Soft Paternalism and Consumer Protection

Contents

1. Introduction.

2. Soft Paternalism and Behavioral Economics.
2.1. Behavioral economics and deviations of rationality
2.2. Soft Paternalism
2.2.1. Definition and Origin.
2.2.2. Examples
2.2.3. Extent of intervention

3. Consumer Protection
3.1. Policy guidelines
3.2. Cooling-off periods
3.3. Providing information.

4. Risks and considerations
4.1. Slippery Slope
4.2. Additional Risks.

5. Conclusion

Bibliography

Figures

Figure 1: Consumer policy making steps

Figure 2: Consumer policy tools to target the demand and supply side of markets

1. Introduction

Economic models assume that human beings are rational, utility-maximizing agents with a set of consistent preferences. However, findings in behavioral economics suggest that when making decisions, people deviate from the full rationality assumed by economists; for instance, their preferences may change depending on how a situation is framed, or they may use heuristics that systematically lead to suboptimal outcomes (Sunstein & Thaler, 2003a). Based on behavioral economics, some scholars such as Richard Thaler and Cass Sunstein have proposed the use of soft paternalism to steer people into making welfare-enhancing decisions. Soft paternalism is a non-intrusive form of paternalism, where people retain their freedom of choice, but are steered in a certain direction by altering the environment within which they decide, making them better off by their own standards (Sunstein & Thaler, 2003a). Although the use of behavioral insights in policy making is not new, the introduction of the concept of soft paternalism has caused discussions in the polity. The Organisation for Economic Co-operation and Development (OECD) Consumer Protection Committee has discussed the effects of soft paternalism and behavioral economics on consumer protection and developed a toolkit to help design policies. In this work, I will explain the concept of soft paternalism, the implications it has for consumer protection, and what risks should policy makers take into account when considering soft paternalism.

In the first part of this work, I will describe the insights from behavioral economics dealing with deviations from the full rationality attributed to the homo oeconomicus in economic models. I will subsequently explain the concept of soft paternalism, as proposed by Sunstein and Thaler (2003a), (2003b), and (2008), Camerer et al. (2003), and Loewenstein and Haisley (2007). In the second part, I will go through the implications of soft paternalism and behavioral economics on the steps for considering consumer protection policies recommended by the OECD (2010). I will explore the demand-side focused policies of information provision and cooling-off periods. In the fourth part, I will explain risks and considerations that should be taken into account when considering soft paternalism, focusing on the risk of moving to harder forms of paternalism as presented by Whitman and Rizzo (2007) and (2009).

2. Soft Paternalism and Behavioral Economics

2.1. Behavioral economics and deviations of rationality

“ Economic research should use reasonable assumptions about agents ’ cognitive abilities. Economic models should make predictions that are consistent with micro-level data on decisions, including experimental evidence. Finally, economists have much to learn from psychologists. ” - Benjamin & Laibson, Good Policies for Bad Governments, 2003

In recent years, the field of behavioral economics has gained attention from academics as well as the general population. Daniel Kahneman was awarded the Prize in Economic Sciences in Memory of Alfred Nobel in 2002 for “having integrated insights from psychological research into economic science […]” (The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel, 2002). Some of the literature on the subject, such as Predictably Irrational by Dan Ariely; Thinking, Fast and Slow by Daniel Kahneman; and Nudge by Richard Thaler and Cass Sunstein have been quite popular on the bookshelves. These particular books explain the recent findings of behavioral economics, such as biases and heuristics, in layman’s terms; Predictably Irrational and Nudge also provide suggestions for policy makers and private parties on how to act based on these findings.

Standard economic theory assumes that people try to maximize consistent, well-defined preferences under complete information with unlimited cognitive abilities and will-power (Kooreman & Prast, 2010). Their preferences are internally consistent, context independent, and stable. In case there is uncertainty, they assign probabilities to outcomes and change them in accordance to Bayesian rules (Camerer & Loewenstein, 2002). Behavioral economists disagree with this image of human beings. They have identified some situations where economic agents do not show full rationality.

Heuristics are cognitive shortcuts that may lead to systematic suboptimal outcomes. Some of these heuristics, outlined below, are the anchoring effect, availability heuristics, the framing effect, the endowment effect, the status quo bias, and hyperbolic time discounting. These biases demonstrate ways in which humans can deviate from full rationality.

The anchoring effect happens when people make decisions based on an arbitrary “anchor”; they take an initial value and make adjustments to it until they arrive at an answer (Tversky & Kahneman, 1974). For example, when considering buying a certain good, the price of the first one seen can serve as an anchor to judge all subsequent goods of the same category. If humans were fully rational, the initial anchor would not make any difference in the final result; however, the initial starting point does influence the final results in practice (Tversky & Kahneman, 1974).

The availability heuristic is the judgment of probabilities based on how easily some of the instances can be remembered i.e. based on salience (Tversky & Kahneman, 1974). For example, a person might consider the risk of premature death caused by excessive consumption of alcohol or nicotine to be low if she knows someone who has lived a long life, even if she consumed alcohol and smoked cigarettes her entire life (Samson & Wood, 2010). The media has a strong effect in causing availability; which may explain why many people consider the risk of an airplane accident to be much higher than it actually is (Samson & Wood, 2010).

Framing effects occur when “different descriptions of the same problem highlight different aspects of the outcomes” (Kahneman, 2002). How a situation is framed can lead to humans deciding different alternatives if the information is presented differently. To exemplify this effect, a patient may perceive the outcome probabilities of an operation differently if the operation is presented as having a 95% chance of survival or a 5% chance of death, even if the same information is given in both cases.

The endowment effect happens when people value an item more highly due to the fact that they own it. They would be willing to pay less to acquire an object than they would to sell it (Kahneman, Knetsch, & Thaler, 1991). This is discrepancy is caused by loss aversion, which suggests that people derive less pleasure from acquiring something than they suffer from renouncing to it (Kahneman, Knetsch, & Thaler, 1991). Loss aversion can also help explain the status quo bias.

The status quo bias is the strong tendency of human beings to stay with the default position; choosing not to choose (Samuelson & Zeckhauser, 1988). The status quo bias is closely related to loss aversion, since loss aversion causes people to perceive deviations of the status quo as losses, which gloom larger than the potential gains of the new situation (Kahneman, Knetsch, & Thaler, 1991).

Hyperbolic time discounting is the tendency of human beings to value rewards that arrive today more than they do rewards that will arrive in the future, to a larger extent than standard economic theory would predict (Frederick, Loewenstein, & O'Donoghue, 2002). Someone might prefer €100 now instead of €110 tomorrow, but would prefer €110 in 31 days over €100 in 30 days; showing inconsistency in her time preferences. This deviation of rationality may lead to people wishing to make far-sighted decisions in the future, such as eating healthy, but when said future arrives making short-sighted ones, like eating unhealthy (Camerer & Loewenstein, 2002).

A distinction that should be made in situations within which people decide is between hot and cold states of mind. Hot states of mind are those states where humans cannot think “clearly”, such as fear, pain, or sexual arousal¹ (Samson & Wood, 2010). When in a cold state of mind, people tend to underestimate how being in a hot state of mind affects their behavior (Loewenstein, 2005); this is called a “hot-cold empathy gap”.

If human beings are prone to systematic mistakes in their decision making process, the outcomes of their choices may be sub-optimal. Being aware of the shortcomings of human cognitive abilities, proponents of soft paternalism argue, policy makers can use this knowledge to lead people into making choices that are better for themselves.

2.2. Soft Paternalism

“ The paradox of libertarian paternalism is that it terrifies both libertarians and paternalists. ” - Benjamin Wallace-Wells, Cass Sunstein Wants to Nudge Us, 2010

2.2.1. Definition and Origin

Soft paternalism², libertarian paternalism³, asymmetric paternalism^{4 5}, or light paternalism⁶ is the attempt to influence people’s behavior to improve their lives, without considerably⁷ limiting their freedom of choice. The first instance of soft paternalism can be found in “Libertarian Paternalism” by Richard Thaler and Cass Sunstein (2003a); in the same year they published a longer essay explaining the concept further, “Libertarian Paternalism is Not an Oxymoron” (Sunstein and Thaler, 2003b). Other works on the subject have been written by Camerer et al. (2003) in “Regulation for Conservatives: Behavioral Economics and the Case for Asymmetric Paternalism” and Loewenstein and Haisley (2007) in “The Economist as Therapist: Methodological Ramifications of ‘ Light ’ Paternalism”. In 2008, Richard Thaler and Cass Sunstein published the book “Nudge: Improving Decisions About Health, Wealth, and Happiness” which explains in simple terms the idea of soft paternalism with examples and recommendations for both private and public parties. Nudge, the book, was widely read and generated much discussion on the idea of soft paternalism, not only in academia, but among the broader population.

To illustrate this concept, Sunstein and Thaler (2003a), give the example of a cafeteria administrator who has to arrange in the order in which the food is presented; taking into account that more people will select the choice that is more salient. She could choose the order at random, she could choose the order so that her customers would be as obese as possible, or she could arrange the food in a manner that would make her customers better off (Sunstein & Thaler, 2003a)⁸. In case she chooses the food arrangement were healthy meals are more salient, a segment of her customers will choose “better” options for themselves without any coercion; whereas someone who rationally decides to eat unhealthy will be free to do so. This policy would thus satisfy the definition of soft paternalism, since it influences people’s behavior to improve their lives, without significantly restricting their freedom of choice. Benjamin and Laibson (2003) capture the essence of soft paternalism when they write: “[soft] paternalistic policies improve consumer welfare by enhancing an individual’s likelihood of maximizing her own welfare.”

Across the soft paternalistic literature, when behavioral economics shows the existence of time or preference inconsistencies, proponents of soft paternalism usually consider decisions made in cold-states of mind and long-term decisions to be the better ones. However, they do not offer a clear answer as to why one preference should be the correct one. Skeptics of soft paternalism, such as Whitman and Rizzo (2007) and Hill (2007) point this out in their works. For the purpose of discussion, I will grant the assumption that decisions made in a cold state of mind, that weigh the considerations for future selves more heavily, are the better choices.

Sunstein and Thaler (2008) call the person responsible for making decisions that affect the context in which people decide a choice architect; they argue that choice architecture is inevitable. Planners are forced to make design choices, such as choosing a default option (Sunstein & Thaler, 2003a). If human beings were fully rational, the existence of a default option would not alter their final choices; however, since humans deviate from full rationality, the default option has a large effect on the choices people make (Kahneman, Knetsch, & Thaler, 1991).

Sunstein and Thaler (2008) argue that soft paternalism may be the “Third Way,” i.e. a middlepoint between liberal and conservative; between libertarian and paternalist (Sunstein & Thaler, 2008). They posit that those who greatly value freedom of choice, e.g. libertarians or classic liberals, should find soft paternalistic policies acceptable, since they preserve freedom of choice and engaging in choice architecture is sometimes inevitable (Sunstein & Thaler, 2003a); whereas paternalists should find the policies attractive as well, since they guide people into making decisions that are in their best interest.

This work does not deal with private uses of soft paternalism but only with those alternatives that could be used by governments to steer people into making certain decisions.

2.2.2. Examples

An example of a soft paternalistic policy can be found in the Save More Tomorrow (SMarT) plan, designed by Thaler and Benartzi (2004), where people decide in advance that a portion of their future salary increases will be saved for their retirement. It is argued that people fail to save enough money for retirement for reasons that, according to proponents of soft paternalism, can be attributed to deviations of rationality (Thaler & Benartzi, 2004). The failure to save can be partially attributed to loss aversion, since saving money in 401(k) plans⁹ can be seen as a loss of income; status-quo bias, since the default is to not contribute anything; and hyperbolic time discounting, since current consumption, facilitated through income, is weighted much more heavily than future consumption (Loewenstein & Haisley, 2007). The SMarT plan can help de- bias people’s behavior. Since the savings take place in the future, they do not fall victim to hyperbolic time discounting. Once in the program, the process repeats itself until the maximum contribution rate is reached, making the status-quo bias work in employees’ favor. It should be noted that this program is not actively enforced by the state, but has been adopted by some private companies voluntarily.

Another example of a soft paternalistic policy is requiring choosing gambling liquidity limits (Benjamin & Laibson, 2003). Gambling can cause individuals with self-control problems to gamble away their savings or significantly indebting themselves, which are assumed to be undesirable outcomes. If a policy requires gamblers to choose a liquidity limit by themselves, ex-ante, in a cold-state of mind, they will be unable to choose in the heat of the moment to spend more money than they initially intended to (Benjamin & Laibson, 2003). This allows rational gamblers who wish to risk large amounts to do so, while helping gamblers with self- control issues to make better choices, by their own standards, since they set their own limits.

The distinction between soft paternalistic policies and harder forms of paternalism lies in maintaining freedom of choice. In an example given by Loewenstein and Haisley (2007), if employees of a company wish to reduce their weight¹⁰, a hard paternalist would ban soda machines altogether, whereas a soft paternalist would keep the machines in the building, but devise a system that did not allow those who state that they wish to lose weight to use them (Loewenstein & Haisley, 2007). This would protect people from hyperbolic time discounting, since they choose their goals ex-ante, and are not able to violate their preferences, as stated by themselves, when faced with temptation. This system can be seen analogous to Ulysses’ actions in the Odyssey, when he decided to be tied up in order to not fall victim to the temptation caused by the mermaids’ songs.

2.2.3. Extent of intervention

When engaging in choice architecture, the costs of deviating from the desired outcome may increase, or some choices could be foreclosed. In the case of the cafeteria manager, if the food is located far away from the cash register, the transaction costs associated with eating unhealthy meals increase. In the case of the SMarT plan, people may have to go through many bureaucratic forms to opt-out of the program. In gambling liquidity limits, a gambler may rationally decide, in the light of new information, to gamble more than he initially intended, but be unable to do so due to the enacted policy.

If the costs of choosing alternatives contrary to the preferred option of the soft paternalistic policies are high, freedom of choice could be limited. If in order to protect consumers and de- bias their behaviors, a consumer protection agency requires cooling-off periods¹¹, the freedom to consume a good or service immediately is reduced. If a cooling-off period lasts a year, is it still soft paternalistic? What would be the optimal duration? It is important to understand how proponents of soft paternalism view the boundaries of government actions when considering soft paternalism, since harder forms of paternalism are considered undesirable¹². In the third part of this work I will explore the possibility of soft paternalistic policies leading to harder forms of paternalism.

To which extent should freedom of choice be limited is a matter subject to debate. Sunstein and Thaler (2003b) admit that the line between libertarian, i.e. soft, and non-libertarian, i.e. hard, paternalism is blurred. They proceed to call a soft paternalist who values freedom of choice greatly, and is thus determined to make it relatively costless for people to choose their preferred alternative, a libertarian paternalist; whereas someone who is confident in her judgment and is willing to impose larger costs on those who deviate should be called a libertarian paternalist. In an exchange of opinions between Rizzo and Thaler in 2007, Thaler emphasizes only nudging by policy when it is inevitable, i.e. when a choice has to be made necessarily, as in the example of the cafeteria (Rizzo, 2007). Thaler is critical of nudging when such a decision is not inevitable and there is no reason to believe that a nudge is really needed (Rizzo, 2007). Camerer et al. (2003) acknowledge that sometimes suboptimal decision making cannot be counter-acted by cooling-off periods, providing information, changing the default, or any other soft paternalistic policy alone; therefore, they argue, there may be benefits in limiting consumer choices in some cases. Although proponents of soft paternalism remain skeptical of harder forms of paternalism, they allow room for limiting freedom of choice in their work.

Camerer et al. (2003) propose an equation to determine the net benefits of a given policy: (p * B) - [(1 - p) * C] - I + > 0. They argue that if the benefits for those who are not fully rational outweigh the costs for those who are, a policy can increase social welfare. In this formula, p is the percentage of those boundedly rational, B are the benefits for boundedly rational people, C are the costs for fully rational agents, and I are the implementation costs of a policy. However, based on this equation, some policies that would not be considered soft by many could be justified; such as sin taxes, which would significantly raise the cost of rationally engaging in a given activity. O’Donoghue and Rabin (2006) argue that sin taxes can improve total social surplus, satisfying Camerer et al.(2003)’s equation, if there is a percentage of the population that lacks self-control, while imposing little costs to those who are rational, if there is redistribution of the tax revenue.

Even if policy makers choose Camerer et al. (2003)’s formula to measure the net benefits of a policy, how to measure the benefits B is not an easy task. One of the causes of this difficulty is that revealed preferences cannot be used to judge welfare, according to proponents of soft paternalism (Sunstein & Thaler, 2003a). Loewenstein and Haisley (2007) state: “Clearly, the traditional welfare criterion used by economists, which involves satisfying people's preferences to the maximum extent possible, cannot be used to evaluate policies that are premised on the view that people don't always choose what's best for themselves.”¹³ The costs, and all other elements of the formula, are difficult to assess as well. If the benefits and costs of a given policy are open to interpretation, some policies which significantly restrict freedom of choice can be enacted.

[...]

¹ An example of decision making under sexual arousal can be found in an experiment by Ariely and Loewenstein (2005), where participants in a “cold” state of mind said they would not wish to hypothetically engage in certain sexual activities but under arousal said they would hypothetically engage in them.

² I will only use the term soft paternalism for simplicity. However, all the ones named can be used interchangeably.

³ Libertarian paternalism is a term coined by Cass Sunstein and Richard Thaler (2003a). They define it as: “An Approach that preserves freedom of choice but that authorizes both private and public institutions to steer people in directions that will promote their welfare.” (Sunstein & Thaler, 2003a)

⁴ Asymmetric paternalism is a term coined by Camerer et al. (2003). They define an asymmetrically paternalist policy as: “A regulation is asymmetrically paternalistic if it creates large benefits for those who make errors, while imposing little or no harm on those who are fully rational” (Camerer et al., 2003)

⁵ It should be noted that under the definition of asymmetric paternalism, more limitations of individual freedom are justified if the benefits for those who are boundedly rational outweigh the costs for those who are fully rational.

⁶ Light paternalism is a termed coined by Loewenstein and Haisley (2007). They define it as “In contrast to traditional 'heavy-handed' approaches to paternalism, light paternalistic policies aim to enhance individual choice without restricting it.” (Loewenstein & Haisley, 2007)

⁷ The extent to which freedom is limited in soft paternalism is subject to debate and will be addressed in this work in the next pages

⁸ Sunstein and Thaler (2003) assume that, in this case, the healthier choices are the ones that would make

consumers better off by saying: “Would many object to putting the fruit before the desserts at an elementary school cafeteria if the outcome were to increase the consumption ratio of apples to Twinkies?” (Sunstein & Thaler, 2003). In “The Economist as Therapist” Loewenstein and Hailey (2007) also point to the obesity rates to argue that people are making sub-optimal decision. Indeed, they are skeptical about the view that “people are obese because they have calculated that the pleasure from the extra food, or the pain from the foregone exercise are sufficient to compensate for the negative consequences of obesity” (Loewenstein & Haisley, 2007)

⁹ A defined-contribution pension program in the United States of America

¹⁰ In this case, the goal of losing weight is stated by the employees themselves (Loewenstein & Haisley, 2007). This is called a self-officiating policy. It should be noted that hard paternalists would be willing to ban the soda machines if they believe people are better off without them, even if the employees do not express any wish to reduce weight or improve their health.

¹¹ With cooling-off periods, consumers are delayed from taking action for some duration or have the right to nullify a contract during a determined period

¹² One of the strongest cases against (hard) paternalism can be found on “On Liberty” by John Stuart Mill (1859). Mill argues that people should retain their freedom of choice, since they are in a better position than any other party to assess what is better for themselves. He states: “the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others. His own good, either physical or moral, is not a sufficient warrant. He cannot rightfully be compelled to do or forbear because it will be better for him to do so, because it will make him happier, because, in the opinion of others, to do so would be wise, or even right.” (Mill, 1859) Soft paternalism properly applied would not violate this, since those steered would retain their freedom of choice, thus they are not coerced against their will.

¹³ Loewenstein and Haisley (2007) discuss in the same work some of the criteria that could be used to assess welfare, such as experience utility, limiting welfare to valid choices, informed decision utility, and capabilities.