Economists love to talk about unintended consequences. Anytime someone mentions some well-meaning policy, you can count on an economist to point out the collateral damage. Do you want to impose rent control to help people afford their housing? We will point out constricted housing supply. Or, give people a holiday on the gas tax to help them afford their commute? That will just increase gas prices, eating into those savings. Maybe you want to encourage business investment by giving a larger capital depreciation allowance? Fine, but get ready for lousy automated customer-service robots replacing the now relatively expensive people.
We just can’t help ourselves from pointing this stuff out, despite the fact that it makes us unpopular at parties (sigh). So, when a recent article by a team of researchers led by Sita Slavov was published in The Journal of Public Economics pointed out an unintended consequence of state Auto-IRA policies, I was hardly surprised. Except that the unintended consequence they were pointing out was…gasp…positive!
Some Background on State Auto-IRAs
The first state Auto-IRA program was launched in Oregon in 2017 and since then nine other states have followed suit. Because Oregon’s program is prototypical, it is worth describing. Employers not offering a 401(k) are required to automatically enroll their employees in a Roth-IRA at a rate of 5 percent. The employee can opt out at any time.
This invested money is put in a safe, money-market fund for 60 days and then into a target date fund unless the individual chooses another option out of a limited set. The Roth structure and immediate placement in a money-market fund is key to the design. People who want their money back can get it without tax or penalty and, for two months, without risk of loss. The thinking here reflects the reality that workers whose employers don’t provide retirement accounts tend to be lower-income and more likely to need their money in a pinch. And, I should know the underlying thinking, as I was a member of the consulting team that helped Oregon design their program.
Private Sector Opposition
As part of the Oregon team and having also consulted with Illinois, Connecticut, and Washington State, I talked to many people in the 401(k) industry. Some were supportive, but more often they seemed to view these programs as competition. A conservative taxpayer group in California sued to stop that state’s program, and the first Trump Administration rescinded an Obama-Era regulation making it clear that the state programs were allowed under Federal Law.
On the one hand, I could understand the fear. State Auto-IRAs were giving employers not currently offering a retirement plan an easy way to offer a savings vehicle. So, the state programs could get in the way of recruiting new business. An argument that was frequently made to me by these folks was that because State Auto-IRAs are in some ways inferior to 401(k)s – they have lower limits and do not allow employer matching – that state plans could lead to worse, not better, coverage.
On the other hand, these complaints felt misguided and just a bit disingenuous. If the financial industry’s 401(k)s were so superior, then why were these state programs such a threat? And, it is not like the 401(k) industry has had much success reaching the sorts of smaller firms with lower-income workers that state Auto-IRAs were meant to serve. After all, in an upcoming issue brief that I wrote with two colleagues, we show that the coverage rate for employer-sponsored retirement plan has been flat for decades. Those working on these projects began to argue the reverse – what if these programs could stimulate coverage? After all, 401(k) providers could now say to these small employers – “hey, you have to offer this IRA anyway, so why not try one of our plans?”
The Private Sector Was Wrong
It turns out, that’s exactly what has happened. The Pew Charitable Trusts had already conducted research showing that Auto-IRAs didn’t “crowd out” (i.e., replace) existing 401(k) plans. But, the most recently published work suggests actual “crowd in.” The authors of the study compared employers in states that required Auto-IRA enrollment to similar employers in states that did not have these programs. The study found that, depending on the state considered, 8 to 23 percent of firms that hadn’t previously offered an employer-sponsored retirement plan were induced to do so by the existence of state-sponsored Auto-IRAs. In other words, state Auto-IRAs have helped private-sector 401(k) providers.
State Auto-IRAs certainly weren’t created to expand access to employer-sponsored plans. They were created to help people without those plans save. Yet, here we are, with an unintended consequence economists can finally talk about without getting socially shunned. In a world where the Trump Administration seems hell bent on damaging the economy, at least some states have gotten something right.