How Behavioral Science Ultimately Fails Retirement Savers: A Noble Experiment

• Author: Dana M. Muir
• Date: 01 December 2019
• DOI: http://doi.org/10.1111/ablj.12150
• Published date: 01 December 2019

American Business Law Journal

Volume 56, Issue 4, 707–770, Winter 2019

How Behavioral Science Ultimately

Fails Retirement Savers: A Noble

Experiment

Dana M. Muir*

Behavioral scientists boast that their insights have increased savings in 401(k) plans.

Evidence shows that careful use of default decision settings and nudges can prevent

decisional errors and encourage behavior that aligns with public policy while retaining

individual power of choice. Indeed, even the Swedish National Academy of Sciences

highlighted the effect of his behavioral science work on retirement savings when it

awarded the 2017 Nobel Prize in economics to Professor Richard Thaler. This article

shows, though, that, unlike the three plan default settings attributed to behavioral sci-

ence insights, the default setting that I term automatic retention fails. That failure

means too many savers take affirmative actions to move (rollover) assets from their

401(k) plans to individual retirement accounts (IRAs). Primarily because of rollovers,

IRAs hold more than 11.5% of U.S. household financial assets. However, usually the

decision to roll over is not an optimal choice for the saver. This last mile problem

undermines the goal of the first three default settings to help employees build long-term

retirement savings. This article examines research on what causes default settings to be

slippery, as is the case for automatic retention, instead of sticky, as is the case for the

other 401(k) default settings. It then evaluates three categories of potential interven-

tions to mitigate the popularity of rollovers: aggressive regulation, expansion of fidu-

ciary obligation, and use of incremental impeding altering rules. It concludes that

adoption of incremental impeding altering rules would be both politically feasible and

effective in increasing the stickiness of the automatic retention default setting.

*Robert L. Dixon Collegiate Professor of Business and Arthur Thurnau Professor of Busi-

ness Law, Stephen M. Ross School of Business, University of Michigan. dmuir@umich.edu.

I am grateful to the Ross School of Business for research support and to William Park for

research assistance. I received helpful comments from participants at conferences held at

the University of Connecticut School of Law, University of Illinois College of Law, and the

National University of Ireland, Galway. Special thanks to Professor InaraScott for her care-

ful editorial work. All errors remain my own.

©2019 The Author

American Business Law Journal ©2019 Academy of Legal Studies in Business

707

In his applied work, Thaler demonstrated how nudging—a term he

coined—may help people exercise better self-control when saving for a

pension…

1

INTRODUCTION

Even the Royal Swedish Academy of Sciences has hailed the positive influ-

ence of behavioral science on how individuals save for retirement.

2

Though perhaps the most prominent example, it is only one e xample of

the extent to which advocates of behavioral science have pointed to retire-

ment plans as proof that careful use of default decision settings and

nudges can prevent decisional errors and encourage behavior that aligns

with public policy while retaining individual power of choice.

3

This article

argues that, although using behavioral science to increase retirement sav-

ings is a noble one, one of the default settings too frequently fails to

achieve its intended results.

Theimportanceofsavingforretirementcanhardlybeunder-

estimated. A 2019 projection estimates the current retirement deficit

for middle-to-retirement age U.S. households at $3.83 trillion.

4

This

article focuses on the 401(k) plan, which is the type of retirement plan

1

Press Release, The Royal Swedish Academy of Sciences, Press Release: The Prize in Eco-

nomic Sciences 2017 (Oct. 9, 2017), https://www.nobelprize.org/prizes/economic-sciences/

2017/press-release/ (emphasis in original).

2

See id.

3

E.g., Lauren E. Willis, When Nudges Fail:Slippery Defaults,80U.CHI.L.REV. 1155, 1158

(2013) (“The most celebrated, and most cited, example of a policy default is in the area of

retirement saving.”); see also Rebecca Moore, Survey Says:Least/Most Effective Retirement

Industry Developments in the Last 25 Years,P

LANSPONSOR (Jan. 22, 2018), https://www.

plansponsor.com/survey-says-least-effective-retirement-industry-developments-last-25-years/

(“THE MOST effective [development] for improving retirement outcomes is the adoption

of automatic enrollment for defined contribution (DC) plans (48.9%).”); cf. Ryan Bubb &

Richard H. Pildes, How Behavioral Economics Trims Its Sails and Why, 127 HARV.L.REV. 1593,

1607, 1616–32 (2014) (stating “BLE [behavioral law and economics] lays claim to retire-

ment savings as perhaps its greatest policy-reform success” before explaining specific prob-

lems with BLE’s application to retirement savings).

4

EMP.BENEFIT RES.INST., EBRI PROJECTIONS SHOW IMPROVEMENTS (Mar. 14, 2019), https://www.

ebri.org/docs/default-source/infographics/23_ig-rspm-14mar19.pdf?sfvrsn=9d763f2f_4.

708 Vol. 56 / American Business Law Journal

pri vat e sector employers most frequently offer to their employees.

5

The

defining element of a 401(k) plan is that the plan must permit employees who

are eligible to participate to decide whether to contribute pretax earnings to

their individual accounts.

6

Therein is one of the key reasons that, historically,

401(k) plans failed to provide benefits to many employees—those employees

simply did not choose to contribute.

7

Behavioral science insights made dramatic inroads in addressing the

problem of nonparticipation. Plans that use default settings for auto-

matic enrollment, automatic escalation of savings rates, and automatic

allocation of assets to a diversified investment product—nudges, to use

the term of behavioral science and the Nobel Academy—do cause more

employees to save for their retirement, more money to be saved, and

increase investment diversification.

8

This article shows, however, that

much of the benefit gained by those default settings may be lost in the

last mile because of the failure of a fourth default setting, which I term

automatic retention, to keep savers from moving assets out of 401(k)

plans.

9

As such, this article shows that failure of the automatic retention

default undermines the progress made by structuring plan architec-

turebasedonbehavioral science research.

This article begins in Part I by discussing the success of default set-

tings in 401(k) plans and the relevant behavioral science concepts. Part

II investigates the frequency with which savers actively decide to move

assets out of a 401(k) plan and roll over those assets to individual

5

For an overview of the development of 401(k) plans, see infra notes 13–19 and accompanying

text. Because of the popularity of 401(k) plans, this article concentrates on them although the

risksofIRArolloversfromothertypesofdefinedcontribution (individual account) plans, such

as 403b and 457 plans that are sponsored by government entities, are similar to the risks dis-

cussed here for 401(k) plans. Technically, account holders may in 401(k)-style plans receive a

distribution from their plan and contribute it to an IRA within a specific time limit or the

account assets can be transferred directly from the 401(k) plan to an IRA. Both actions have

the same consequences, and this article refers to both of them as rollovers.

6

Dana Muir & Norman Stein, Two Hats,One Head,No Heart:The Anatomy of the ERISA Set-

tlor/Fiduciary Distinction, 93 N.C. L. REV. 459, 497 (2015).

7

See infra text accompanying notes 21–22.

8

See infra Part I.A.

9

The default is statutory; the law provides that, except for small 401(k) balances, even when

the plan participant terminates employment, her account assets will remain in the plan

unless she actively chooses otherwise. I.R.C. § 411(a)(11)(b) (2018).

2019 / Behavioral Science Fails Retirement Savers 709