Auto-IRA May not Help as Many as Government Hopes

Estimates by researchers for the Center for Retirement Research at Boston College indicate that if predictions are made only using data on the workers who currently have access to a plan, the policy effects would be greatly overestimated. If matched 401(k)-type plans were provided to all full-time employees, 75% of individuals would participate; however, controlling for selec­tion, the predicted participation rate drops to 58%.  

For low-income workers, the study found 57% would participate if all were offered a 401(k)-type plan, but controlling for selection, the figure drops to 34%.  

In 2008, the Obama campaign proposed a “Plan to Strengthen Retirement Security” which included a proposal to require employ­ers with 10 or more employees to automatically enroll their employees in Individual Retirement Accounts (IRAs). As a default, 3% of each worker’s earn­ings would be invested in a low-risk portfolio, but workers could choose to change the defaults or opt out of the plan.  

The researchers noted that the purpose of the Auto-IRA is to increase the pension participation of all workers, but particularly low-income workers; yet, it is unclear how many of these workers would participate. The researchers study using the Survey of Income and Program Participa­tion indicates that 60% of low-income workers currently eligible for voluntary 401(k) plans choose to participate in those plans, and the study report said other research suggests that when workers are automati­cally enrolled in a 401(k) plan, they are even more likely to participate.   

However, the findings are based on individuals who have a 401(k) available to them. The researchers reason that it seems likely that these individuals may have sought out jobs offering such plans with an inten­tion of participating, and if so, workers who did not seek employment offering pensions may be less likely to participate in the IRA plan, limiting the program’s success in expanding pension participation.