By Mark Miller
CHICAGO, March 7(Reuters) - Are employers stingier with their 401(k) matching contributions when they automate the enrollment process?
Auto-enrollment has been a hot trend in the 401(k) world over the last few years. The idea is to nudge workers to save who might not otherwise do so through a simple behavioral change: action needs to be taken to opt out, instead of opt in. A study by behavioral economists Shlomo Benartzi and Richard H. Thaler, released today in Science magazine, says that 56 percent of companies offer some kind of auto-enrollment program, up from 19 in 2005.
Auto-enrollment boosts participation rates sharply, but there has also been a downside. The average initial default contribution rate in auto-enroll plans is just 3 percent - far too low to generate meaningful account balances. Many workers contributing at that level also are leaving money on the table when the employer's matching contribution is higher.
A growing number of plans are adding auto-escalation features that typically bump up workers' contribution rate by a percentage point annually. The study in Science also found that 51 percent of employers now offer that feature, up from 9 percent in 2005.
The result of these programs is supposed to be increased savings rates. However, a new study points to a another problem: Some employers who adopt auto-enrollment offer smaller matching contributions. The report, by the Urban Institute and the Center for Retirement Research at Boston College, found that matching contributions in auto-enrollment plans were averaged 3.2 percent of workers' salary, compared with 3.5 percent for voluntary-enroll plans.
Most industry players and analysts dispute that finding - emphatically. But the best evidence suggests that two trends actually are developing: Smaller retirement plans are stingier with their matching contributions, while big plans are more generous. That will make it more difficult for workers in smaller plans to accumulate retirement savings - and roughly one-third of U.S. 401(k) savers are enrolled in smaller plans (2,500 or fewer participants), according to data from the Employee Benefit Research Institute (EBRI).
The Urban Institute study is based on data from the National Compensation Survey of the U.S. Bureau of Labor Statistics (BLS). It includes 14,000 private sector retirement plan records and it is reflective of the national mix of plans in terms of industry and plan size. The survey records allowed the researchers to compare not only benefit costs, but also wages and other benefit types.
"Nothing comes for free," says Nadia Karamcheva, research associate in the Urban Institute's Income and Policy Benefits Center and co-author of the study. "We expected to find reduced wages or other benefits, but the only thing we found was lower match rates."
Meanwhile, a 2010 study of 1,000 large plans by EBRI and Aon Hewitt, the employee benefits consulting firm, found that firms actually boosted their match rates as a percent of salary when they adopted auto-enrollment.
Large companies often add auto-enrollment when they freeze or close a defined benefit pension program, says Alison Borland, vice president of retirement solutions and strategies at Aon Hewitt.
"They feel responsibility to keep employees on track for retirement, so many increase their matching contribution and add auto-enrollment at same time," she says.
Vanguard also disputes the Urban Institute findings.
"We don't see plans that are adopting auto enroll reducing their employer match," says Jean Young, a senior research analyst for the Vanguard Center for Retirement Research. "And we haven't seen any general evidence that the value of the employer match is declining."
Fidelity Investments also says plan sponsors are not cutting matching contributions to cover auto-enrollment participation costs. But a break-out of plan data that Fidelity provided does show a significant gap in contribution rates. In 2012, Fidelity-administered plans that use auto-enrollment had an average employer contribution rate of 3 percent of salary, compared with 3.5 percent for plans without auto-enrollment.
But the match rates vary depending on plan size. For example, among plans with 100 to 250 participants, auto-enroll plan contributions averaged 2.6 percent of salary, compared with 3.3 percent at plans without auto enrollment. At very large plans, the opposite was true: auto-enroll plans with 25,000 to 50,000 participants had a 4.6 percent average employer contribution, compared with 4 percent for non-auto enroll plans.
The Urban Institute study reflects similar findings: Auto-enroll matching contribution rates of 2.8 percent at plans with fewer than 500 workers, compared with 4.1 percent in plans with 1,000 to 2,500 workers.
These differences may sound small, but they make a big difference in total contribution rates and in retirement outcomes when compounded over many years of saving. Higher participation rates are welcome news, but the questions about match rates and auto-enrollment will bear watching in the years ahead.