PLANSPONSOR - September 2017 - 28

Retirement Plan Applications
Data analytics can help employers make retirement plan-design
decisions in many areas. These include:
* Shifting automatic enrollment features. Data analytics
" can show a plan sponsor a projection of which employee groups
a change in 'auto'-features will help-not just [a projection] for
the company as a whole, but for different demographic groups, "
says Jonathan Price, a Segal vice president in New York City. For
example, an employer that is thinking of implementing autoescalation
may want to know how that would affect employees in
some age groups more than others.
Data analysis also can project how much a plan-design
change would affect specific work force segments. " Let's say
a plan has 85% participation now and the employer's target is
90%, " Price says. " An employer can look at: Working with the
tools we have, how can we get to 90%? "
Other considerations for employers beyond data-analytics
results include the corporate culture, long-term work force planning,
and the human resources (HR) and finance departments'
goals for the plan-particularly in light of overall compensation/
spending, he says.
Employers also can utilize data analytics to assess whether
to shift the timing for auto-enrollment, Dill notes. Say, an analysis
for an employer with high turnover shows that people leave
mostly in the first three years of employment and that employees
passing that threshold usually stay on, long term. " Some of those
organizations are considering the idea of not defaulting new
employees into the plan right away: The employer may decide to
do auto-enrollment after three years, " she says. " Rather than pay
a match to people who won't stick around, some have decided to
wait until they know it's very unlikely the employees will turn
over. " A worker may still opt in to the plan, but automatic enrollment
would be held until later.
* Changing the match formula. Studying the data on participation
levels by demographic segment helps an employer see
how its current 401(k) match formula influences participant
contributions. " The data usually shows that the match rate
sends a strong signal to employees about how much they should
contribute, " Dill says. " Whether a match formula is 50 cents on
the dollar or dollar for dollar isn't the key factor in how much
[they] contribute. The much stronger signal is where the match
ends. " So, matching 50% on 6%-or 30% on 10%-signals that
employees need to contribute more than a 100% on 3% match.
Analysis often finds, too, that employees max out their contribution
at the employer's match level, says Laurie Vance, senior vice
president in Fidelity Investments' benefits consulting group in
Boston. " So now we see companies looking at, 'How many of our
employees are getting stuck at the match level? And does it make
sense to change the match formula?' " Fidelity uses predictive
modeling to forecast the likely impact of a match-formula change
on participants' contributions and on the employer's costs.
Data analytics also assists employers in understanding how
a match adjustment likely will affect different employee groups,
28 PLANSPONSOR.com September 2017
including their most-valued employees. " We help an employer
think through, which talent at their organization is the most
difficult to recruit and retain? " says Marc Howell, vice president
of custom retirement solutions at Prudential Retirement in
Philadelphia.
For example, a hospital may face the most challenges
recruiting and retaining nurses. " So we say: Here is the money
[you] have available. How can you best structure your retirement
benefit to recruit and retain those employees? " That may mean
varying an employer's match formula over different job classifications
and geographic regions, within discrimination-testing
requirements. " So for that health care provider, some groups may
get a more competitive match than do other groups, " he says.
" We can show them, 'Yes,
there is a cost to doing
re-enrollment, but there is also
a cost to having employees
approaching typical retirement
age who can't retire.' "
* Deciding on re-enrollment. Several factors, including an
employer's budgeting realities, come into play when thinking
about re-enrolling nonparticipating employees. " The biggest
stumbling block is usually cost, " Vance says. " So we look at,
what is the total cost to the employer for the next year if everyone
sticks with the re-enrollment? Then we can model forward, what
will be the participation impact and the cost impact if the plan
re-enrolls every eligible employee for the next five years? "
Prudential can do predictive modeling on retirement
patterns of an employer's work force based on a multitude of
factors, which allows the employer to see whether its employees
will likely retire on time, Howell says. It also can give an employer
analytics that estimate both the ongoing cost of performing
re-enrollments and the longer-term costs it faces as employees
cannot retire. " We can show them, 'Yes, there is a cost to doing
re-enrollment, but there is also a cost to having employees
approaching typical retirement age who can't retire,' " he says.
A decision to re-enroll nonparticipating employees every
year improves employees' savings, but, for employers, the decision
goes beyond analyzing numbers. " It's also a philosophical
decision, " Dill says. " It's more for an organization that might be
very paternalistically oriented toward its employees, so it feels
OK saying, 'We're going to make it hard to not participate.' But
that is a slippery slope: Some employees might not be contributing
because they can't afford to, and they might end up taking
on more [personal] debt if they are re-enrolled into the plan. As
a sponsor, you might be solving one problem at the expense of
http://www.PLANSPONSOR.com
PLANSPONSOR - September 2017

Table of Contents for the Digital Edition of PLANSPONSOR - September 2017
