Summary
• RCS data used to gauge impact of auto-escalation: This report uses data from the 2007 Retirement Confidence Survey (RCS), fielded several months after the enactment of Pension Protection Act of 2006 (PPA), to ascertain how high workers are likely to allow their default 401(k) contributions to go in plans with an automatic escalation feature. The result is a first approximation for the expected impact of automatic escalation under the PPA safe harbors for a number of different assumptions about worker and employer reactions.
• Previous EBRI research: EBRI has done extensive modeling of the retirement income prospects for future generations of retirees, with results ranging from very bleak for substantial portions of the U.S. population to fairly positive for 401(k) participants with continuous coverage throughout their working careers.
• Replacement rate estimates for automatic enrollment: Assuming that all 401(k) plan sponsors would adopt automatic enrollment immediately (in 2005), the median replacement rates for the lowest-income quartile increased to 37 percent (from the 23 per-cent baseline) even under the conservative assumptions of a 3 percent default contribution rate and a money market default investment. When the default contribution rate was increased to 6 percent and the default investment was changed to a life-cycle fund, the median replacement rate for this group increased further to 52 percent.
• Auto-escalation is likely to result in major improvement for low-income workers: Results from the 2007 RCS suggest that the introduction of automatic escalation will result in a significant increase in 401(k) accumulations—especially for low-income workers—compared with estimates previously determined for automatic enrollment. Under the assumptions and scenarios modeled in this article, the automatic escalation feature is likely to increase overall 401(k) accumulations between 11–28 percent for participants in the lowest-income quartile, and between 5–12 percent for those in the highest-income quartile.