Plan Sponsors Should Be Encouraged to Use Auto Features

The use of automatic enrollment is expanding, a survey from the Defined Contribution Institutional Investment Association (DCIIA) finds.

Plan sponsors of the larger plans surveyed (greater than $200 million in assets – 185 plans) continue to adopt automatic enrollment, with 62% of survey respondents indicating that they utilize this feature, compared to 56% in 2012 and just 44% in 2010. Among plans with $50 million to $200 million in assets, 59% use auto enrollment, with $5 million to $50 million in assets, 38% use it, and with less than $5 million in assets, 24% do.

During a webinar about the research, Catherine Peterson, global head of insights programs at J.P. Morgan Asset Management, said the vast majority of plans say the participant opt-out rate for auto enrollment is less than 10%; around 70% report it’s less than 5%.

A 3% default deferral rate is used by 37% of plans, down from 55% in 2010 and 47% in 2012. Meanwhile, the proportion of plans using a default rate of 6% or more increased to 27% in 2014, up from 20% in 2010 and 24% in 2012. Peterson said this is one area plan sponsors need to be encouraged to do—move to a higher default rate.

Since 2010, the level of automatic deferral escalation used by large plans has leveled off (46% in 2010, and 48% in both 2012 and 2014). “There is a risk of having participants defaulted and remaining at low contribution rates,” Peterson noted.

Use of plan re-enrollment, whereby participants’ assets are reset into the plan’s default investment option unless the participant opts out, has increased from 6% in 2010 to 13% in 2012 and 19% in 2014, but it remains an underused practice to improve participant asset allocation (see also “Re-enrollments Remain a Poorly Leveraged Plan Booster”).

“We haven’t seen nearly enough use of re-enrollments, considering the huge benefit of getting people into a proper investment allocation over time,” Peterson said. “Other research shows individuals in target-date funds (TDFs) have better returns than do-it-yourself investors over time.”

The DCIIA Plan Sponsor Survey on Auto Features reveals the positive impact of using automatic plan features. A before-and-after picture of participation rates shows nearly 50% of plans had 75% or less participation before using auto enrollment. After auto enrollment, only 20% had 75% or less participation rates. Plans with greater than 90% participation moved from 18% to 45%. “Many plans still only use auto enrollment for new hires, so it will take more effort to get more plans above 90% participation,” noted Lori Lucas, executive vice president and defined contribution practice leader at Callan Associates in Chicago.

A picture of average contribution rates before and after using auto escalation shows 66% of plan had average deferral rates of 6% or less before using the feature; 41% had average deferral rates of 6% or less after using it. Lucas noted that the survey showed the economic theory of cognitive dissonance among plan sponsors; 50% of employers say the optimal savings rate is between 5% and 10%, 34% say it is between 11% and 15%, yet 66% of plan sponsors surveyed are defaulting participants at a less than 5% deferral rate.

The reported reasons for not closing the savings gap: For plans with $200 million to more than $1 billion in assets, it was that they already felt they were doing what they should; for smaller plans, it was that they felt employees preferred their wages be paid to them.

Are the larger plans really doing enough to close the savings gap? Lucas feels they could be doing more.

The primary barrier to adopting auto enrollment cited by plans in the DCIIA survey is the additional cost from matching contributions (30% of large plans cited this). Fifteen percent think their plan participation is already high enough, 22% say it’s not necessary because their defined contribution (DC) plan is supplemental to a defined benefit (DB) plan, and 15% feel auto enrollment is too paternalistic.

By plan size, the barriers can be quite different, noted Joshua Dietch, managing director at Chatham Partners. That it is too costly is more likely the barrier to adopting auto enrollment for plans with $200 million or more in assets. Thirty percent of plans with less than $50 million say it is too paternalistic. More than one-quarter of small plans have not even considered it; “maybe they didn’t even know it was an available option,” Dietch speculated.

For large plans, the feeling that auto escalation is too costly from a match perspective has dropped. Dietch thinks this may be because of the trend of using a stretched match formula. For small plans, 31% have never considered auto escalation, and they are two times more likely than large plans to think participants will complain. Fifteen percent of small plans feel their average deferral rate is already high enough.

As for barriers to engaging in a reenrollment, 18% of plan sponsors said they are comfortable with participant allocation, but Dietch noted this has dropped from 36% in 2010 and 35% in 2012. Thirteen percent feel there is too much risk in doing a reenrollment.

Catherine Collinson, president of the Transamerica Center for Retirement Studies, shared with webinar attendees the DCIIA’s recommendations for automatic plan features:

• Automatically enroll all new and existing employees;
• Set the initial default deferral to no less than 6%;
• Use auto escalation with a default increase of 1% to 2% up to a maximum of 15%;
• Investigate the use of a stretched match formula to encourage higher rates of deferral in a cost-effective way. Collinson shared the example of switching from a match of 100% up to 3% of deferrals to 50% of up to 6%, and she warned that plan sponsors should not stretch the match so far that participants will think they can’t save enough;
• Consider reenrollment to help participants invest their retirement savings in a risk-appropriate manner using the default investment option; and
• Providers should offer decision tools that can help plan sponsors optimize these benefits.

Collinson noted that in the survey, plan sponsors expressed a desire for more information to help in plan design decisions. There is a reported lack of understanding among survey respondents of the risks and unintended consequences of implementing or optimizing automatic plan features. Very few plans reported that they modeled potential outcomes when considering implementing or modifying automatic plan features. There is a perceived lack of effective plan sponsor tools to analyze alternatives appropriate for each plan’s unique characteristics and objectives.

The survey report may be downloaded from the DCIIA’s website.