Alicia Munnell firmly believes there’s a retirement crisis (see “Exclusive Interview with Alicia Munnell: Auto-Enrollment Without Auto-Escalation Reduces Deferral Rates,”FiduciaryNews.com, February 17, 2016).
She understands employees aren’t saving enough.
She’s also frank enough to admit that one of the causes is the behavioral finance researcher’s favorite nudge – auto-enrollment.
Auto-enrollment has been a boon to all those seeking to increase employee participation in their company’s 401(k) plans. And this is true. Plans that offer auto-escalation have witnessed a marked increase in the percentage of employees saving for retirement.
You’d think that’s a good thing.
Alicia Munnell knows it’s not. It’s not just a gut instinct sort of think, either. She’s studied the numbers.
The fact is, we’ve made 401k plans such a no-brainer that employees aren’t even thinking about them anymore. No. Like so many lemmings, they’re merely following the default decisions of the plan sponsor (actually, the professionals that advise the plan sponsor).
To say they “set it and forget it” gives them too much credit. Employees merely forget it. The plan sponsor sets it.
Therein in lies the problem, according to Munnell. Employees are told the plan works on auto-pilot and they don’t have to do anything. It’s easy. No effort. Just forget it.
Oh, it’s not like they’re forced to save. They can always opt out. But that requires effort. And behavioral researchers insist that “effort” is the problem. And those academics are correct. Employees don’t want to exert the effort. That’s why they didn’t save when saving required effort. That’s also why they don’t stop saving when stopping requires effort.
Unfortunately, there’s a dark side to this Faustian Bargain. Sure, employees don’t want to exert the effort to stop the saving. But they also don’t want to spend the energy to increase their savings rate.
That’s a problem. Most auto-escalation policies put the tab at 3 percent of the employee’s salary. That’s barely enough to put a dent into what’s needed to meet the employee’s retirement goal.
Munnell suggests we start this deferral rate at 6 percent, but she realizes even this isn’t enough. She knows that you need to couple on an auto-escalation policy, lest auto-enrollment go wasted.
Indeed, if done correctly, the marriage of auto-enrollment to auto-escalation can effortlessly (there’s that word, or at least a derivative of it) get to the level of saving most retirement planners suggest (i.e., between 15 percent and 20 percent).
At that level of saving, for a sustained period, employees can grow their defined contribution plan account to a size that can replace the need for Social Security or any other pension.
We know this to be true because we’ve already seen it happen to disciplined savers. These people didn’t need auto-enrollment. They didn’t need auto-escalation. They had the internal strength of commitment to do it all on their own, without the guidance of some paternalistic hand.
These birds, though, are too few and far between to build a healthy public retirement policy on.
That’s why we need auto-enrollment and auto-escalation. These two levers can exist without much interference in an employee’s natural budgeting process.
And, as a safety governor, if employees cannot adjust spending to account for auto-enrollment and auto-escalation, they always have the option to opt out.
Actually, now that I’m thinking about it, auto-enrollment is just like a payroll tax. And auto-escalation is just like politicians periodically raising that payroll tax.
Payroll taxes are how we currently fund Social Security (to the tune of 15 percent of an employee’s salary). Maybe a lot of people complain about payroll taxes, but, in the end, they all have learned to live with them (for the simple reason there is no opt out provision).
Hmm, this might sound like a crazy idea, but if we wanted to “fix” an unsolvable dilemma like the looming bankruptcy of Social Security, getting everyone acclimated to auto-enrollment and auto-escalation in their defined contribution plans is exactly how we would start.