The famed behavioral economist compares autoenrollment and escalation to autopilot—get the altitude and heading wrong and it’s all gonna crash. What’s just over the horizon for 401k saving nudges, and what dynamic will digital innovation play? Here’s what advisors need to know.
We’re well aware of the sensational nature of the question, but nevertheless muster the courage to ask.
“Do you feel you should have shared the Nobel Prize with Richard Thaler?”
It’s a point raised by several top 401k advisors since the committee’s November announcement of Thaler’s win, most of them referencing UCLA behavioral economist Shlomo Benartzi’s own groundbreaking work in the discipline, as well as his frequent collaboration with Thaler.
Benartzi seems genuinely surprised by the query.
“Wow!” he says before quietly repeating it. “Wow. I am flattered to hear that some people feel this way and, um …”
He’s still a young man, so there’s time.
“I’m still a young man, so we’ll leave it there,” Benartzi diplomatically concedes.
It’s the only time he sounds unsure of himself. The high-profile academic has made it his mission to discover why people behave the way they do, especially in the area of 401k retirement investing, and how to cost-effectively remove inherent biases and barriers that prevent participants from saving more.
A deep dive into the human psyche would surely drive anyone mad, which is why it probably requires a scholar’s touch. Yet the real-world results from Benartzi’s research play a large part in the auto-revolution (enrollment, deferral, escalation), arguably at a time it’s needed most due to longer lifespans and the defined benefit’s demise.
While insight into the behavioral aspects of retirement saving is fascinating and effective, a shot of steroids in the form of technology is making it exponentially more so, and the focus of Benartzi’s current work.
Jumping off the widely discussed concept of behavioral “nudges,” or the theory that subtle and indirect suggestions can positively influence participant outcomes (which also happens to be the title of Thaler’s bestselling 2009 book), Benartzi is currently hyping digital nudges and the critically important benefits they provide in helping 401k advisors pump enrollment rates.
For an idea of just how important, his research shows that digital nudging is more effective by “a factor of 100” than traditional nudging. But caution is warranted, he adds. If 401k advisors and sponsors push too hard or advocate for deferral rates too high, damage can be done (more on that later).
“Digital nudging is anything from how you design a website, app or text message,” he says. “It might be the content, the timing or the visuals. It might be the technology you employ when sending a 401k statement. It could be a personalized, 60-second video rather than a 40-page document.”
It’s combining “all of the aspects of the digital world together with a bit of psychology” and (what else?) nudging.
Benartzi immediately points to three advantages of doing so:
1) Speed—“You can do it very quickly,” he simply states. By “it” he means sending, receiving and analyzing the results of enrollment and engagement efforts, and determining if time and resources are better allocated elsewhere.
2) Scale—It’s easier to ramp up and reach more people than with traditional nudge methods.
3) Adaptability—We normally miss out on the ability to learn from, and adjust for, what’s happened in the past. Not so with digital nudges.
“We can run a lot of experiments and A/B tests to try different language and looks, as well as the time of day when we ask people to save more,” Benartzi explains.
“We can see almost immediately what works and what doesn’t, and then finetune the nudges from there.”
It was something not possible in the offline world. Sure, the research and learning capabilities were possible and even prevalent, but it took a heck of a lot longer, was more expensive, and a lot less precise.
“When you mailed someone a postcard, you didn’t know when (or if) they received and opened it. Not only did all your measurements take longer, they were exponentially more expensive. Because you didn’t have that precision, you couldn’t learn as much.”
In a recent article for Harvard Business Review, Benartzi ran through the digital nudges he’s recently employed, including the power and process of framing an email, electronic “mental accounting” (a person’s tendency to make different choices about the same scenario depending on how it’s presented) and the impact of immediate mobile feedback.
So what’s he working on now, and what excites him for the future?
“As you probably know, I am an academic advisor to Voya,” he says. “We’re working on personalized videos not only to participants, but to plan sponsors and advisors, with embedded calls to action. It’s important because people tend to ignore everything that’s not relevant to them, but they respond to this because it’s customized and because of its visual, dynamic nature.”
Employers and plan sponsors are therefore better able to engage prospects and participants, and link that engagement to specific actions. Think of it as a one-click process, like something Amazon uses, with an intuitive yes-no, accept-decline option for increasing savings by, say, 1 percent.
But it should be as simple as possible, to which Benartzi agrees, with a caveat.
“Yes, you want to make it simple, but it first needs to be engaging. If it’s really simple, but they don’t know where and what it is, they won’t take action and there’s no point. There are more components to this puzzle, one of which is also the use of big data to make targeted and smart suggestions.”
For example, he points to the United Kingdom and discussions surrounding merging data from their single-payer system with retirement information to make personalized recommendations for the purchase of annuities, based on the health and welfare of the individual.
“If you’re sick, you’ll be offered an annuity that pays more every month, and the process would all be automated,” he adds.
While it probably wouldn’t fly with HIPAA and the more libertarian preferences of their American cousins, his point, and its potential, is hard to argue.
“There needs to be a combination of simple and engaging, because we certainly wouldn’t want people buying annuities without understanding what they are. If we can communicate complicated concepts in digestible pieces while taking digital attention spans into account, it would be critical.”
But is this retirement saving technology a silver bullet that will save us from ourselves?
Yes and no, Benartzi counters. He argues digital nudging is far more impactful than supposedly positive behavioral influences like tax benefits, matching contributions and financial education.
“It obviously won’t solve everything, but if you’re deciding where to spend limited resources in order to move the needle, there is absolutely no doubt in my mind that it has to be a digital nudge because the rest of the stuff is more expensive and less effective.”
Surprisingly, tax minimization inherent in retirement savings plans, often touted by advisors as a reason (if not the reason) for participating, barely result in higher enrollment rates. So, all of the drama over potential contribution cuts and rigmarole in recently passed tax reform won’t have much of an effect, we naively ask?
“No, of course not,” Benartzi answers, somewhat exasperated. “Think about it this way. The vast majority of Americans don’t understand the tax system because it’s so complicated. It’s very difficult to have something we don’t understand affect our behavior. If you add inertia on top of that, and the fact that many people in a 401k plan are saving because it’s on autopilot, they’re not going to change their behavior because of some new tax code that nobody understands.”
Which makes sense, but what are some of the absolute “must-have” features that 401k sponsors should include that would change behavior to ensure the best chance for participant success?
Emphasizing the accumulation phase, he says “no matter how you look at it,” it’s automating the plan through automatic enrollment, automatic increases and automatic portfolio choices for the participant. Offer those and there is a very good chance “they will reach the finish line.”
“Having said that, what we are generally missing is that the devil is in the details. A lot of plans implement automatic features, but they do it the wrong way. Historically, for example, the most common default saving rate with automatic enrollment was 3 percent. And later, we found a study suggesting that people can stomach 6 percent, so why not double it actually?”
So, he did.
His recent work with Voya included testing the suggested rate above 6 percent, even going as high as 11 percent. He found that a further boost could be had by increasing defaults from 6 percent to 7 percent. But humans being humans, the law of diminishing returns will eventually strike.
“We’re still playing [with the optimal deferral rate]. What we find, which is interesting, is that once you push it beyond 7 percent, people scale it down, which is something we did not expect. It doesn’t backfire, it’s not like they run away, it’s not that they push it down too low, but if you set it up to 9 percent, you do not seem to have any increase beyond what people do with 7 percent.”
Which is exactly what the industry should want, discovering the point at which people adopt the default, but also the point at which people engage and recognize it might be too much for their given situation.
“In a hypothetical, crazy world, you don’t want people to set their default saving rate to 50 percent. You don’t want people to pile on credit card debt, or [mortgage] their house in order to save 50 percent. But the fact that at some point they engage turns out to be, in my opinion, very good news because they personalize the solution to what fits them.”
It’s enough to get their attention but not enough to turn them off or to have to make even worse decisions?
“Correct, and it’s all in the details. If people get stuck saving 3 percent then yes, automatic enrollment makes them start saving, but they are going to save very little, so you haven’t really solved any of the crisis. So, across-the-board, whether it is automating the process of getting people into the plan, automating the increases or automating the investments, the devil is in the details. We’ve made a lot of progress, but there is a lot more to be done.”