When Nudges Fail: Slippery Defaults

Now that my last few posts have bludgeoned consumer financial education and at least bloodied disclosure, and given that my suggestion of comprehension requirements is completely untested as a means of consumer protection for financial products, what about “nudges”? 

One nudge I have taken a close look at is the use of policy defaults. Defaults are settings or rules about the way products, policies, or legal relationships function that apply unless people take action to change them. Although some defaults in the law are mere gap-fillers and others, as pointed out by Ian Ayres and Robert Gertner, penalize one or more parties with the intention that the parties will contract out of them, policy defaults aim for stickiness. The idea behind policy defaults is to set the default to a position that is good for most individuals, under the assumption that only the minority who have clear preferences to the contrary will opt out. 

The poster child for policy defaults is automatic enrollment in employer 401k retirement savings plans, which has boosted participation dramatically. Inspired by this apparently unmitigated success, policy defaults have become hot.  In the consumer finance area, legally-required default terms have been put in place for checking account overdraft coverage, credit card over-the-limit practices, and some home mortgage escrows. Consumer privacy on the internet (e.g., “do-not-track”) appears poised to be the next arena. 

But how broadly applicable are the results obtained from the retirement savings automatic enrollment default? My research shows that policy defaults may not be as effective in increasing welfare as many have hoped, in at least two respects. First, defaults created by the law are not always sticky, and can even be slippery. Second, those who opt out are not consistently the ones who are better off outside of the default. 

Why are other defaults so much slipperier than automatic enrollment? With respect to the policy default I have examined in depth, checking account overdraft coverage, the key is whether the interests of the parties are aligned. The overdraft policy default is that a bank cannot charge a customer for overdrafting her checking account at an ATM or through a non-recurring debit transaction unless the customer has explicitly opted out of the default and into overdraft coverage.  The aim of the default is to reduce use of overdraft, particularly by lower income consumers who have paid the bulk of overdraft fees in the past. In the retirement savings context, the employer, pension plan provider, and most employees generally benefit when employees stick with the automatic enrollment default. But overdraft fees are very lucrative for banks, and banks therefore oppose the overdraft policy default.

That one party opposes the default matters because the strength of the mechanisms that make defaults sticky varies dramatically depending on, e.g., the framing of the default and the process for opting out. When firms have significant control over the process for opting out or the context in which the defaults are presented, firms can undermine the stickiness of policy defaults. Further, individuals who generate the most revenue for firms by opting out may be the very people who would benefit most from the default position. Thus, firms may design the opt-out process and surrounding frame not only to make a policy default slippery in general, but specifically to encourage those who are worse off outside the default to opt out. In these situations, policy defaults may give the appearance of helping consumers, while doing little more than helping firms defuse political demands for substantive consumer protection. 

For a complete picture of how this works, you will have to read my article, When Nudges Fail: Slippery Defaults (80 U. Chi. L. Rev. forthcoming, 2013), but here are a couple of examples. 

One reason defaults can be sticky is because opting out usually requires some amount of time and effort.  In the case of the checking account overdraft default, however, banks make it more burdensome for consumers to stick with the default than to opt out. When the default went into effect, banks bombarded customers - and particularly those customers who had been frequent overdrafters in the past - with marketing urging them to opt out. Only by opting out could consumers stop the stream of phone calls and emails or navigate directly to on-line banking without first clearing a computer screen asking them to opt-out. 

Defaults can also be sticky due to judgment and decision biases that favor the status quo. For example, people will choose what they already possess over a change due to loss aversion and the endowment effect, so where a default is a position people perceive themselves as possessing, they will be more likely to stick with it. But bank marketing portrays the overdraft default as a change and opting out as the endowed, status quo position. One bank urges consumers to “STAY PROTECTED with … ATM and Debit Card Overdraft Coverage” and frames the choice to opt out of the policy default as follows:

                                                                    or

Third, defaults can garner traction though an implicit advice effect: people often believe that defaults have been set to the position that a party with superior knowledge believes is best for people. But banks explicitly advise their customers to opt out, robbing the default of any implicit advice it might otherwise have carried.

Not every party that opposes a policy default will be able to convince consumers to opt out. The opponent must have the ability to shape the frame in which the consumer considers opting out and the process for opting out. Further, where people’s preferences are well-established and easy to map onto available choices, neither defaults (as Liz Emons has demonstrated) nor attempts to drive people away from defaults will matter much. But where policy defaults encounter an opposed party, access to the consumer by that party, a confusing decision environment, and consumer preference uncertainty - as is the case for both the overdraft default and most of the proposed internet privacy defaults - policy defaults will fail.