The Ambiguity of Tax Deferral

A failure to agree on whether or not there's 'government money' in tax-deferred savings accounts has complicated our debates over the DOL fiduciary rule, the need for a harmonized fiduciary standard, and the so-called Rothification of 401(k) plans.

When you look at your 401(k) or 403(b) or rollover IRA, what do you see? Do you see a two-sleeve account, where you contributed about 80% and the federal government contributed about 20%? Or do you feel like the account contains only your money, with no help from your Uncle Sam?

To economist Steve Zeldes of Columbia University, it’s obvious that Uncle Sam contributed part of your balance. But advisors who manage brokerage IRAs tend to believe that the money is all yours. Similarly, many affluent retirees over age 70½ wonder why the IRS requires them to withdraw three or four percent of their tax-deferred savings each year and pay ordinary income taxes on it. 

Just as there’s more than one way to look at the image that illustrates this article, there may be more than one way to think about tax deferral. Indeed, our national arguments over the Department of Labor’s (DOL) fiduciary rule, the potential “Rothification” of defined contribution plans, and the need for a harmonized DOL-SEC fiduciary rule, may stem from the fact that we don’t all think about it the same way.

An academic view

At the annual academic forum of the Defined Contribution Institutional Investors Association (DCIIA) in New York a few weeks ago, Zeldes presented a research paper he’s been refining for a year or so. The paper (available on SSRN at ) asserts that our “Traditional” 401(k) system, in which contributions are tax-deductible today but taxed at withdrawal, leads to a U.S. government subsidy to the private asset management industry of about $13.5 billion relative to a pure Roth system in which all contributions are made with after-tax money and never taxed again.

Zeldes explained that when 401(k) plan participants defer $1,000 a month into their plans, they contribute about $800 and the government contributes about $200. By excluding $1,000 from their gross paycheck, they reduce their tax withholding by about $200 (assuming they are in the 20% tax bracket). The $1,000 contribution to the plan cost them only $800. The other $200 came from Uncle Sam, and results in a boost to the size of the 401(k). But it’s not a permanent boost; the government will collect it back when the participant withdraws the funds later in life and pays taxes on the principal and gains.

When you consider that the asset management industry charges retail (not institutional) fees on the assets purchased with that money, you wind up with a subsidy, reasoned Zeldes and his co-author, Mattia Landoni. They assumed that of the $14.4 trillion in tax-deferred accounts today (excluding defined benefit plans) the government’s contribution is worth about $2.9 trillion.

The two economists then assumed that asset managers and other financial intermediaries charge about 72 basis points (50 bps for investment management and 22 bps for trading costs) a year on that $2.9 trillion and that, as corporations, they pay taxes on profits at the rate of 35%, so that the overall fee would be 0.72% x (1 – .35) = 0.47% or about $13.5 billion a year. Forty-seven basis points, the reasoning went, is perhaps ten times as much as a big institutional investor like Uncle Sam should have to pay.   

“The government could achieve savings equivalent to $13.5 billion per year by forcing the conversion of all existing tax-deferred retirement accounts into Roth accounts,” Zeldes and Landoni wrote in their paper. This $13.5 billion, a cost to the government, is an annual subsidy to the asset management industry.”  

A $200 gain or $800 loss?

The several hundred executives from asset management and related industries listened calmly and impassively to Zeldes’ presentation. Presumably, they do not think of themselves or their industry as partially underwritten by the federal government. No one in the audience seemed to register surprise or offense at the implications of the paper. (Zeldes told RIJ afterwards that it was only a coincidence that he was presenting the paper at the same that Congress was wondering—in the face of massive resistance from the financial services industry—whether to “Rothify” the retirement system by making contributions taxable and withdrawals tax-free.)

It’s possible that few or none of them accepted Zeldes’ assumption that 20% of the money in tax-deferred accounts came from the government. It’s possible that plan participants do not perceive that a benevolent government simultaneously enhances their retirement account by $200 and charges them $200 less in taxes (per $1,000 contribution).

Perhaps plan participants believe that they put $1,000 into their tax-deferred accounts, and that the government took $200 less in taxes in their paychecks—which they spent on groceries and so forth—without enjoying a sense of extra wealth. Perhaps they felt the pain of an $800 drop in their take-home pain without feeling any pleasure that it didn’t drop by $1,000. 

Skin or no skin

Anecdotal evidence suggests that most people regard the money in tax-deferred accounts as theirs and their alone, with no government “skin” in their “game.” This question came up at a conference for asset managers a few years ago, when that industry was excited about capturing rollovers from 401(k) plans. Asset managers looked forward to charging retail fees, rather than the relative lower fees charged in 401(k) plans, on trillions of dollars in rollover assets.

A speaker at the conference asked members of the audience if they saw any difference between qualified money and other money. Only one person raised his hand to answer. “Yes. It’s tax deferred,” he said. Brokerage firms and advisors don’t seem to regard Uncle Sam as a minority partner in brokerage rollover IRAs either.  

An official in a retirement trade group, when asked about this question, saw no government skin in the game either. Describing the difference between a Roth 401(k), a traditional 401(k) and a conventional taxable savings account, he said, “It’s just a matter of whether somebody wants to pay their taxes up front or later.” His view may be based the fact that traditional and Roth accounts produce the same amount of after-tax income in retirement, if the saver has the same tax rate at the time of contribution and the time of withdrawal).

As for retirees who must take taxable required minimum distributions (RMDs) from their 401(k)s or IRAs starting at age 70½, anecdotal evidence suggests that they don’t understand why, if they didn’t need the withdrawal for current expenses, the IRS forces them to take a withdrawal and pay taxes on it.

Even if their tax rate in retirement is lower than their tax rate during their working years, they don’t seem to feel the difference, let alone take pleasure in it. When asked, they don’t appear to recognize that RMDs are the only way that the government can make sure that tax-deferral achieves its public policy goal of producing retirement income.   

In the United Kingdom, the government emphasizes its role in tax-deferred savings. On one government website, a diagram explicitly shows savers that their tax-deferred accounts have three sleeves: their own contributions, Her Majesty’s contribution, and their employers’ contributions. In the U.S., the government makes no similar effort to frame the discussion.

The meaning of tax deferral

You might ask if it matters how we think about tax deferral. I think the consequences run into billions of dollars and years of time spent inefficiently if not wasted.

The huge, expensive, drawn-out conflict over the Department of Labor (DOL) fiduciary rule, and the current debate over “harmonization” of fiduciary standards between the DOL and the SEC, and the fight over whether to Rothify 401(k) were arguably all created, or at least exacerbated, by a failure to agree on the way we think about tax-deferred accounts, and whether the government has a stake in them. 

Brokerages fighting against the fiduciary rule can accuse the DOL of regulatory “overreach” because they don’t believe that there’s government money in rollover IRAs. The DOL can argue that there is government money in those accounts. Or it can argue that the government has an interest in those accounts merely because the accounts remain tax-deferred. That is: the money remains government-subsidized pension money, just as it was in 401(k)s. The DOL can even claim that it has an obligation to regulate it.  

The inability to resolve the difference in our perception of tax deferral is creating delays and indecision not just for the DOL fiduciary rule but also in the current battle over tax reform. If we knew, for instance, whether 401(k) participants actually regard tax deferral as important incentive to save, the debate might have more clarity. Similarly, agreement on the significance of tax deferral might also illuminate the debate over whether there should or shouldn’t be a higher fiduciary standard for brokerage IRAs than for taxable accounts.

Instead of addressing these issues directly and openly, different sides have taken rigid positions without first agreeing on important facts. The facts, unfortunately, can be ambiguous—as in the case of the picture at the top of this article. We could be so engrossed in the ambiguity between the newspaper and the landscape that we don’t realize that we’re looking at a portrait of the famous Abbey St. Michel in Normandy, France.   

© 2017 RIJ Publishing LLC. All rights reserved.