, it seems to me, “not optimal.”

this post originally appeared here.