Skip to Content

The IRS Takes a Big Step Toward a Small Reduction in RMDs

But these changes will only help so much.

Back in January, I wrote about legislative proposals to postpone required minimum distributions and an executive order President Donald Trump wrote directing the IRS to make technical changes in the RMD calculation that also was designed to reduce these required distributions. Although legislation to push back RMDs to later ages has made it out of the House of Representatives, it remains stalled in the Senate. Unlike Congress, the IRS has been working on the issue and took an important first step toward reducing RMDs for most retirees.

This will make a lot of people happy because no one really likes RMDs, and it's easy to understand why: Who wants to take money they do not need out of their 401(k) and IRA accounts? That's particularly true for retirees who are bumped into higher marginal tax rates because of these withdrawals, especially when these RMDs increase the percentage of Social Security benefits that are subject to taxation.

As I noted in January, the current RMD rules are dated, going back to 2003, and they use even more out-of-date mortality assumptions. In fact, the IRS used mortality data from the year 2000 when it designed the table to spread 401(k) and IRA withdrawals out over a couple's lifetime. Almost 20 years later, we can expect people to live a little longer than they once did.

Back in January, I observed that if the IRS recalculated these two-decade-old mortality tables and updated the rules, it would be modestly helpful but wouldn’t make that big a difference. I noted that the current rules already make a very generous assumption that a spousal beneficiary is 10 years younger than the IRA owner, regardless of his or her actual age. (People with spouses more than 10 years younger do use a separate table.) However, for single people, the changes could potentially reduce their RMDs by bit more, depending on their age.

Now that the IRS has formally proposed a change, it turns out this intuition was right. Single retirees would see their RMDs drop about 10% from age 70.5 (when required distributions commence) to age 80, although after that the updated calculations have less effect on distributions, which are only slightly lower than under the current rules.

Couples would see much less of a change, with just a 6% to 7% decrease in their early years, depending on their age. Interestingly, (and not surprisingly), the RMDs fall more compared with current rules for couples when the account owner is in his or her mid-80s. This is because most of the improvements to the mortality table come from ages 70 to 80, and the spouse is assumed to be 10 years younger than the account owner. However, those reduced RMDs could be worth less by the time someone attains these later ages, since the owner would have been drawing on the balance for some time. (It depends, of course, on the investments in the account and how well they perform.)

A last wrinkle is that, for people who live an unusually long time (into their 90s), the RMDs actually go up as the IRS would now assume that mortality rates, contingent upon reaching these ages, are worse than they assume they are today. (This is in part a function of more people reaching these ages because of improvements in life expectancy for people in their mid-70s.)

It seems very likely that this rule will become a final regulation sometime in 2020. As proposed, it would apply to distributions starting in January 2021, so a little relief for retirees who don’t like RMDs should be coming in a little over a year.

More on this Topic