The Retirement Savings Problem

Published June 19, 2017

Americans seem to have a problem saving for retirement.  This problem has implications for our own lives and for employer-provided pension plans.

Over 60 percent of American workers currently have employer-provided pensions.  The self-employed and workers without pensions can use Individual Retirement Accounts (IRAs).  Both pensions and IRAs defer taxes on contributions and investment earnings until retirement.

Employer-provided pensions are of two main forms, defined benefit (DB) and defined contribution (DC) plans.  DB plans promise a given retirement income, typically based on lifetime earnings, funded by employee contributions and earnings on pension assets.  DC plans deposit contributions into an employee account.  The money is fully portable, and can even be withdrawn before retirement.  Retirement income depends on the accumulated contributions and investment earnings.

Both DC and DB plans each have advantages.  DC plans give more control and responsibility to the individual, who must choose to contribute money (often with an employer match) and resist the temptation to withdraw money.  But under DC plans, employees bear what is known as the investment risk, the potential that investment returns will be less than expected, resulting in insufficient funds for retirement.  Employers bear this risk under DB plans.

Both private and public sector DB plans have been chronically underfunded.  Employers must make additional contributions to shore up an underfunded pension.  This places a business at a disadvantage relative to competitors without such legacy costs, resulting in numerous corporate bankruptcies.  Public sector plans have survived, but taxpayers may still have to make good on the commitments in the future.

The promises of current under-funded plans should be honored, but we should try to prevent a recurrence.  DC plans offer one solution, and between 1992 and 2010 they increased from around one third to 62% of employer-provided pensions.  Although DC plans allow easy estimation and limitation of the employer’s cost, I think they may exacerbate the retirement savings problem.

Research by the Center for Retirement Research at Boston College quantifies the savings problem.  The typical or median asset level of employees aged 51 to 56 in DC plans in 2010 was $98,000, which projects to about $11,000 in annual retirement income.  Without knowing the exact plans of persons behind data, it is difficult to conclusively say this is too little savings.  For example, people in this age range may have chosen to pay for their children’s college and then be planning to begin saving a lot.  This figure, however, is still concerning.

What will happen when thousands of DC plan retirees use up their pensions within the first few years of retirement?  I suspect that our government would end up providing some form of assistance.  DB plans are more paternalistic, and would protect people who fail to save adequately with a DC plan.

The retirement savings problem also likely contributes considerably to underfunded government pension plans, as Bloomberg columnist Megan McArdle has argued.  People who think that their own under-funded 401k’s or IRA’s are adequate are likely also likely to believe that meager contributions can fund public pensions.  Furthermore, underfunded DB plans and inadequate contributions to DC plans both leave taxpayers potentially on the hook to provide for the retirement of people who did not save as much as they should have.

Behavioral economics combines research from economics and psychology and has made great strides in identifying frequent mistakes, like not saving enough for retirement.  Behavioral economics also identifies a surprising antidote to many poor decisions, namely default rules.  Default rules specify the choice applied when someone fails to make a choice.  For example, perhaps DC plan employees could be enrolled by default at the highest allowed annual contribution, while retaining the freedom to reduce their contribution.

Freedom has led to the thriving of people and societies, despite our mistakes.  As we learn more about our common mistakes, we will hopefully devise clever ways to capture the personal and societal benefits of freedom while mitigating the harm from our mistakes.  Defined contribution pensions, unfortunately, do not accomplish this goal.