Framing Effects in Retirement Savings

June 2008

This paper examines how supplemental retirement savings decisions of higher education faculty are influenced by the generosity and structure of an individual’s primary retirement plan. Standard economic models of saving predict than an extra dollar contributed to an employee’s primary defined contribution account will reduced the supplemental savings of that employee by one dollar, regardless of the source (employer or employee) of the pension contribution. A behavioral perspective suggests that people may treat them differently, however. In particular, suppose that people assign the employer’s pension contributions to one mental account, and wages to another. Since employee pension contributions and supplemental savings are both withdrawn from the “wages” account, supplemental savings will be offset dollar-for-dollar by the value of employee pension contributions. With imperfect fungibility across accounts, however, supplemental savings will be less sensitive to employer contributions. Empirical analysis finds that supplemental savings rates depend on how compensation is labeled. In particular, supplemental savings are significantly lower when a larger fraction of the primary pension contribution appears as a salary deduction; supplementary savings are reduced by 60-90 cents per dollar of employee contributions to the primary pension, but only by one-half as much per dollar of employer contributions. The difference in these responses suggests that mental accounting (or differential inattention) leads to substantial differences in realized supplementary savings amounts across people with similar compensation streams but different pension formulas. Consequently, two faculty members with the same total compensation and the same total contribution rates to their primary pension will reach retirement age with substantially different amounts of supplemental saving, depending on the share of primary pension contributions labeled as an employee contribution.