Turning ‘Negative Elections’ into Automatic!: a true story
J. Mark Iwry, Nonresident Senior Fellow, The Brookings Institution
Mark ❤ Iwry’s bio runs long and deep. He served as Senior Advisor to the Secretary of the Treasury from 2009 to January 2017, and served concurrently as the Treasury Department’s Deputy Assistant Secretary for Retirement and Health Policy. His portfolio included pensions/retirement (DB and DC/401(k) plans, IRAs, etc.) and national savings policy; related tax policy and other tax and legal aspects; legislative and regulatory implementation of the Affordable Care Act, and other health policy; other employee benefits and compensation, and related legislative, policymaking, and rulemaking responsibilities.
Currently a Nonresident Senior Fellow at the Brookings Institution and a Visiting Scholar at the Wharton School at the University of Pennsylvania, Mark is the originator or co-author of the Saver’s Credit, the SIMPLE IRA, the startup tax credit for small business, the QLAC, auto rollovers, direct deposit and splitting of tax refunds in IRAs and savings bonds, the automatic IRA proposal (with David John), and various other reforms and innovations, including the nationwide movement to involve the states in promoting retirement savings through auto-enrollment of uncovered citizens in private-sector IRAs.
While in government, Mark also spearheaded the initiative to define, approve, and promote the practice of automatic enrollment in 401(k)s and similar plans, which laid the groundwork for automatic escalation, default investment in QDIAs, default rollovers and retirement income, and, in later years, the federal and state auto IRAs. We talk about how the auto enrollment concept first came to light and started to take root. Twists and turns ahead!
Mark, how did the “automatic” enrollment movement in retirement readiness begin?
In 1997, while I was at the US Treasury Department with responsibility for overseeing national retirement policy and the regulation of pensions, 401(k)s, and other tax-qualified plans, an obscure legal question arose. Could 401(k) plans put people into the plan without the employees signing up so long as they could opt out if they wanted to? Could a 401(k) impose this kind of “negative election”?
Most people were unaware of this question, although we heard there actually were a couple of companies using these negative elections in 1997. It was definitely a gray area: people who heard about it tended to be surprised and to wonder whether it was even legal.
It's always a gray area when you do it for the first time, isn't it!
Yes, like lots of things. So when this somewhat esoteric question came to our attention, I remember thinking: Really, how much does this matter? Is there anybody who actually wants to do this? I also wasn’t sure we would have the time or bandwidth to deal with that sort of a side question.
That's how it first struck me, unfortunately. I remember asking the staffer who brought it in, “Should we really add this to the large backlog of projects and issues we're already working on? How did this come up, any way -- someone at IRS heard someone in the private sector asking the question. We already have dozens of issues stakeholders are asking about all the time -- important, pressing questions generating controversy and lively debate in the market. And we’re trying to keep our focus on developing national policy to promote greater retirement security and saving. So why distract ourselves by adding this kind of arcane legal question to our already daunting inventory of projects?
It took a few weeks before the light bulb went off, and I felt pretty thick when it did: Hey, wait, you know that funky idea of putting someone in a plan even though they didn't sign up, and letting them opt out if they want to? Maybe we shouldn't be viewing this as an obscure and relatively unimportant technical question. Couldn't it actually be deployed as a powerful strategy for increasing 401(k) participation?
We were focused on participation because of the decline of the defined benefit plan and the rise of the 401(k). Could we find ways to “DB-ify” the 401(k) to make it a more effective retirement readiness tool, more like a pension plan?
What do you mean by that?
Defined benefit pensions and in fact, classic profit sharing plans or money purchase pension plans don't depend on people to sign up. They just cover employees automatically. Why couldn't a 401k do this?
We started to wonder whether we could come closer to universal participation among those who were eligible for 401(k)s. Maybe the negative election should be viewed not as just a weird, arcane notion, but as a potentially pathbreaking policy strategy.
Therefore, instead of asking whether the 401(k) regulations permit this, maybe we should be asking, could this potentially revolutionize the 401(k) landscape and advance the goals of retirement policy? And therefore, shouldn’t we make sure that the rules will permit it, even if that requires changing them?
That’s a pretty big shift in thinking.
Yes, but Treasury and IRS own the 401(k) regulations. I asked my guys to advise, do you think we can do this administratively or, if we want to allow this, do you think we need to go through Congress, which could take many years? And the conclusion was that we could permit negative elections without an act of Congress since the 401(k) statute simply says contributions by cash or deferred “election” without defining exactly how the election needs to be done.
So that's where we started to get excited about this. But I also started to worry: if we did this, would employer matching contributions go away?
Beware the unintended consequences.
Exactly. The law of unintended consequences is the only law that Congress can’t repeal. In early 1998, when considering whether or not the government should define, approve, publicize, and promote the concept of auto enrollment, what was causing me to lose sleep was the risk of losing employer contributions as a potential, major unintended consequence.
We know employers make matching contributions to have an attractive benefit package and to hire and retain talented workers in competitive labor markets. But they also do it to help them meet the 401(k) non-discrimination rules. If we announced, approved, and were successful in promoting auto enrollment, it too would helping employers pass their 401(k) ADP tests by encouraging more average and lower-paid employees to participate.
I worried whether employers might think, “auto-enrollment is the cheap way to improve our ADP performance by getting more rank-and-file employees to participate. It won’t cost me money the way the match does. So maybe I'll drop the match.”
Over the course of the spring my staff and I persuaded ourselves that this outcome was unlikely. You'd have to cut the match for everybody, not selectively -- a big takeaway for the workforce. So we figured, it's a risk, but a small one. Most likely, if we succeed in encouraging the use of auto enrollment without losing the employer match, that will mean more people saving. And auto enrollment encourages participation especially by the most vulnerable among us, those who are traditionally left behind and who need the help the most.
Sounds promising!
In June of 1998, the Congress and the Executive Branch co-sponsored the “Saver’s Summit” – the first of three Summit Conferences of several hundred invited experts, regulators, policy makers, and industry leaders to share ideas about the private pension system and savings and what to do to improve it.
It was a big deal. President Clinton was going to speak. Speaker of the House Newt Gingrich and Senate Majority Leader Trent Lott were going to speak. So I knew it would soon be “feed the beast” time: the White House would be looking for a new and good proposal that the President could announce.
Now that the statute of limitations has run out, I can confess that I waited until the optimal moment, while the White House was getting nervous that the conference was fast approaching and they didn't yet have a great new idea. I kept saying, don’t worry, we're working on something, and I think you’ll like it.
I finally unwrapped the concept of automatic enrollment at the moment when I knew the White House would be most anxious that time was running out. They loved it.
(By the way, it was not a political decision: I told the White House that we had decided at Treasury and IRS to announce that we're going to let 401(k)s do this, and that ruling was being issued without waiting for the summit conference and the speech.) The purpose was to expand participation in 401(k)s, especially for minorities, women, and lower wage people who, to this day, still contribute and participate less than others. And so it was embraced.
Nice!
Actually, before taking it to the White House, I got together with my staff and said, okay, let's focus on how we rename this thing. We know we can't promote something called a “negative election”. This is Washington -- names matter!
We spent an entire lunch time one day kicking around possible names. We ultimately settled on “automatic enrollment”. It describes what it is. You're enrolling people automatically, but they can opt out. But I wanted to plant another seed, if you will, of making every phase of the 401(k) saving process “automatic” (in other words, using intelligent defaults) to make 401(k)s more like pensions: automatic enrollment, automatic contributions, automatic investing, automatic payouts and rollovers.
In fact, you can’t have automatic enrollment without specifying an automatic, or default, investment. Thoughtful, diversified default investments could steer the system in a more positive direction than reliance on employer stock and simple money market funds, in which too many people were still investing most or all of their retirement savings over the long term.
So we wrote a ruling that defined and approved auto-enrollment. We deliberately minimized any regulation, to make it simple and easy. The main requirement was just for plans to give eligible employees ample advance, written notice.
By the way, this was also the deep background of the QDIA regulations. We designed our Treasury/IRS ruling to illustrate auto enrollment into a default investment that was a balanced fund of diversified equities and fixed income securities (and no employer stock). But the prudence of the default investments under ERISA was in the jurisdiction of my counterparts in the Labor Department, and I encouraged them to simultaneously issue rules – as a companion to our Treasury/IRS auto enrollment ruling -- approving these kinds of diversified funds as default 401(k) investments.
But they weren’t ready – not until 8 years later, when my friend Ann Combs was at Labor and decided to develop the QDIA regulations (just before Congress heard about it and directed Labor to do so!).
Mission (Almost) Accomplished!
President Bill Clinton gave his speech at the Saver’s Summit in June of 1998, including auto enrollment as a centerpiece. It was covered in the New York Times, Wall Street Journal, etc., and after a couple of lonely years pitching the idea of auto enrollment to the 401(k) market, it finally started getting traction, and was adopted in an estimated 34% of larger 401(k) plans.
At that point, I drafted and encouraged Congress to include several legislative auto enrollment provisions in the Pension Protection Act of 2006 in the hope of continuing to encourage the spread of auto enrollment in the market.
And here we are just a few short years later.
Yeah. Although it's not that short: from 1998 to 2021. We’d hoped, as this thing started to take off, that, after ten years, it would be everywhere. By now, it’s estimated that over 70% of larger 401(k)s use automatic enrollment, although, as you know, it’s still working its way into the small end of the market.
What an excellent reminder about patience and persistence. 1998 to 2021 may be a long road, but it’s been a very valuable one.
Thank you so much for the scoops Mark Iwry! As part of his continuing illustrious career, Mark focuses on retirement and health policy, sometimes in the forefront and sometimes behind the scenes. You can connect directly with Mark here. You can follow Mark’s work at the Brookings Institution here. You can also connect with Mark on LinkedIn.
This piece was featured in the May 20, 2021 edition of Retirement Security Matters. For more fresh thinking on retirement savings innovation, check out the newsletter here.
