A mandated shift to Roth-style plans would harm middle-class workers.
The U.S. Congress and White House are talking seriously about passing a major tax-reform package. If they do, what can ordinary investors expect? Due to arcane budget rules, they likely can anticipate that a new emphasis on Roth-style retirement savings will be part of the discussion.
Major tax reform is tough—the last successful effort was in 1986 and before that in 1954. Recent history has not revealed much enthusiasm among members of Congress for the tough choices involved in a large tax-code overhaul. Yet, many experts think 2017 might be the year.
Republicans have offered some clear high-level principles for tax reform, such as reducing marginal tax rates while shrinking deductions and credits, but they have largely left specifics about retirement plans out of their early plans. For example, House Speaker Paul Ryan’s “Better Way” plan would lower the number of tax brackets to three from seven and reduce the marginal tax rates on those brackets, while making changes that would limit the number of people taking itemized deductions to around 5% of taxpayers. Further, the Ryan plan would try to encourage savings and investment in general— not just in retirement accounts—by reducing taxes on investment income.
Retirement Policy Is Important to Any Tax-Reform Plan
Still, changes to the tax treatment of retirement savings must be part of tax reform, even if there are few clear details yet to emerge. That’s because the “tax expenditures” on retirement contributions are among the highest in the tax code, at more than $100 billion annually, according to the U.S. Treasury. A tax expenditure is the amount the government “spends” by letting people or corporations reduce their taxes for various activities. The largest tax expenditures are for healthcare and mortgage interest, followed by traditional contributions to retirement vehicles such as 401(k) plans and IRAs.
Of course, the retirement tax expenditure is a little different from other tax expenditures because the government will collect taxes on traditional retirement savings someday. Traditional retirement savings are taxed in the future when people draw them down for retirement. However, because of the 10-year budget-scoring window, when the Joint Committee on Taxation and the Congressional Budget Office score this tax revenue, this future government revenue is mostly excluded.
10-Year Budget-Scoring Doesn’t Work for Retirement
The 10-year scoring window is just what it sounds like: Congress evaluates the effects of changes in the tax code based on revenue projections for the next 10 years, and typically no further. This makes a lot of sense for most policy evaluation. The CBO and the Joint Committee on Taxation cannot accurately predict further than 10 years into the future (let alone even 10 years into the future), making such efforts ridiculous. At the same time, at least trying to make these predictions forces policymakers to consider the medium-term impact of their legislation, which is important. So, 10-year scoring generally makes sense.
Unfortunately, 10-year scoring can have major distortionary effects on policymaking for retirement-savings incentives.
The reason is that 10-year scoring will show large tax expenditures for traditional retirement contributions but not for Roth contributions. Any shift toward Roth, in which the taxes are paid immediately on contributions but not on withdrawals, will generally create more revenue inside the 10-year window. Any shift toward traditional contributions, which are tax-deferred, will decrease revenue. Because Congress wants more revenue to offset other tax changes, increasing Roth contributions and decreasing traditional contributions will be very appealing to legislators trying to find a way to accomplish the medium-term goals of tax reform.
At this time, there are no concrete proposals on tax treatments for 401(k)s or other retirement savings. However, there is some recent precedent for a shift to Roths.
In 2014, then-U.S. Rep. Dave Camp, R-Mich., introduced H.R.1, a major tax reform bill that would have simply required that all contributions to employer-sponsored plans above $8,750 be Roth contributions. Assuming this plan were updated in 2017 dollars, it would require that all contributions in excess of $9,000 be designated as Roth contributions. (For the technical-minded reader, the law used one half of the 402(g) limits as the limit on pretax contributions, and it would also extend the limit for one half of catch-up contributions.)
Obviously, that plan failed to become law, but it reveals the interest lawmakers have in exploring a shift to Roth contributions. While that plan would have affected fewer than 20% of taxpayers, it is very possible that Congress could go further and simply require all future contributions to be Roth-designated. In either case, employer contributions to retirement plans would likely still be deductible as a business expense, so as not to discourage employers from creating new retirement plans or cause employers to drop their existing plans.
Who Wins and Who Loses?
People often note that Roth contributions and traditional contributions are equivalent, as long as tax rates in retirement are the same as they are during a saver’s working life, but this claim is not quite right. The tax benefit on a traditional IRA contribution is that it allows a worker to avoid paying his or her marginal tax rate on a retirement contribution—the highest rate a worker pays in our progressive tax code. The tax benefit on a Roth contribution may be as low as the effective rate—the weighted average of the taxes paid on all the tax brackets. It must be emphasized that the benefits of Roths can be on the effective rate, depending on the other retirement income a worker has and how he or she withdraws income for Roth accounts.
Unlike traditional contributions, Roth contributions reduce take-home pay by the full amount of the contribution, as the tax benefit on the Roth is deferred until withdrawal. In contrast, traditional contributions reduce a worker’s paycheck by much less, as the worker benefits from immediate tax savings for federal and, in most cases, state or local income tax.
To examine the tax-equivalent difference from a shift to Roth contributions, the following analysis assumes workers will not reduce their take-home pay, but will adjust their traditional contributions according to their marginal federal tax rate as they shift to Roth. (Because they vary widely, we do not use state income taxes for this analysis.)
Our analysis shows that middle-class workers would likely experience the biggest income losses in retirement from a shift to Roth vehicles. For example, a middle-class 30-year-old worker in the 25% bracket earning $50,000 annually could see a sharp 7.8% annual reduction in aftertax income in retirement from a shift to Roth contributions, assuming tax brackets remained the same ( EXHIBIT 1 ). This example focuses on a younger worker because those are the ones for whom the change in tax timing means the largest changes in eventual income. Of course, younger workers who greatly increase their wages might benefit from the Roth, if they end up with more retirement income than they had at age 30.
Higher-income and lower-income workers would be hurt much less than middle-income workers under a shift to Roth contributions. As illustrated in EXHIBIT 2 , higher-income workers have much higher effective tax rates in retirement, increasing the value of Roth accounts in the drawdown phase. Lower-income workers are also hurt less because their tax burden is quite low during their working lives, reducing the benefits of traditional contributions. In fact, many low-income workers may benefit from a shift to Roth, because they may have no federal tax burden at all if they are eligible for the Earned Income Tax Credit or Child Tax Credit during their working lives.
While our analysis does not use a microsimulation model (of the type that the Employee Benefits Research Institute and The Urban Institute use) to estimate all the likely winners and losers from a shift to Roth, it seems clear from the analysis on prototypical savers that many middle-income people would lose from a shift to Roth contributions.
Extending the analysis, the deleterious effects of a shift toward Roth contributions would be mitigated for workers with employer matches and for those who have already saved substantial traditional retirement savings account balances ( EXHIBIT 2 ). The reason is that workers with employer matches (which we assume will still be treated like traditional 401(k) contributions) or with account balances will benefit from their mix of account types in retirement. Because they will have Roth and traditional accounts, they will be able to pay their lower marginal tax rates on traditional withdrawals, while using their Roth withdrawals in place of traditional withdrawals that would otherwise be taxed at higher marginal rates.
Indeed, higher-earning workers with an employer match and traditional retirement account savings would benefit from the shift to Roths under current tax rules, as their account mix would result in lower taxes in retirement.
Roth Withdrawals, Social Security, and Sensitivity to Inflation
One critical aspect of this analysis is the assumption that current rules regarding Roth withdrawals and their effect on Social Security benefits taxes will remain unchanged. Currently, for a single taxpayer collecting less than $25,000 in retirement, Social Security benefits are not taxable. (For a couple, the threshold is $32,000.) At higher levels of retirement income, more of the benefit is considered taxable income. For taxpayers earning more than $34,000 in retirement, up to 85% of the benefit can be taxable income ($44,000 for a couple).
Roth withdrawals do not count toward these tests for the taxability of Social Security benefits, somewhat offsetting the disadvantages associated with them. However, these thresholds are not indexed for inflation—unlike most other parameters in the tax code. This creates an increase in the percentage of Social Security benefits that are taxed as withdrawals, and the Social Security payments themselves increase due to inflation. As a result, projecting the effect of the shift to Roth is highly dependent on inflation assumptions.
The higher inflation is (assuming real returns remain constant), the less valuable the Roth savings are as a mechanism to reduce taxes vis-à-vis traditional defined-contribution savings.
For example, the 30-year-old in EXHIBIT 1 would have 7.8% less income in retirement if she had to use a Roth rather than a traditional defined-contribution plan. (We assume employer contributions are traditional and inflation is 2%, but the Social Security tax rules are not indexed for inflation, as is current law.) If inflation were 3.5% annually, her income in a Roth system would be the same, but her income with traditional contributions would fall, leaving her just 5.9% worse off from a shift to Roth. (There is a point at which her Social Security benefits alone would be sufficient to incur a tax, but it will take a very long time to get there.)
On the other hand, the value of the traditional account is much higher if inflation is extremely low for a sustained period.
Other Concerns: Behavioral and Practical
Although our analysis shows that a shift to Roth could be harmful for purely rational retirement savers, another concern is that workers may not respond well to Roths. If workers do not see an immediate tax benefit, they may simply decide not to contribute.
On the other hand, it is also possible that workers might decide to contribute at the same levels they had before the shift because of the “anchoring effect” of their existing contributions, meaning they boost their aftertax income in retirement and reduce their take-home income today. In any case, a shift to Roth could be a major behavioral experiment that could be a disincentive for people saving for retirement.
Additionally, if workers do adjust their contributions so the decrease in their take-home pay after contributions is neutral, that could mean they forgo their employer match. For example, a worker in the 25% bracket (income between $37,950 and $91,900 for a single worker in 2017) who contributes 6% of his salary to a retirement plan would see his paycheck decrease by 4.5%, or less depending on state and local income taxes. If he shifted to a Roth, to see the same reduction he could only contribute 4.5% of his salary at most, but he would pay no taxes in the future on the contribution. However, if his employer matched his contributions up to 6%, he would be losing his match, or he would have to reduce his paycheck further to maintain the match, which might not be financially feasible.
A stronger emphasis on Roth is a policy preference driven by arcane scoring rules, and it might lead to unintended consequences. Many workers may end up with lower incomes in retirement from a shift to Roth, particularly those with lower savings today and lower employer contributions. Further, we don’t know the extent to which people rely on the immediate tax benefits of traditional contributions to save for retirement, nor do we fully understand the possible behavioral disincentives a shift to Roth could mean.
At a time when few are willing to make predictions about policy, it seems likely that if Congress pursues tax reform, 10-year budget scoring will make Roth contributions too appealing to ignore. Policymakers should carefully consider who the likely winners and losers would be of such a shift to Roth.
A more modest change, such as the one Camp proposed in 2014, would affect fewer people. Still, even in that case, any shift to Roth is an intergenerational transfer, trading tax revenue today for forgone revenue in the future.
This article originally appeared in the August/September 2017 issue of Morningstar magazine.