Auto-Enrollment Could Work Much Better

New Vanguard report confirms auto-enrollment leads to lower saving rates.

Each year Vanguard publishes “How America Saves,” a compilation of an enormous amount of data on the 401(k) plans that the company administers.  Because I have been concerned about how auto-enrollment has evolved, I turned to that section in this year’s report.

Some background on auto-enrollment.  When 401(k) plans first came on the scene, participation rates in the plans were very low.  Research in the field of behavioral economics concluded that many people lack the long-term planning skills required to defer gratification involved in saving today for consumption in the future; that, when faced with a difficult decision, they are racked by inertia and tend to procrastinate; and that they generally take the easiest route – the default – not to sign up for the plan.  A series of studies also showed that changing the default from opting in to opting out vastly increased participation, and that even after three or four years the great majority of those automatically enrolled continued participating. 

In 1998, the government changed the rules to allow employers to require workers to opt out rather than opt in.  And in the Pension Protection Act of 2006, Congress, recognizing the power of automatic enrollment, created a safe harbor and made other changes to encourage its widespread adoption.  In the wake of the legislation, the share of plans with auto-enrollment increased substantially at first but, during the past several years, growth has been much slower.  Today, about half of plans have adopted auto-enrollment, according to data from the Plan Sponsor Council of America (PSCA).

One problem with auto-enrollment is that the same inertia that makes the approach effective for participation can lock people into low levels of contributions.  The typical default contribution rate is only 3 percent, and left on their own, people tend to stay at this low level of contribution.  To combat this problem, the 2006 legislation also tried to encourage auto-escalation – that is, increasing a participant’s contribution rate by at least one percentage point annually up to 6 percent of compensation   Employers also have the option of increasing the default further, up to 10 percent.  Unfortunately, the PSCA data show that only 35 percent of plans with auto-enrollment have auto-escalation in the default contribution rate, so many who are enrolled at low contribution rates remain at those levels. 

Back to the Vanguard data.  Vanguard reports that employees subject to auto-enrollment have a participation rate of 89 percent compared to only 61 percent for employees hired under plans with voluntary participation.  That’s the good side of auto-enrollment. 

The dark side of auto-enrollment shows up in the contribution rates.  The elective employee deferral rate has declined steadily from 7.3 percent in 2006 to 6.9 percent in 2014.  Auto-enrollment without auto-escalation in the default contribution rate likely does more harm than good.  The time has come to fix the problem.  Congress should change the law to require all 401(k) sponsors to adopt auto-enrollment for their entire workforce, with deferral rates set at a meaningful initial level – that is, higher than 3 percent – and to adopt auto-escalation in the deferral rate.  For most workers, 401(k)s are their only supplement to Social Security.  These plans need to work better, and this modest legislative change would make a big difference.