February 20, 2021

Could contributions to Health Savings Accounts (HSAs) actually “crowd out” contributions to 401(k)s?

A chart book just published by the Employee Benefit Research Institute (EBRI) suggest that the answer is—yes. It’s a conclusion—or at least a provocative question—that runs counter to the prevailing body of research on the subject, which has tended to find that, at least in aggregate, HSAs bolster the overall level of retirement savings. 

That said, in recent years a school of retirement counsel has suggested that it makes sense to consider—particularly if financial resources are limited—to take an integrated HSA/401(k) approach to maximize the benefits of the two in retirement planning. The thinking goes that the HSA and its “triple tax” advantage[1] is superior to the 401(k) (certainly with regard to paying for health care expenses in retirement)—with one exception—the match that typically accompanies the 401(k) deferral, but doesn’t always in the case of the HSA. So, the counsel goes, contribute to the 401(k) until you get the full match, then contribute to the HSA as much as you can (there may be financial constraints, after all, and there are contribution limits to the HSA), and if you are able to contribute still more, go back to the 401(k).

As rational as that counsel is, it’s admittedly complicated, and doing it “right” involves far more analysis and work than one typically finds among retirement savers. 

It is not beyond the realm of reason to think that financial constraints alone might well produce the result put forth by EBRI—that workers just might just not have “enough” money to retirement expenses in both an HSA and 401(k), and—if only because health savings accounts often support current health care expenses—to defer salary there rather than the 401(k). Indeed, in that sense the EBRI report affirms a certain rational perspective, even if it seems to run counter to the “prevailing wisdom.”

There are, however, it seems to me reasons to pause in jumping too quickly to the conclusions presented in the “chart book” EBRI offering which, it should be stated, lacks the narrative nuance more typically associated with the organization’s research.[2]

The conclusions were drawn from what amounts to a single year of experience (though over two years)—and focused on workers who had a 401(k) in Year 1, and both a 401(k) and HSA in Year 2 (2016), though it did what many are unable (or unwilling) to do, draw from a cross-sectional experience of more than 45,000 workers from EBRI/ICI’s 401(k) database and EBRI’s HSA database. EBRI found not only that workers reduced their contributions to the 401(k), but that more than half (56%) did. More significantly, they also found that as income increased, the percentage of participants who reduced their contributions increased.[3] Moreover, they also found that those with a higher rate of employer to employee HSA contributions were more likely to reduce those contributions.

I would not have been surprised to learn that lower income workers, feeling the “squeeze” of financial constraints and perhaps forced from a traditional health care program to the contributory designs of an HSA, to cut back on their 401(k). Indeed, in that sense the prior research that found no reduction by the vast majority of workers was, if anything, counterintuitive.[4]

But higher income workers? Not that they don’t (also) have financial constraints—but it was, to me, a bit of a headscratcher. 

In fairness, the reductions EBRI cites, at the median anyway—are modest; $315 among those contributing more than $4350 to their HSA—and a meager $8 among those contributing $1,000 or less (on the latter see footnote ii below). Medians, while more reliable than averages in discerning certain trends, aren’t without their opportunity for obscurity. In the EBRI report that relatively modest median result belies the range in results EBRI also provided; 1 out of 10 had a decrease of more than $5,127, while another 1 in 10 had a $1,143 increase in 401(k) contributions.

Setting aside that disparity, the decline among higher income individuals—their potential financial constraints notwithstanding—seems counterintuitive. Presented with such a range of outcomes, it’s dangerous to speculate, but such behaviors call out for some kind of understanding. One possibility is, of course, that they chose to implement the integrated strategy outlined previously, and simply forgot to go “back” to the 401(k) (or ran out of funds to do so). 

That said, it occurs to me that they might simply have bumped up against the 402(g) limits (not to mention an ADP/ACP or a 415 cap)—and thus the reduction might have had nothing to do with the HSA. At this point we can’t tell—and with only the experience of one year following their first contribution to an HSA—we have no idea whether it’s a “blip” or a potential trend. 

All in all, it seems to me that there’s still a lot of “smoke” in the results. While we have a “what” (one that is quite broad in its results) and a “when” (that is equally narrow in timeframe)—we have very little yet upon which to draw reasonable conclusions as to the “why.”   


[1] See What’s Holding Back HSAs?   

[2] Not that, in many cases a picture is worth a thousand words, but pictures/graphs sometimes require more explanation than a PowerPoint heading can provide, and such seems to be the case here.

[3] They also found a reduction among lower income workers, but that reduction was $8/year at the median, which seems almost to be a payroll rounding result. EBRI comments that this means they were “essentially unaffected.” 

[4] At least until one considers that the contributory decisions for a 401(k) and HSA are generally made at different times, and communicated via different/distinct mechanisms. In the case of the EBRI analysis, since these individuals already had a 401(k) in place, a reduction in deferral would likely have meant filing a change with payroll or via a 401(k) website which again seems like more energy that most dedicate to such things.