Brigitte Madrian Testifies to the Senate Committee on Finance

September 17, 2014

Remarks by Brigitte Madrian, Aetna Professor of Public Policy and Corporate Management prepared for the hearing, “Retirement Savings 2.0: Updating Savings Policy for the Modern Economy” for the Senate Committee on Finance - Tuesday, September 16, 2014.

Thank you for the opportunity to speak to you today and share my thoughts on how we can strengthen America’s retirement savings system. By way of background, I am the Aetna Professor of Public Policy and Corporate Management at the Harvard University John F. Kennedy School of Government. I have spent the past 15 years studying individual savings behavior and the policy interventions and plan design features that impact savings outcomes.
Public policy has historically promoted private saving for retirement using financial incentives. In the United States, the primary inducement to save is the exemption of retirement savings plan contributions (up to a limit) from taxable income. The Joint Committee on Taxation places the magnitude of this tax expenditure in 2014 at $127.2 billion annually (Joint Comm. Tax. 2013). Lower-income taxpayers are also eligible for the Saver’s Credit, as a further enticement to save. In addition, public policy encourages employers who sponsor retirement savings plans to provide their own financial inducements for employees to save, namely the provision of an employer match.
A large body of literature has examined the responsiveness of savings to financial incentives. A rather consistent finding from this literature is that the behavioral response to changes in incentive to save is not particularly large. In a recent paper, I surveyed the academic literature on the impact of one kind of financial incentive, matching, on savings plan participation and contributions. The studies using the most credible empirical methods find strikingly similar results in a variety of different contexts using a variety of different data sources: A matching contribution of 25% increases savings plan participation by roughly 5 percentage points. This is a modest effect at best.
Conditional on participating in a savings plan, financial incentives can impact how much individuals save. But this effect does not come from the magnitude of the financial incentive so much as from the fact that at some point the incentive expires. For example, in many 401(k) plans, the employer provides a match, but only up to a certain fraction of pay, say 6%. The Saver’s Credit gives eligible low and moderate income households a financial incentive to save for retirement, but only for the first $2000 contributed to an IRA or workplace savings plan. When financial incentives to save are limited to savings below a certain threshold, this threshold becomes a focal point as individual decide how much to save. For example, data from 401(k) plans shows that savings plan participants overwhelmingly choose contribution rates that are either multiples of 5 (5%, 10%, 15%) or the match threshold. This finding suggests that the match threshold may be a much more important parameter in a matching scheme than the match rate.
The relatively small impact of financial incentives on savings plan participation suggests that a failure to save is not primarily the result of inadequate financial incentives. Rather, there are other barriers to saving not addressed by traditional policy solutions. The literature on behavioral economics and savings outcomes points to a myriad of frictions that impede successful savings outcomes: procrastination, a lack of financial literacy coupled with the complexity of determining how much to save and how to best to invest for retirement, inattention, and the temptation to spend. In many cases, countering these frictions leads to increases in savings plan participation and asset accumulation that surpass the effects of financial incentives.
Before discussing policy alternatives to financial incentives that are informed by behavioral economics, let me note that from a behavioral economics standpoint, the tax code is particularly ill-suited to generating financial incentives to save. First, the tax code is complicated. It is difficult for the average taxpayer to even assess the financial incentives he or she faces through the tax code. For example, in a research project that I am working on, my coauthors and I have found that most individuals do not accurately understand the tax implications of saving in a Roth vs. a regular 401(k) or IRA. For a low or moderate income taxpayer, assessing the incentives of the Saver’s Credit without the help of a tax professional would likely be a daunting task, indeed, I attempted to do so in preparing these remarks and gave up after 10 minutes on the IRS website! Second, individuals are more responsive to immediate than to delayed financial incentives, but many of the financial incentives to save that operate through the tax code are delayed. The benefits of tax deferred compounding are delayed, as are the benefits of tax deductions or credits that aren’t processed through payroll deduction to reduce tax withholding throughout the year. Ironically, what could perhaps be a very effective financial incentive to encourage individuals to enroll in a workplace savings plan—a small but immediate financial reward—is actually not allowed in 401(k) or 403(b) plans under current law.
If financial incentives are not a savings panacea, what is? By far the most effective method to increase savings plan participation is automatic enrollment. Savings plan participation rates are substantially higher when the default is enrollment in the savings plan (that is, individuals must opt out if they prefer not to save) than it is when individuals must take action to participate in the savings plan. The impact of automatic enrollment on participation rates can be sizable and is greatest for groups with the lowest savings rates initially: younger and lower- income workers. The Pension Protection Act of 2006 has successfully encouraged a rapid expansion in the number of employers using automatic enrollment in their savings plans, and this has resulted in an increase savings plan participation rates.

The success of automatic enrollment in employer-sponsored savings plans results from two factors: (a) individuals recognize the need for retirement income above and beyond what they will get from Social Security and therefore want to save, and (b) automatic enrollment simplifies what individuals already want to do. Other initiatives that simplify the savings process have also been shown to increase savings plan participation substantially, although not to the same degree as automatic enrollment. For example, “Quick Enrollment” tools that provide individuals with a pre-selected contribution rate and asset allocation bundle that they can easily opt into have been shown to increase participation rates by 10-20 percentage points.
Expanding the reach of automatic enrollment is the most promising policy step we can take to increase the fraction of Americans who are saving for retirement. This means continuing to increase the number of employers with savings plans who use automatic enrollment, increasing the number of employers who offer a savings plan to their employees by providing a simple and low cost way for small employers to offer a savings plan (only half of workers are in a job where they are even eligible to participate in a savings plan), and providing simple savings alternatives for individuals who are self employed or whose employers do not and are unlikely to ever sponsor a savings plan. Policy initiatives that support these measures include auto-IRA proposals and legislation to facilitate the creation of multiple employer plans with limited fiduciary liability.
Paradoxically, we have a savings system that, in the absence of automatic enrollment, makes saving complicated while at the same time making it very easy for individuals to tap into their retirement savings well before retirement. Another policy response to encourage retirement wealth accumulation is to reduce the leakage from our retirement savings system by discouraging pre-retirement distributions. A sizeable fraction of individuals completely cash out their retirement savings plan balances when they change jobs, and many others roll their money into IRAs where fees are often higher than in employer sponsored plans. Measures to reduce the impact of leakage include facilitating 401(k) loan repayments even after individuals have left an employer where they had taken out a loan, limiting the amount that individuals can withdraw before retirement, and increasing the penalties on pre-retirement withdrawals.
In conclusion, the lessons from the behavioral economics research are clear: if you want individuals to save, make it easy. If you want individuals to save more, make it easy. If you want employers to help their workers save, make it easy. And if you want individuals to spend less, make it hard!


Joint Comm. Tax. 2013. Estimates of federal tax expenditures for fiscal years 2012–2017. Rep., Joint Comm. Tax, Washington, DC
Madrian BC. 2013. Matching contributions and savings outcomes: a behavioral economics perspective. In Matching Contributions for Pensions: A Review of International Experience, ed. R Hinz, R Holzman, D Tuesta, N Takayama, pp. 289–310. Washington, DC: World Bank
Madrian BC. 2014. Applying insights from behavioral economics to policy design. Annual Review of Economics 6 (2014): 663-688.
Vanguard (2014). How America Saves: 2013.

Brigitte Madrian, Aetna Professor of Public Policy and Corporate Management

Brigitte Madrian, Aetna Professor of Public Policy and Corporate Management

Photo Credit: Martha Stewart


John F. Kennedy School of Government 79 John F. Kennedy Street
Cambridge, MA 02138
617-495-1100 Get Directions Visit Contact Page