Repeal the tax incentives inherent in 401(k)s to shore up Social Security? It’s a controversial idea raised by a trio of academics that immediately rallied defined contribution defenders.
One prominent supporter, Andrew Biggs, a Senior Fellow at the conservative think tank American Enterprise Institute and a member of the Social Security Advisory Board, makes his case below.
ICI senior staff members Peter Brady and Sarah Holden argue against it, using their own research to counter his claims.
Point: Reassess the Federal Tax Preference for Retirement Savings
In a recent working paper co-authored with Alicia Munnell and Michael Wicklein of the Center for Retirement Research at Boston College, we argued that the federal tax preference for retirement savings has been ineffective.
A better use for the money would be to maintain Social Security’s solvency. Eliminating the retirement tax preference would address roughly three-quarters of Social Security’s long-term funding gap, presenting a tempting source of resources for federal policymakers.
That proposal was, to say the least, controversial. Objections to our argument are worth considering, given the stakes for retirement savings and the federal budget.
But it helps to first clarify how the federal tax preference works. Contributions to employer-sponsored retirement plans and IRA accounts are exempt from income taxes. However, withdrawals from retirement plans are taxable.
Fiscally speaking, this deferred taxation more or less compensates the federal budget for the initial loss of income tax revenue, plus interest. However, it does not make up for lost capital gains tax revenues.
In effect, the tax preference lowers the capital gains tax on retirement plan assets to zero. The U.S. Treasury estimates that the net cost of the tax deferral, meaning the initial loss of revenues net of taxes paid later, was $185 billion in 2019, equal to about 0.9% of GDP. The Congressional Budget Office produces a very similar figure.
The tax preference intends to incentivize Americans to do something they otherwise would not do—save more for retirement. In general, a lower effective capital gains tax should encourage saving.
The problem we raise, however, is that households can reap the benefits of the tax preference without saving more simply by shifting savings from taxable investment accounts to non-taxable retirement accounts.
According to the Tax Policy Center, 59% of the total tax expenditure is received by households in the highest two income quintiles. Those households typically hold financial assets outside of retirement accounts that significantly exceed their retirement plan savings. It would be irrational for them not to locate assets strategically.
And, numerous studies find this to be the case. While we noted one highly-regarded study focused on Denmark, which found that each dollar of retirement tax expenditures generated only one cent of additional savings, we cited other studies that reached similar conclusions. For instance, one 2004 study tracked people over time as they enrolled in IRAs, finding that household spending did not decline, which implies that savings did not increase.
What happened is that savings in taxable investment accounts fell by an amount very similar to the maximum allowable IRA contribution. The authors found similar evidence of asset-shifting for tax-favored accounts in the United Kingdom. A 2014 study of tax-favored savings in Spain found that household savings did not increase. A 2022 study reached similar conclusions for Latvia.
Now,the tax preference might induce higher savings among lower-income individuals who lack any other savings to offset. But remember, the retirement tax preference effectively exempts a saver from paying capital gains taxes. These aren’t levied on households with incomes below $89,000, while the median retiree household has an income of only about $60,000.
As a result, the tax preference provides only limited incentives for low earners to save more. This partly explains why only 4% of the federal tax preference is received by households in the bottom two-fifths of the income distribution.
It also is possible that the federal tax preference expands retirement plan coverage, as employers would offer plans so their more highly paid employees could receive the benefit.
This may be so, but evidence for the argument is sparse. I performed a preliminary analysis of retirement plan coverage by state, reasoning that—all else equal—coverage should be higher in states that levy high-income taxes than in states with low or no income taxes. I found no statistically significant difference. And so, this question is unresolved.
Regardless, the United Kingdom’s policy of offering supplemental retirement plans to employees who are not offered a plan at work, and automatically enrolling employees in those plans, generates better results at a fraction of the cost. Since the U.K. approach was implemented a decade ago, private-sector retirement plan participation has risen from the low 40% range to above 80%.
Rolling back the federal retirement tax preference would indeed be a major change, and so should not be taken lightly. But so would be increasing the Social Security payroll tax, which reduces incentives to work and already is the largest tax that most workers pay. Similarly, eliminating the ceiling on wages subject to payroll taxes, as many propose, would effectively increase the top marginal tax rate by 12 percentage points, giving the U.S. one of the higher top rates in the developed world.
The cost of having delayed Social Security reform for decades is that policymakers face difficult choices. The federal tax preference for retirement savings is expensive, appears to do little to increase total savings, and provides most of its benefits to households who are at little risk of poverty in old age. Under the circumstances, policymakers should give it a second look.
Andrew G. Biggs is a Senior Fellow at the American Enterprise Institute.
Counterpoint: Do Not Sacrifice 401(k) Plans to Shore Up Social Security
A recent paper published by Andrew Biggs and Alicia Munnell proposes to sharply curtail the tax incentives for employer-sponsored retirement plans and use the revenue raised to shore up the funding of the Social Security system.
Biggs further defended his proposal above. The paper’s thesis is that the tax incentives for America’s voluntary retirement plan system do not appear to work and primarily benefit high earners.
This is easily dispelled.
The facts are that most workers accumulate resources from retirement plans at some point in their careers and eventually receive retirement income from these plans. And the benefits of tax deferral are not restricted to high earners.
It is true that most tax measures, expressed in dollars, will be skewed to high earners simply because both income and taxes paid are highly skewed. If you express the benefits as a change in average tax rates, the same Tax Policy Center estimates Biggs cited show that the benefits are a similar share of income for the top three income quintiles—that is, the top 60% of the income distribution.
In fact, they’re higher for the fourth quintile—folks with income between $92,300 and $167,000 in 2019 dollars, so not exactly rich—than for the top quintile, the highest-income 20% of the population.
American retirees rely on the combination of Social Security benefits and retirement plan income. And it is the Social Security system’s design, not the tax system’s, that primarily determines who benefits from retirement plans.
Social Security benefits replace a higher share of wages for low-income earners. As a result, these workers rely more heavily on Social Security in retirement, and middle- and higher-income workers rely more on employer plans and individual retirement accounts (IRAs).
As the Biggs-Munnell study notes, ICI research illustrates that higher-income workers benefit more from retirement plans because a higher share of their wages are deferred for retirement—not because they benefit more on each dollar deferred.
The facts are that most workers benefit from employer plans and IRAs, with middle- and upper-middle-income workers getting a substantial share of their retirement income from these plans.
Analysis of tax data by ICI economists found that 75% of 72-year-olds receive income from retirement plans. For those who were middle- or upper-middle-income before retirement, income from retirement plans typically makes up one-third to one-half of their total income at age 72.
The combination of Social Security and employer plans has allowed millions of workers to maintain their income in retirement. The study from ICI economists also found that, adjusted for inflation,
the typical American retiree maintains more than 90% of their average age 55–59 spendable income. While more reliant on Social Security, lower-income people have the highest replacement rates of their pre-retirement income.
The two parts of the retirement system work together to provide retirement security to American workers. It’s unhelpful to pit the two components of the retirement system against one another. Ironically, by substantially restricting the ability of workers to supplement Social Security benefits in retirement, the proposed policies risk reducing support for the U.S. retirement system as a whole.
Many of the policies Biggs and Munnell propose would drastically reduce the incentive for employers to offer retirement plans to their workers. For example, they suggest reducing the combined (employer plus employee) contribution limit to $20,000 or $10,000. That would be lower than the initial contribution limit established in 1974 in nominal dollars.
Their study is not overly concerned about the fate of retirement plans, partly because it dismisses employers’ important role. The symbiotic interaction of employers, workers, and the financial services industry has produced a successful employer plan system and a highly competitive market for retirement plan investments and services.
U.S. policymakers designed Social Security and employer plans to be complementary, allowing workers across the income spectrum to plan for retirement. Biggs and Munnell should not be so quick to dismiss the importance of employer plans. The data show that most retirees rely on a combination of Social Security and retirement plan income—and we should not undermine retirement plans in the hope it can help us fix Social Security.
The bottom line is that diverting tax incentives away from America’s vibrant 401(k) system to provide short-term funding for Social Security would harm millions of American workers endeavoring to save for a financially secure retirement. It’s the wrong move, and the data show why.
ICI will remain committed to the success of the retirement system for all Americans, including the millions of Americans contributing paycheck-by-paycheck to a voluntary plan. They deserve nothing less.
Peter Brady is ICI’s Senior Economic Adviser, and Sarah Holden is ICI’s Senior Director of Retirement and Investor Research.
