401(k) auto features don't help savings as much as thought, researchers find

Two of the most widely used mechanisms to improve participation in 401(k) plans – automatic enrollment and automatic escalation of contribution rates – may be much less effective than previously thought, a recent study found.

There are several reasons for that, and some have to do with turnover at the companies that sponsor 401(k)s, according to the paper.

“Our results suggest that the job transition moment is a key weakness in the US retirement savings system,” the authors wrote. One way of addressing that would be to make 401(k) assets harder to access until retirement, they said.

About 40 percent of private-sector workers who participate in savings plans are in ones that use auto enrollment or escalation, the authors noted.

Those avenues have been credited with getting many more people to save and invest for retirement than otherwise would have. The power of inertia, or making no decisions about their participation in employer-sponsored retirement plans, leads most workers to the default investment options and contribution rates in their 401(k)s.

Many large employers have used those features for years, and they are even now required by law. The Secure 2.0 Act mandates most new 401(k) plans to auto enroll employees and use auto escalation, by which their contribution rates increase annually up to certain limits. And many states have moved to require businesses to either provide retirement plans or enroll their workers in publicly administered IRAs.

But despite the success of selling auto enrollment and escalation, they appear to only be about a third as effective as prior studies have shown, according to the new research paper, “Smaller than We Thought? The Effect of Automatic Savings Policies,” published by the National Bureau of Economic Research.

Previously, auto enrollment was credited with boosting savings rate by 2.2 percent of income – but the new paper puts the figure at just 0.6 percent. When combined with auto escalation, the rate goes up to 0.8 percent, the authors found.

Even while lower than researchers thought, the effects of auto enrollment and escalation are net positives, said James Choi, professor of finance at the Yale School of Management and one of the paper’s authors.

“If you thought you were going to nudge your way to savings nirvana,” you might have overestimated the effectiveness of that strategy, Choi said. “It helps a little bit … Nudging by itself is not going to get you out of the [retirement savings] crisis.”

There are four main reasons why the effects on savings behavior are lower than thought.

The first is that workers in plans without automatic features tend to choose higher savings rates over time, so the benefits of those features might be relatively small. However, the second reason is that workers opt out of auto escalation at a surprising high rate, as high as 60 percent at the first year and even higher thereafter, the authors found. By the third year, only 29 percent of workers were in the default contribution rates set by auto escalation.

“The presumption has been that if you make auto escalation the default, 90 percent of people would accept [it],” Choi said.

The other two reasons have to do with leaving jobs. People often quit before the matching contributions provided by employers are fully vested, which has the net effect of reducing their retirement savings. Further, 42 percent of 401(k) balances are taken as cash outs when people leave one employer.

“Ultimately, it's all about leakage. The automatic savings structures have in my opinion done a magnificent job of getting more individuals to save (and invest - though not part of this study) sooner and better than ever before,” said Nevin Adams, former chief content officer at the American Retirement Association, in an email. “The issues highlighted in the study are about what happens when transitioning from one employment situation to another. Like that moving sidewalk that ends when you're not looking for it, stumbles can result.”

It could help to make it harder to cash out when leaving a job, Choi said.

However, limited access to 401(k) assets before retirement has also been considered a deterrent for participation, especially for lower-income households where the priorities of emergency savings and retirement plan contributions are in conflict. Some account assets can be accessed early without penalty in certain cases, such as for families affected by natural disasters or facing hardships.

Employers are also starting to offer emergency savings accounts alongside 401(k)s. Whether that helps in reducing 401(k) cash outs is unknown, given its newness and a lack of research on the topic, Choi said.

Another way the US could help fix the retirement system would be to follow examples by Singapore or Australia, both of which have high mandatory contribution rates for their defined-contribution system, he noted.

That could be difficult to accomplish politically, at least for some time, he said. But even more so would be expanding Social Security, another way of helping improve retirement savings, he said.

Another way of helping to fix the savings system would be to require automatic contribution or escalation levels based on something other than time of employment, he said. For example, a 45 year old with a new job would likely be contributing far too little if automatically enrolled at a 3 percent contribution – a standard rate that many employer use for their plans.

“You change companies, now you’re back at square one,” Choi said. “You could think about creating some kind of system where the default contribution rate doesn’t depend on how long you’ve been with the company. It could be age-based.”