Auto-Pilot Retirement - Ready for Takeoff?


In late October, the Government Accountability Office released a report on auto-pilot retirement practices and proposals. The report reviews data on 401(k) plans that have adopted automatic enrollment, and discusses the Obama administration’s budget proposal to mandate payroll deduction IRAs for firms with ten or more employees that lack other plan coverage.

The report found generally that automatic enrollment tends to increase the proportion of employees participating in 401(k)s but does not necessarily ensure adequate accumulations. It concludes that automatic payroll deduction IRAs would similarly increase coverage, but that cost considerations may make these inappropriate for high-turnover or smaller firms.

The report is particularly timely since new auto-pilot proposals are now being readied for introduction on Capitol Hill as soon as healthcare reform is passed.

But before these proposals take off, policymakers may want to take a hard look at some of the assumptions on which they are based – and their long-term implications.

As the GAO report notes, auto-pilot initiatives trace their origin to the work of behavioral economists. These theorists contend that if you create structures which nudge people into making the “right” decisions, you can achieve desired public policy objectives while preserving individuals’ freedom of choice.

An explanation of this “choice architecture” concept can be found in a book called Nudge: Improving Decisions about Health, Wealth, and Happiness written last year by behavioral economist Richard Thaler and Cass Sunstein, now a top official at the Office of Management and Budget. They call their approach “libertarian paternalism.”

Thaler and Sunstein apply choice architecture to a wide range of situations, such as encouraging children to make healthy food choices in a school cafeteria, organ donation, and choosing among prescription drug programs.

Their arguments and those of other nudge theorists are interesting and in some instances persuasive, but the application of their theory to retirement security deserves closer scrutiny.

The type of nudge they advocate for retirement savings is to default individuals into “good” choices, and then count on inertia to reduce the likelihood that they will opt out.

But there is a key difference between many of the default situations cited by Thaler and Sunstein and retirement security. In these other situations individuals do not lose anything by being nudged into the desired behavior.

For example, if food selections in a cafeteria are structured to encourage a child to select a salad rather than french fries, the child will not be harmed by the selection of a salad. Similarly, if drivers’ license applicants are defaulted into an organ donation program or Medicare recipients are assigned to appropriate prescription drug programs, no negative consequences will result from their inaction.

But when it comes to saving for retirement, some individuals can be hurt by defaulting them into the “right” choice of enrollment in a 401(k) or IRA (and, in some auto-pilot proposals, increasing contributions, adding matches, and making balanced investment decisions). These are workers who cannot easily afford to have money taken out of their paychecks and set aside for retirement.

If these employees are defaulted into a plan, and inertia keeps them there beyond an opt-out window (usually 90 days), the arrangement can turn out to be a poor deal for them. That is because they are more likely than other employees to withdraw their funds before age 59-½. If they do, they will have to pay a 10 percent penalty tax plus regular income taxes on the money they take out of their accounts, thus forfeiting much, or even most, of their investment earnings. In some situations, they might actually wind up with less after taxes than they would have had they never contributed at all.

In many cases, the compulsion to pull out their funds will come from major unanticipated expenses. But creditors owed smaller amounts are also likely to exert pressure since they are sure to know that cash-outs are possible (either hardship withdrawals in a 401(k) or other withdrawals from an auto-IRA).

But far more troubling than the loss of much-needed money in the short-term, is the fact that at the end of their careers, these workers will be made to feel that it is their fault that they don’t have enough money to live on in retirement. After all, they made an affirmative choice not to save.

In a recently published essay for the New America Foundation, “All Power to the Choice Architects: A Liberal Critique of Libertarian Paternalism,” political scientist Alan Wolfe makes a similar point about the application of default options in a retirement context: “policies along these lines are likely to reinforce invidious distinctions between people.” Those who save will be defined as “good” citizens, while those who don’t “will be viewed as harmful” to the system.

It is precisely because retirement is remote, and other financial demands inevitably loom as more immediate and compelling, that all taxpayers are called upon to subsidize 401(k)s and IRAs – and pensions. It is also why significantly increasing retirement savings will require more than a “nudge.”

Others have noted different concerns about auto-pilot retirement proposals. Some worry that the proposals will exclude too many employees. Some fear that firms will drop existing plans (or reduce matching contributions). Still others worry about possibilities of fraud or increased wealth inequality.

Does this mean that auto-pilot arrangements are a bad idea?

Not necessarily. The mandated nudges now under consideration would certainly help those employees who can afford to save but simply don’t “get around” to setting up an IRA or signing up for a 401(k) (or increasing their contributions, or reallocating their investments.) But even these better-off employees would have no assurance that their accumulations will be either safe or sufficient.

In short, auto-pilot arrangements will offer modest incremental improvements for some of today’s workers – but not for others. The far greater challenge will be to develop comprehensive reforms to benefit all of tomorrow’s workers.