Sticky expectations: responses to persistent over-optimism in marriage, employment contracts, and credit card use.

• Author: Williams, Sean Hannon

Most people underestimate the likelihood that they will experience negative events and overestimate the likelihood that the law will protect them if those events occur. Many of these mispredictions are highly resistant to change even in the face of accurate and available information. This Article illustrates the consequences of these "sticky" expectations using examples from marriage, employment, and credit card regulation. In each of these areas, erroneous expectations create costs. The largest and most common cost is the failure to adequately self-insure against future negative events like divorce, job loss, or high debt. But proposals for correcting irrational expectations can be costly, in part because unrealistic optimism can also create benefits. This Article develops a Calabresian cost-benefit framework to help us to assess those costs and benefits sensibly, arguing that policy makers should seek to minimize the sum of the cost of disparities between expectations and reality and the cost of reducing those disparities. This approach can help to determine whether it is worth implementing legal reform to close the gap between expectations and reality, and if so, whether to do so by attempting to change the expectations or by changing the law to correspond to existing expectations. This framework provides reasons to rethink existing proposals aimed at informing or debiasing people through law.

INTRODUCTION I. STICKY EXPECTATIONS A. The Above-Average Effect, Comparative Optimism, and Overconfidence B. The Self-Serving Bias II. A CALABRESIAN APPROACH A. Debiasing Strategies 1. Dispositional Heterogeneity 2. Situational Heterogeneity B. Regulating in the Face of Limited Knowledge: Asymmetric Paternalism III. CASE STUDIES A. Marriage and Divorce 1. Evidence of Sticky Expectations 2. Is Debiasing Possible? 3. Collateral Costs of Debiasing 4. Alternate Proposals B. Employment Contracts 1. Evidence of Sticky Expectations 2. Is Debiasing Possible? 3. Collateral Costs of Debiasing 4. Alternate Proposals C. Credit Card Usage 1. Evidence of Sticky Expectations 2. Is Debiasing Possible? 3. Collateral Costs of Debiasing 4. Alternate Proposals CONCLUSION INTRODUCTION

Many people have erroneous beliefs about whether and under what circumstances the law will shield them from the consequences of a negative event. In some contexts, the vast majority of the population systematically mispredicts both the content of the law and its likely effects on their lives. Many of these mispredictions are highly resistant to change even in the face of accurate and available information. This Article illustrates the consequences of these "sticky" expectations using examples from marriage, employment, and credit card regulation. In each of these areas, erroneous expectations create costs. The largest and most common cost is the failure to adequately self-insure against future negative events like divorce, job loss, or high debt. This Article develops a Calabresian cost-benefit framework to assess possible legal responses to sticky expectations. This framework provides reasons to rethink existing proposals aimed at informing or debiasing people through law.

In many aspects of their lives, most people are irrationally optimistic about their futures. They do not think that negative events will happen to them, (1) and sometimes also believe that the law will protect them if the negative event occurs. (2) For instance, couples entering marriages radically underestimate their likelihood of divorce. (3) They also radically overestimate the likelihood that, if they do divorce, the law will provide a financial safety net for the poorer partner by awarding alimony. (4)

This irrational optimism can create substantial costs. First, it means that people do not adequately plan for divorce or other negative events, and fail to take steps that they might otherwise have taken to insure against their costs. Divorce leaves many women substantially poorer than they were during the marriage, (5) in part because they often leave the workforce during marriage and thus have less post-divorce earning potential. Had they been less optimistic during the marriage, some of these women might have chosen to maintain their work skills, thus self-insuring against the costs of divorce. Irrational optimism can also inflict emotional costs: when expectations are frustrated, people experience needless emotional harm. Loss aversion magnifies these harms--most people are extremely hesitant to accept unexpected losses and will enter high-risk gambles in the irrational hope of averting them. (6) These gambles often take the form of litigating weak claims, which impose unnecessary litigation costs on the parties and the legal system. (7)

Legal scholars have long recognized, of course, that people are sometimes irrational. But they have too often assumed that the law can change people's irrational expectations--and that doing so would necessarily be a good thing. (8) In reality, some expectations are highly sticky even in the face of accurate contrary information. For instance, informing people of the overall divorce rate and the relevant statistics on alimony does not alter their optimistic predictions about their own futures. (9) Moreover, even when it is possible, correcting over-optimism can impose collateral costs. For instance, optimism may itself be an important ingredient of a successful marriage--and so correcting couples' misperceptions may make marriages unhappier and divorce more likely. (10) A potentially discomforting conclusion emerges: a state seeking to increase the welfare of its citizens might, in some situations, have an interest in promoting misperceptions that stem from unrealistic optimism.

This Article adapts Guido Calabresi's famous argument that the law should minimize the sum of the cost of accidents and the cost of preventing accidents. (11) In the context of sticky expectations, the law should again seek to minimize the sum of two costs: the cost of the disparity between subjective expectations and reality, and the cost of responding to that disparity. Policymakers could take one of three general approaches in order to accomplish this goal. First, they could pursue debiasing strategies; that is, they could attempt to correct the erroneous expectations. Second, they could pursue insulating strategies; that is, they could change the law to align it more closely with the prevailing expectation or reduce the costs of the disparity in some other way. Third, they could allow the disparity to persist. Each choice carries costs. This Article focuses primarily on the costs and benefits of debiasing strategies. It also offers preliminary thoughts on several insulating strategies.

Sticky expectations predominantly result from optimistic biases. Two such biases are particularly relevant to the examples in this Article. First, people tend to believe that they are less likely than the average person to experience negative events like divorce and job loss. (12) This tendency is especially pronounced when people perceive that they have some degree of control over whether the negative event occurs. (13) This pattern is generally referred to as the above-average effect. Second, people tend to interpret ambiguous information in self-serving ways and therefore become overly confident in their predictions. (14) This pattern is generally called the self-serving bias.

Some expectations are stickier than others. Their stickiness is largely a function of their psychological source. The above-average effect is difficult to correct. Because people do not feel that they are average, providing information about the average person is seldom useful. Even providing personalized risk information has only fleeting and inconsistent effects on the strength of the above-average effect. (15) Self-serving biases, on the other hand, may be easier to correct. Lab studies show that people can be debiased by being forced to consider other plausible interpretations of ambiguous information. (16) Although real-world expectations may be stickier, it is possible that similar but intensified interventions could unstick them.

Although most sticky expectations result from a combination of the above-average effect and the self-serving bias, in many situations one of these two biases will play the dominant role. This will affect the choice and potential effectiveness of debiasing strategies. For instance, sticky expectations about marriage are primarily a reflection of the above-average effect. Spouses accurately predict the probability that the average person will get divorced, but believe that these statistical base rates are not probative of their own probability of divorcing. (17) Similarly, spouses accurately predict that courts usually do not award alimony and yet also overwhelmingly predict that they will be among the minority to benefit from this post-divorce safety net. (18) Like all above-average effects, these expectations are difficult to debias. And as noted above, even if debiasing were possible in the marriage context, it might not be advisable because preliminary data suggest that there is a causal relationship between optimism and long-term relationship satisfaction. (19)

This analysis of marital optimism sheds light on the potential pitfalls of a number of recent proposals. For example, Jeffrey Stake has suggested making prenuptial agreements mandatory in order to skirt the bluntness of a single default rule. (20) To the extent that mandatory premarital bargaining creates conflict and reduces marital optimism, its cost may exceed its benefits. Similarly, proponents of premarital counseling should carefully examine the content of these interventions in order to ensure that they do not needlessly burden marital optimism.

It may be possible to mitigate the costs of excessive marital optimism (such as poor financial planning) without incurring the costs of dissipating that...