The effect of pension design on employer costs and employee retirement choices: Evidence from Oregon

We use administrative data from Oregon's Public Employees Retirement System (PERS) to study the effect of pension design on employer costs and employee retirement-timing decisions. During our 1990-2003 sample period, PERS calculates each member's retirement benefit using up to three different formulas (defined benefit (DB), defined contribution (DC), and a combination of DB and DC), and PERS pays the maximum benefit for which the member is eligible. We show that this “maximum benefit” calculation results in average ex post retirement benefits that are 54% higher than if they had been calculated using only the DB formula and that employees receiving DC benefits are significantly more likely than employees receiving DB benefits to retire before the plan's normal retirement age. Monte Carlo simulations verify that the higher costs could have been predicted at the start of our sample period. Exploiting exogenous plan changes, we show that employees respond to within-year variation in their retirement incentives and, consistent with peer effects, that they respond more strongly to these incentives when more of their coworkers face similar incentives. Finally, consistent with the emerging literature on financial mistakes by households, we show that a small but noteworthy fraction of retirees would have benefited from shifting their retirements by as little as one month.