Roth 401(k)s: Appealing but Impractical

Tax-deferred retirement accounts help the financial industry but cost taxpayers an added $20.7 billion a year, according to two economists. Would an all-Roth world be more socially beneficial?

The lives of retirees would probably be simpler if Roth IRAs and Roth 401(k)s replaced tax-deferred traditional IRAs and 401(k)s. Minimum annual distributions (“RMDs”) after age 72 (under the new tax rules) wouldn’t be necessary, and retirees would be relieved of the tax bills that the annual distributions bring in their wake.

Nor would anyone feel compelled to convert a traditional IRA to a Roth IRA—a labor-intensive process often recommended to high net worth retirees as a way to minimize taxes in retirement.

Economists Stephen Zeldes of Columbia University and Mattia Landoni of Southern Methodist University suggest another reason for switching to an all-Roth regime. (Read their paper here or here.)

These two economists observe that tax deferral has caused the accumulation of $3 trillion of additional assets (the present value of taxes on future RMDs) into 401(k)s and traditional IRAs, including rollover IRAs. The money is there, in effect, because the government hasn’t taken taxes out of it yet. In the two economists’ opinion, the government is paying too much in fees on that money.

Zeldes and Landoni calculate that if Roth 401(k)s and Roth IRAs replaced the status quo, or if the government paid institutional-level fees on its investment in 401(k)s and IRAs, the government would save enough money to give every saver a 6% match on their contributions.

“We estimate that tax deferral increases demand for asset management services by $3 trillion, causing the government to pay $20.7 billion [per year] in corresponding annual fees. Tax deferral in our model produces a larger asset management industry, higher taxes, and lower social welfare,” Zeldes and Landoni write.

I agree that tax deferral creates complexity, and that the savings from federally-enhanced economies of scale should benefit individual investors, not intermediaries. But the retirement industry has consistently fought to preserve the status quo, for the same reason the two economists want to change it.

The benefits of the current system to the retirement industry are huge, if difficult to calculate precisely because of feedback effects. To get a handle on it, I look at the estimated “tax expenditure” for individual and group retirement—the taxes the government doesn’t collect on contributions or gains—and assume that it translates into extra assets under management and fee revenue that the financial industry would not have received in the absence of tax deferral.

For the five year period, inclusive of years 2019 through 2023, the Joint Committee on Taxation estimates that the tax expenditure for defined contribution plans will be $775.6 billion, while the tax expenditure for traditional IRAs will be $100.9 billion and for Roth IRAs $44.5 billion. The Center for Retirement Research at Boston College recently calculated that fees reduce 401(k) account balances by about 10%. Fees on rollover IRAs at brokerage firms may be higher than fees in 401(k) plans.

(In their model, Landoni and Zeldes assume that the foregone taxes for Roth IRAs and traditional IRAs end up the same in the long run. That’s why they prefer to focus on the impact of fees, where a switch to Roth accounts would be demonstrably cheaper for the government.)

Ironically, it’s their lack of RMDs, which are desirable for the owner, that disqualifies Roth accounts as an viable alternative to traditional accounts under current law. Lack of RMDs makes Roth accounts effective vehicles for inter-generational transfers, but not good vehicles for the provision of retirement income.

Here’s why. If the public policy goal of tax deferral (and the tax expenditures associated with it) is to help people generate more income during retirement, then RMDs are essential to the achievement of that goal. They force the quasi-annuitization of savings and discourage retirees from using tax-deferred accounts for bequests. Ultimately, the government hopes that, if more Americans save, fewer will become dependent on government entitlement programs in their old age. There’s a method to the madness.

Landoni and Zeldes suggest that non-taxable RMDs could be added to Roth accounts to help achieve the policy goal described above. But that would remove a selling point of Roth accounts—flexibility in taking distributions. Financial advisers like the flexibility of a Roth. They have long advised retirees to tap tax-free Roth IRAs last—after spending down taxable and tax-deferred assets. Such delays naturally increase the chance that a Roth IRA will pass to a beneficiary.

Many people glibly say that RMDs exist because “the government wants its money back.” I doubt it. The federal government has shown that it can create all the money it needs whenever necessary. If Uncle Sam were so desperate for our tribute, you would see revenue agents going house-to-house every spring, demanding every last nickel or dime.

You might even see angry taxpayers pour boiling water on the tax collectors’ heads from second-story windows—as happened in my own neighborhood during the John Fries tax rebellion of 1798.

© 2020 RIJ Publishing LLC. All rights reserved.