Art by J.CardielloThere are widespread reports that House Republicans—in an effort to raise revenues for unrelated purposes—are considering converting the current 401(k) system to Roth-only contributions: requiring that all participant contributions be made on an after tax basis. Those contributions would then—under “Roth treatment”—accumulate and be paid out tax-free (subject, one assumes, to some sort of holding period requirement). Many are reacting (very) negatively to this idea, arguing that it will result in a significant reduction in retirement savings and that, particularly, an all-Roth system will discourage participants from making 401(k) contributions.
I’m finding the discussion of this a confusing mixture of cynicism and misunderstanding—or at least a lack-of-being-clear—all round. So I’d like to get some things straight about Roth versus non-Roth contributions and then discuss what seem to me to be the obvious problems with it.
Retirement savings taxation math
So, first, I want to start with the math. Under the current “non-Roth” system, a participant can make a “pre-tax” contribution—say, the current $18,000 maximum—to a plan, leave it there for 15 years (or one year or 40 years). When this non-Roth saver takes her money out (contributions and earnings), she will pay taxes (at her then income tax rate) on the entire distribution. (That, by the way, will be the first time she has paid any taxes on this money.)
Under the Roth system, an identical participant can take an identical $18,000, pay taxes on it first (at her income tax rate in the year of contribution), then put the (after-tax) balance into the plan. When this Roth saver takes her money out after 15 years, she will take it (contributions and earnings) tax-free.
Can we all agree that—assuming that the tax rate for these savers is same at the time of distribution as it was at the time of contribution, and assuming that they both left the money in for the same period and earnings were identical—these two savers, at distribution, will both have the same amount of money, after taxes. That is, it doesn’t matter if you pay all your taxes at distribution, or you pay all of them at contribution. Your benefit is the same: the non-taxation of trust earnings.
Why do participants dislike Roth?
Now, I have some significant problems with “Roth only,” which I’m going to get to. But something I find a little puzzling (and in some respects interesting) is that many are saying (and with some evidence) that participants (and particularly low-paid participants) don’t like Roth contributions. And, I thought we just agreed that it doesn’t affect participants.
Here are my thoughts about why that might be—that participants don’t like Roth contributions:
(1) Plans calculate contributions and matching contributions on the basis of gross (that is, pre-tax) pay. So that whole thing about starting with “the identical” $18,000, paying taxes on it, and contributing the balance doesn’t work. OK. But, isn’t there a way to fix that?
(2) Participants don’t understand the math. Maybe. But, again, can’t the math of this be explained?
(3) Most participants expect to be in a lower tax bracket when they take their money out, so that that assumption we made (that tax rates are constant) doesn’t hold. I actually think this may in fact be the case. I would certainly like to see more research on it. If all the talk we hear about how participants are not saving enough is the case, then it seems to me that such a participant expectation may be pretty rational: they will be paying taxes at a lower rate in retirement, because they will have less income in retirement.
Whatever the case, we are going to need to get some clarity on: participant attitudes toward Roth contributions; whether (if they are negative) they can be changed; and whether (if they are negative and can’t be changed) the adverse effect on retirement savings of a Roth-only regime is acceptable.
Those are my thoughts—at this point (and before we’ve seen any actual proposal)—on the effect of Roth-only on participants. NEXT: Three other problems