Over the years, I’ve seen some convoluted ways to rationalize undermining the tax preferences of workplace retirement plans and substituting government tax credits — but a new one just may take the cake.
“A Behavioral Contract Theory Perspective on Retirement Savings,” authored by Ryan Bubb and Patrick Corrigan from the New York University School of Law and Patrick L. Warren from Clemson University’s John E. Walker Department of Economics, starts off by assuming that workers are rational, though perhaps not rational in the way you or I might consider to be rational. However, I’ll accept as logical their assertion that rational workers will prefer saving through an employer-provided plan, rather than accepting a job that does not provide such a plan.
They also claim to provide an analysis that “provides novel explanations for the use of low default contribution rates in automatic enrollment plans, the shift away from defined benefit annuities toward lump sum distributions in defined contribution plans, and the offering of investment options with excessive fees.” Well, it’s certainly novel. More on that in a minute.
The authors claim that employers that provide qualified retirement plans provide more after-tax compensation (in the form of employer contributions not subject to FICA) to employees than employers that do not, and that if workers are rational (there’s that “rational” reference again), then competition for workers in the labor market should therefore provide incentives for employers to offer such plans. So, all other things being equal, workers would prefer to work for an employer that offers a plan than one that doesn’t. So far, so good.
Now we all know that some workers don’t save, and some don’t save enough. But the paper maintains that “if workers undersave (or make other retirement savings mistakes) due to behavioral biases,” then the labor market will give employers incentives to design plans that “cater to and exploit, rather than resolve, those behavioral biases.”
They claim that some workers undersave due to myopia — and that also results in employer plan designs that “exploit the myopic,” and that they do so specifically by — wait for it — offering matching contributions, which the study’s authors claim “naïve myopic workers overvalue.”
Moreover, they claim that this matching results in what they term “cross-subsidization” of rational workers, which, in turn, lowers myopic workers’ total compensation. Said another way, matching contributions mean that the naïve, myopic part of the workforce is not only undersaving (because they overvalue the contributions), but actually receiving less compensation (since the employer is paying them less to offset the cost of the contributions). Are you following this?
And while you may be under the impression that employers offer matching contributions either to increase NHCEs’ contribution rates or to meet one of the safe harbors so that they can provide greater tax-advantaged compensation to HCEs, according to the researchers, you would be wrong. They claim that employers offer those matches precisely because myopic workers overvalue them. Sound like a devious plot to you? Now you’re getting it…
Not only that, they also claim that matching contributions crowd out what they consider to be the superior — and non-redistributive — “commitment device” of non-elective contributions. That’s right, contributions that do not require any particular action on the part of the worker other than to exist and be eligible. Apparently it’s the reward for specific behaviors that is distorting things.
Indeed, these researchers believe that the mere existence of the match means that workers will overestimate how much they will, in fact, save for retirement — and, since they think they will accumulate more than they actually will, they will save less — and employers save money, not only because they pay less (the offset for the matching contributions), the workers will choose to save less — and get less of a match.
Put simply, the paper claims that the labor market gives employers incentives to craft plan designs that cater to what biased workers perceive to be of value, and that the same behavioral biases that produce worker mistakes in saving for retirement will also typically lead workers to prefer plans that fail to correct their mistakes (apparently by encouraging them to save for retirement, but not as much as they need) and that can even exacerbate them. The result, they say, is an equilibrium set of choice architectures (plan design) that fails to effectively address the basic retirement savings policy problem of people not saving enough for retirement. Choice architecture that, in their estimation, is making things worse, not better. Because employers offer workplace retirement plans — and have the temerity to match the contributions of workers who take advantage of these programs. The horror!
The authors’ main message is that when workers undersave due to those behavioral biases, then “employers have incentives to design employer contributions in a way that fails to address the undersaving problem and indeed that can even actively exploit the undersavers.”
They not only claim that matching contributions are harmful to rational workers, but maintain that they are “unambiguously so.” How? Well, they claim that matching contributions distort the worker’s incentive to save by encouraging too much savings (which is apparently different than enough savings), and the worker would receive more “utility” from being paid this money as wages.
Oh, and as another example of employer bad behavior in plan design (and ostensibly another reason employers can’t be relied upon in this role), the authors note that most employers choose an automatic enrollment default deferral rate of 3%. The authors concede that the use of automatic enrollment can result in retirement savings by employees who would otherwise have initially contributed nothing to the plan. However, they caveat this admission by noting that some who would have saved more than that default rate if they had had to fill out an enrollment form only save at the default rate.
The solution for this? Substituting for the current system of employer-provided pension plans a new federal defined contribution plan, designed by a federal agency and not linked to any particular job, of course. Oh, and not only do they claim that this could improve savings outcomes, but that it would do so “at little to no fiscal cost to the government.”
Except, presumably, for the part where the government trades the temporary deferral of tax revenues for the permanent forgiveness of a government tax subsidy.
FWIW, I find it hard to believe that this kind of analysis would survive serious scrutiny outside of the rarefied air of academia. But stranger notions have.
- Nevin E. Adams, JD