Kate Tokeley[*]

7.1 Introduction

The first stage of this project identified the range of issues that confront the New Zealand policy maker when deciding whether or not to introduce a paternalistic law to protect consumers from harming themselves. For the purposes of this project a "paternalistic consumer law" is any law that has the goal of encouraging or coercing consumers to act in a way that the government determines is in the consumer's best interests.[1] This definition is wide enough to cover laws that might also have a co-existing non-paternalistic goal of preventing or discouraging consumers from behaving in ways that harm other people.

The latest findings of behavioural economics suggest that consumers in a free market do not always make rational choices that maximise their self-interests. Consumers suffer from impulsiveness, excessive optimism, status quo bias, a lack of will power, poor ability to adequately weigh up short-term and long term benefits and an inability to process large amounts of complex information.[2] They also have a limited time within which to make purchasing decisions. Consumers often act in ways that seem to decrease their welfare. For example, they eat too much fatty food, smoke cigarettes, do not save enough, borrow too much, enter contracts with seemingly unfair terms, fail to negotiate cooling-off periods, use dangerous recreational drugs, engage in dangerous recreational sports and buy unsafe products.

Evidence that consumers make mistakes, however, does not necessarily justify unlimited paternalistic interventions to correct these mistakes. The crucial question for New Zealand policy makers is whether, and for what reasons, legal paternalism to protect consumers from these poor decisions is justifiable. The spectrum of possible paternalistic intervention ranges from soft paternalistic measures which aim to nudge or encourage the consumer to change their behaviour (for example, taxes, advertising bans, opt-out government savings schemes and other default rules[3]) to harder paternalistic measures that coerce the consumer to behave in a certain way (for example, bans, mandatory cooling-off periods, prohibitions of unfair terms and interest rate caps). Examples of potential areas of future law reform in New Zealand that will give rise to issues of paternalism are given in the "Issues and Background" chapter on this topic in the first volume of the New Zealand Law Foundation Regulatory Reform Project.[4] Some key areas include unfair terms prohibition, the imposition of responsible lending duties on suppliers of credit, credit interest rate caps, legislation to reduce obesity levels, regulation of natural health products, requirements for cigarettes to be in plain packaging[5] and adjustments to the KiwiSaver scheme.

The overall objective of this chapter is to establish a framework to guide policy decision-making about paternalism in New Zealand consumer law.[6] The framework is intended to encourage policy makers to engage in an informed debate about the topic. Without such a framework there is a danger that a serious debate about paternalism is neglected or that the debate reverts to an exchange of broad, unexamined assertions. Those opposing the proposed law brush it aside as being too paternalistic or an example of the "Nanny State" and those in favour make vague assertions about the need for "consumer protection".

This chapter represents stage three of the project. It gives conclusions and recommendations on the topic. The chapter is divided into the following parts: Part 2 considers the tension between consumer freedom and paternalism and outlines the general approach advocated by the chapter. Parts 3 to 5 identify the various factors that are relevant to determining whether a paternalistic intervention is justifiable and, if so, whether it should be soft or hard paternalism. Each of the factors is examined to assess how it might strengthen a justification for paternalistic intervention. Part 6 examines the relationship between consumer welfare and consumer preferences.

If policy makers are satisfied that the particular consumer problem is one where pursuing the goal of paternalistically protecting consumers is justified in principle, then the question becomes one of designing a regulation that is likely to be effective and avoiding unintended negative consequences. The primary focus of this chapter is the development of a framework for establishing a prima facie case for paternalistic intervention. Nevertheless, the discussion is not complete without some mention of the issues of effectiveness and unintended consequences. Part seven briefly considers these issues.

7.2 The tensions between consumer freedom and paternalism

Whether a paternalistic legal intervention can ever be justifiable has been the subject of much legal, philosophical and economic debate for hundreds of years. Stage One of this project reviewed the existing theories on the legitimacy of legal paternalism. Each theory is based on a single premise; paternalism is seen as either a good idea (traditional paternalism); a bad idea (anti-paternalism); justifiable only if it retains some consumer choice (libertarian paternalism); or only if the benefits to irrational consumers outweigh the costs to rational consumers (asymmetric paternalism).[7] It was argued in Stage One that the search for a unifying theory offers inadequate assistance to the policy maker.

Neither of the extreme positions on paternalism provide guidance for policy makers seeking to assess the legitimacy of paternalism in any given situation. Modern democratic societies such as New Zealand value individual liberty too highly for uncritical paternalism to be acceptable. However, uncompromising anti-paternalism offers little help to those policy makers who are open to the idea that paternalism might be appropriate in some situations. This project aims to assist policy makers to identify factors that might provide a justification for a paternalistic intervention.

Stage One of this project also argued that the current middle ground "single premise" theories on paternalism are unable to offer complete answers to the question of when to intervene.[8] Libertarian paternalism, for example, is in favour of soft paternalism, and yet Thaler and Sunstien, the most well known proponents of the theory, admit in the penultimate chapter of their book Nudge that hard paternalistic measures, such as bans of unsafe products or cooling-off periods, might be justified in some cases.[9] The book gives little analysis of what factors might be relevant to establishing a rationale for coercion. This is understandable given that the central thesis of libertarian paternalism rests on the premise that only soft forms of paternalism are legitimate because these measures respect individual liberty. Neither does the book provide any guidance on how to filter out the types of "poor" consumer decision-making where even soft paternalism is inappropriate.

New Zealand already has many paternalistic laws and there are likely to be many more policy debates about the legitimacy or otherwise of introducing additional paternalistic measures.[10] Examples of current paternalistic laws that are widely accepted include airline safety standards, medicine safety regulations, the ban on sale of hard drugs such as heroin, cooling-off periods in door-to-door sales and the prohibition on contracting out of the Consumer Guarantees Act 1993.[11] These laws restrict consumers' freedom of choice. Consumers are not free to: take a cheaper but possibly less safe flight; purchase heroin in the supermarket; negotiate a cheaper price with no cooling-off period; or contract out of statutory guarantees as to the quality of consumer goods and services. This chapter asks what makes these paternalistic restrictions on consumer freedom seem more acceptable than other examples of paternalism. Suppose the state wanted to enact laws that remove consumers' freedom to purchase chocolate or ugly clothes or go on a hot air balloon or climb Mount Everest. How is the policy maker to distinguish between arguably legitimate interventions and those interventions that are illegitimate? It is not the goal of this chapter to argue for an exact point at which to draw the line between individual liberty and paternalistic legal intervention. The goal is to identify the kinds of factors that should be addressed in any debate on the topic. People of different political persuasions will draw the line at different points and will attribute different weight to the importance of each factor.

The starting point for making decisions about the appropriateness of paternalism is to acknowledge that paternalistic intervention is at odds with individual liberty, the sanctity of contract and the value of personal responsibility. It requires the government to make a potentially controversial assessment about what it believes to be in the best interests of consumers. It then restricts consumer freedom in order to change consumer behaviour. Both paternalistic "nudges" (soft paternalism) and paternalistic "pushes" (hard paternalism) compromise consumer freedom of choice, albeit to different degrees. For example, a tax on fatty foods, a form of soft paternalism, does not completely eradicate consumers' freedom to choose to eat those foods, but it does take away the freedom to purchase them at market value. Likewise, the "nudge" of automatically enrolling people in an opt-out retirement saving scheme reduces consumers' freedom to easily choose their own savings plan.[18] Opting-out of the approved plan requires time and effort.

Any debate about the legitimacy of paternalism should begin with the assumption that consumer freedom is generally a good thing. A paternalistic intervention ought therefore to proceed with caution. Academic commentators have recently expressed concern that the New Zealand Government is too often in a hurry to pass legislation and fails to allow time and opportunity for policy deliberation.[19] Paternalism is an issue which requires policy makers to slow down and allow time for informed debate and the development of an in-depth set of justifications for intervention.

This chapter advocates moving from a single premise approach to a multifactorial approach which better recognises the complexity of the problem. It recommends that various factors are assessed in order to determine if a paternalistic intervention to protect consumers might be legitimate and, if so, whether it should be soft or hard paternalism. The factors used to justify intervention in one case might be different from the factors that justify intervention in another.

The following parts of this chapter explore this issue of developing a set of factors relevant to determining legitimacy.

7.3 Overview of factors

Policy makers need to define the consumer behaviour that they are proposing to alter by way of legal paternalism.[20] Then they should consider the following factors when determining whether a paternalistic intervention might be legitimate:

· the magnitude of potential consumer harm;

· the probability of consumer harm;

· the irreversibility of potential consumer harm;

· the degree to which addiction is affecting consumer choice;

· the degree to which consumers want to be protected;

· the degree to which consumers are dealing with complex large quantities of information they are unable to process;

· the degree to which the problem is affecting children, young adults or other potentially disadvantaged groups;

· the degree to which there are additional, non-paternalistic reasons for enacting the law; and

· the probability of non-legal responses, such as education or support programmes, failing to provide solutions to the problem within an acceptable time frame.

Each of these factors requires policy makers to make a detailed assessment of the extent and nature of the consumer problem. Consideration of each factor will produce results on a spectrum ranging from high to low. Where a particular consumer problem gives results at the higher end of some or all of the spectra there is a more compelling argument for paternalistic intervention than if the answers are at the lower end of the spectrum. So, for example, where the magnitude of potential consumer harm appears extreme there is a stronger case for intervention than where it is minor. Consideration of all these factors is essential to mounting a challenge to objections to paternalism.

The two main objections to paternalistic consumer laws rest on one of two arguments, both about freedom. The first argument says that paternalistic laws violate consumers' fundamental right to freedom of choice – even the freedom to make mistakes that harm themselves.[21] This argument is about the importance of liberty itself, not about how good consumers are at using their freedom to make good decisions. It says that consumer freedom of choice has intrinsic worth. Making choices is what makes us human, it provides us with dignity and the ability to be self-reliant. This emphasis on the importance of freedom is also tied in with ideas about personal responsibility and the sanctity of contract.

The second argument is about the relationship between freedom and welfare. It says that freedom of choice is the most reliable way of enhancing consumer welfare.[22] Therefore, paternalistic laws are objectionable because they make the mistake of assuming that the government is better equipped than consumers to determine what is in the consumers' best interests. According to this theory, the revealed preferences, as evidenced by consumer choices, are the best indicator of consumer welfare. Some scholars also point out that the government decision-makers themselves have cognitive limitations and are prone to error.[23] One further potential objection to legal paternalism is that it is not legitimate to attempt to resolve a consumer problem with a paternalistic legal response if there are non-legal responses that could achieve the same result in an acceptable time frame.

The following sections of this chapter discuss each of the factors in turn and make an assessment as to how they might be relevant to combating objections to legal paternalism.

A further difficulty arises when attempting to apply cost-benefit analysis to situations involving unstable consumer preferences in respect of products that produce short-term benefits and long-term costs. Calculating consumer benefits and costs in these cases requires policy makers to choose between various different methods of calculation. For example, calculations could be based on consumers' revealed preferences (as evidenced by their choices), or consumers' stated preferences, or government and expert assessments of consumer benefits and costs. Each method could give different results and rests on different assumptions about consumer welfare. Typically, a traditional cost-benefit analysis will calculate consumer welfare on the basis of revealed consumer preference (as evidenced by consumer choice).[26]

The approach for analysing the legitimacy of paternalism recommended in this chapter allows for an engagement in a discussion about moral disagreements. This includes the disagreement about when and why it is appropriate, or not, to use consumer choice as a yardstick for measuring consumer welfare. The suggested approach starts with the most common objections to paternalism and then considers various factors that in any given situation might overcome these objections. The answer to the question of the legitimacy of paternalism ultimately depends on a value judgment. It requires a value judgment about the limits of freedom and the wisdom of government assuming that, in some cases, it can make better decisions about consumer welfare than consumers themselves. It depends, in part, on one's conception of justice. For some people the idea of liberty to do as one chooses, so long as it does not harm someone else, is seen as such a fundamental right that it cannot be easily overridden by any number of benefits.[27] For others there will be a point at which they are convinced that in the circumstances a paternalistic intervention is justified. Having a checklist of factors to be considered encourages the policy maker to engage in a discussion about the acceptable limits of freedom. The aim is for policy makers to be able to explain exactly why this particular consumer problem is one for which they consider it is legitimate to aim to reduce consumer liberty in order to increase consumer welfare. After this baseline question of the legitimacy of the goal has been addressed, the policy maker can go on to deal with the issue of designing regulation that is most likely to effectively achieve this goal and is cost-effective.

7.4 Factors to be considered

The following factors should be taken into account when determining the legitimacy of a paternalistic legal intervention. Each factor is relevant to combating the common objections to legal paternalism.

7.4.1 The magnitude of potential consumer harm

Some "poor" consumer choices lead to relatively minor harm and therefore should be left to individual choice. For example, consumers often make hasty, seemingly irrational purchasing decisions to buy items such as gadgets they never use, unaffordable holidays or bad haircuts. It would be hard to justify laws that intervene with these types of purchasing decisions because of their minor consequences. Allowing consumers to choose freely in these situations seems preferable to intervening in some misguided attempt to help consumers make better choices.

However, where consumer choices cause extreme harm it is easier to argue that paternalistic laws are justifiable. For example, choosing to use heroin, smoke cigarettes or buy untested medicines can lead to serious diseases or even death. Other consumer problems such as over-indebtedness or problem gambling can lead to financial ruin and serious psychological distress.

The magnitude of harm is relevant both to the arguments about freedom of choice and consumer welfare. If the harm is extreme then it might be argued that, although the consumer choice is voluntary, it may not have been a "free" choice. Perhaps the consumer is addicted to the product or is struggling with the willpower to put long-term goals ahead of short-term pleasure.[28] Problems with alcohol, cigarette smoking, gambling, over-indebtedness, obesity and recreational drugs are all in this category. Where harm is extreme it seems more likely that the consumer choices do not represent the consumer's "true" or underlying preferences.[29] While the consumer might not be under duress or subject to fraud they might nevertheless be unable to make an unconstrained and free choice.

The degree of harm is also relevant to weakening the argument that consumers are better equipped than governments to make choices that enhance their welfare. The more extreme the harm that consumers are causing themselves by their choices, the more persuasive is the argument that their choices are in fact decreasing their welfare and that perhaps the government might be in a position to help. Unlike the consumer, the government does not face the temptation to indulge in immediate gratification. Moreover, the government may be in a better position than consumers to devote the time and resources to examine the situation in more detail.

Another category of purchases where consumers might make a choice that causes them serious harm is when they are purchasing goods or services for which large amounts of complex information needs to be digested before making a judgment. The problem in these cases is nothing to do with willpower. In these cases, even if all the relevant information was made available, consumers would not have the cognitive skills to process and understand the information in order to make a rational decision about their own welfare in the time available.[30] For example, in a free market, consumers would be able to choose the airline safety and price combination that they would prefer. Similarly, decisions about bicycle safety, medicine safety and building safety would be up to consumers to choose the level of risk they deemed acceptable. Expecting ordinary consumers to make a welfare-enhancing decision in these situations is unrealistic. A consumer would need to be well qualified in aviation safety to assess the likely outcome of various airline safety specifications. In these situations an unregulated market could lead to consumers suffering serious harm.

7.4.2 The probability of consumer harm

The probability of consumer harm is important to any argument that attempts to justify paternalism. If consumers engage in behaviour that has a very high chance of causing harm then this is relevant to concerns about both liberty and consumer welfare. It could indicate that the consumer choice is being made because of some sort of consumer cognitive bias (such as failure of willpower, addiction, or a lack of information-processing powers) and is possibly not a truly free choice. It might also indicate that in these cases the consumer might not be in the best position to make choices that maximise his or her welfare.

If there is potential for consumers to be harmed by their choices, but the chances of this harm occurring are extremely low, then it becomes more difficult to argue that the government can legitimately interfere with consumer freedom. For example, while there is a minute risk that consumers could choke while eating grapes it would be hard to argue for a ban on the sale of grapes. Likewise there is some chance a child could break an arm at a skating rink; but this does not necessarily mean that skating rinks should be banned.

Some consumer behaviour carries a high risk of harm occurring. For example, consumers who choose to smoke cigarettes have a high chance of developing lung cancer or some other disease associated with tobacco use.[31] Even moderate cigarette smoking is likely to cause some harm to a smoker's health. The latest American Surgeon-General's Report on tobacco concludes that there is no safe level of exposure to tobacco smoke. Any exposure to tobacco smoke, even an occasional cigarette or exposure to second-hand smoke, is harmful.[32] In other situations the likelihood of harm is lower. For example, consumers who sign a contract with an excessively punitive default clause might be highly unlikely to default and so their poor choice may not carry a high risk of causing harm.

When assessing this factor, policy makers need to be careful to accurately define the "poor" consumer choice that they are concerned about in order to assess whether these choices are likely to lead to consumer harm. If, for example, a person's gambling consists of buying the occasional lotto ticket it is unlikely to cause any consumer harm. However, playing the pokey machines for six hours a day is likely to cause harm to the consumer. Many consumer problems, such as obesity and alcohol abuse, stem less from the consumer product itself than the way the product is used. Eating two squares of dark chocolate a day and drinking two glasses of wine a week is not likely to cause harm; but, consuming two bottles of wine a night and junk food for most meals is likely to result in serious harm. Policy makers may decide that it is only these specific types of high risk behaviour for which legal paternalism is justified. This would then require the regulator to attempt to specifically target the problem behaviour and not restrict the liberty of those consumers making choices that are unlikely to cause them any harm. A law that prohibits liquor stores from being open at 4 am is an example of legislation targeted at the type of consumer behaviour that is most likely to lead to a decrease in consumer welfare.

7.4.3 The irreversibility of potential consumer harm

In some cases the harm that consumers cause themselves by poor decision-making is, at least to some extent, reversible. If a consumer buys a shirt that is too small they can pay more money to get another shirt the next size up. The consumer has still suffered a minor monetary loss but they have solved the problem of having a too-tight shirt and can potentially re-sell the original shirt, albeit at less than the retail value. In some situations, consumer harm is more difficult, but still feasible, to reverse. If a consumer chooses to consume large quantities of junk food and they become overweight, it is possible, even if challenging, to reverse this weight gain by diet and exercise. However, at some point in time weight gain can lead to irreversible conditions such as diabetes and heart disease.

Other consumer behaviours might lead more quickly to irreversible harm. For example, tobacco use can be distinguished from more minor consumer problems by the irreversibility of many health conditions caused by smoking. The most obvious examples of irreversible harm involve permanent disability or death. Significant financial distress (such as a mortgagee sale or bankruptcy) might also be classified as "irreversible" even if many years later the consumer might have built up a greater level of net wealth.

If consumers are behaving in ways that can cause irreversible harm then there is a stronger argument for legal intervention to protect consumers than if consumers are causing themselves minor and reversible harm. If the harm is reversible then the consumer has the opportunity to learn from their mistake.[33] If the harm is irreversible then it is too late. Any market feedback mechanism would come too late to undo any harm caused. Leaving matters such as the safety of medicines or airlines to be optimised by the free-market risks some consumers suffering death or injury.

Where consumer behaviour is causing irreversible harm, the arguments about the importance of preserving consumer liberty are weakened. So too are arguments that the consumers are in the best position to maximise their own welfare.

7.4.4 The degree to which addiction is affecting consumer choice

The argument that consumers choose behaviours that are in their best interests, and so should be left alone without government interference, appears to break down in situations of strong addiction. Addiction can be defined as physical and/or psychological dependence on mood-altering substances or behaviours which temporarily alter the chemical profile of the brain.[34] Addiction explains, at least in part, why some consumers continue to smoke, drink, eat unhealthily, take recreational drugs and gamble despite the negative consequences of these behaviours. Addiction was mentioned in Part 7.4.1 of this chapter as a possible cause of consumers causing themselves severe harm and that this might indicate a lack of genuine free choice. This part of the chapter looks more closely at the question of whether or not addiction impairs the ability of a consumer to make a rational and free choice about his or her own self-interest.

The degree of addictiveness varies.[35] Some forms of addiction are more behavioural (gambling) and some are more chemical (cocaine). Another variation is between the potency of the pleasure produced and the difficulty of withdrawal. Some addictive products produce mild highs and have relatively mild withdrawal effects (for example, coffee or chocolate). Other products generate stronger highs and have stronger withdrawal effects (for example, cocaine, heroin, cigarettes). A further variation is the ability of the addictive substance to impact on rational thinking. For example, coffee improves cognitive functioning, whereas alcohol decreases it.

If a consumer problem is not a result of addiction then it is easier to argue that the consumer behaviour is a result of a genuine free choice and that, in the absence of other factors such as extreme levels of harm, the consumer is maximising his or her own self-interest by exercising that freedom of choice. Paternalism is therefore unjustified. However, where addiction is a factor, the degree to which consumer choice is genuinely free is more difficult to determine.

Consumers who suffer from an addiction seem to lose the ability to think rationally or to choose freely. Some theorists, however, have argued that if a consumer has enough information about the dangers of an addictive product, he or she will "rationally choose" to use the product in a way that maximises welfare.[36] Under this theory, for example, a smoker would take into account the various costs of smoking one more cigarette today (the monetary cost of the cigarette, the increase in addiction and the decrease in future health) and rationally compare that with the pleasure of smoking that cigarette. The choice to smoke will therefore maximise the consumer's welfare. If this theory is correct, paternalistic intervention is unjustified. Regulation should merely make sure smokers are well-informed and that they are not imposing costs on third parties (such as inflicting second-hand smoke on to others).

Several scholars have severely criticised rational addiction theory.[37] The philosopher Jon Elster argues that rational addiction theory is flawed. He maintains that addiction and emotions can short-circuit rationality so that people behave in ways that are not in their self-interest.[38] He uses a subjective definition of "rationality". Behaviour is not considered irrational just because third parties view it as self-destructive. It is irrational because the consumer, given what he or she knew at the time, thinks he or she should have acted differently. Elster argues that the irrationality of addictive behaviour is evidenced by addicts' attempts at self-regulation.[39] For example, a smoker might put photographs of diseased lungs on the wall, an alcoholic might empty the house of all alcohol, a gambler might ask for the casino to deny him future entry into the casino.[40] These actions show that the consumer has a reason to think they should be behaving differently. The failure to quit is not the result of a rational choice based on an assessment of all the costs and benefits. Elster acknowledges that in some cases the decision to continue using an addictive product might be rational.[41] For example, a schizophrenic who smokes in order to medicate the symptoms of the disease may, from his or her perspective, be making a rational choice. Presumably the decision to continue consuming moderate quantities of mildly addictive, relatively harmless substances such as coffee or chocolate might also be classed by Elster as rational.

Economist Ole Rogeberg describes the rational addiction theory as "absurd".[42] He argues that it is implausible to explain the behaviour of all addicts as the result of choices made by well-informed, utility-maximising individuals. Such a theory makes the false assumption that we make fully thought-out, forward-looking, complex calculations before making our choices. It presupposes that "we are aware of or can uncover the reasoning behind our choices".[43] Rogeberg argues that this leads to the false assumption that addictive products pose no welfare problems.[44]

If it is accepted that addictions can, if extreme enough, impact on the cognitive abilities of consumers to make rational choices about their own welfare then the two main arguments against paternalistic intervention are eroded. The argument based on the importance of liberty is weakened because consumers who are addicted to a product or behaviour are arguably not acting "freely". The argument based on consumer choice leading to maximised welfare is also weakened. If a consumer's rationality is impaired by addiction then their ability to make welfare-enhancing choices is compromised. The higher the degree to which addictiveness is affecting consumer choice, the more compelling is the argument that some form of paternalistic intervention might be justified in order to help consumers to make "better" choices.

7.4.5 The degree to which consumers want to be protected

The degree to which consumers want to be protected is relevant to responding to the argument that consumers' freedom of choice is of fundamental value and should not be violated. Where it can be shown that most consumers are in favour of having their liberty curtailed it becomes much harder to claim that liberty is sacrosanct. This factor is also relevant to countering the argument that consumer freedom of choice is the best method of maximising consumer welfare. If consumers themselves are supportive of paternalistic regulation it suggests that they do not feel that they are making good decisions on their own. It indicates that in this instance the government might be in a better position than consumers to maximise consumer welfare. Tim Irwin recommends that regulators ask themselves whether "the people making the apparent mistakes consider, on reflection and with good information, that their decisions are mistakes?"[45] If the answer to this question is yes then there is likely to be more consumer support for a paternalistic intervention.

Some paternalistic regulations are likely to have more widespread consumer support than others. For example, while many consumers might balk at the suggestion that the government wanted to protect their health by banning the sale of fish and chips, most consumers are likely to be relaxed about the idea of being protected by the insertion of a mandatory cooling-off period into a credit contract or a door-to-door sale. Both forms of regulation are paternalistic. They seek to alter consumer behaviour in order to increase consumer welfare and they do so by restricting consumer liberty. Requiring a cooling-off period restricts consumers' liberty to negotiate a lower price for a contract with no cooling-off period. The important distinction between these two examples is that consumers are more likely, in the latter example, to consent to restrictions on their liberty.

Most consumers are also likely to support safety regulations for products such as bicycles, helmets, medicines, automobiles and airlines.[46] These safety regulations are paternalistic in that they take away consumers' liberty to choose their own preferred combination of price and level of safety. Because consumers all have slightly different levels of risk aversion, the level of safety required by the state will not match what all consumers would choose if they were perfectly informed and could accurately assess the risks. Nevertheless, consumers are likely to approve of the state setting standards that ultimately restrict their freedom of choice. They recognise that, if left to their own devices in an unregulated market, there is a possibility, even if they were given full information, they would miscalculate risks, not have the time or expertise to understand the implications of the information and fail to maximise their own welfare.

In other areas where consumer behaviour appears to decrease consumer welfare, the degree of consumer support for paternalism is likely to be more mixed. This variation in consumer support is particularly likely in circumstances where consumers are making trade-off decisions between short-term benefits and possible future negative consequences. In these scenarios there will be some consumers who want no paternalistic intervention, some who would consent to being nudged and some who might be in favour of more coercive attempts to control their behaviour. For example, suppose that government is concerned that many consumers are failing to save for their retirement. Some consumers are likely to see this as none of the government's business. Other consumers might support a default savings scheme from which they can choose to opt out. Other consumers might vote for a compulsory retirement savings scheme. Similarly, while some alcoholics might be against any government interference in their liberty to purchase alcohol, others might be in favour of laws that shorten opening hours for liquor stores.[47] In these grey areas there may be a great variation amongst consumers as to whether they even see themselves as making a mistake. In these cases, unless the potential harm suffered is extreme, it may be more appropriate to avoid hard paternalism and use either soft paternalistic measures or avoid any intervention.

Determining the degree of support for a paternalistic intervention will sometimes be challenging and may involve a degree of hypothesising. Moreover, using a "degree of support" as a relevant factor implies a rejection of the "revealed preference" model where consumer choices made in the market place are considered the most accurate way to gauge what consumers really want and what is in their best interests. In contrast, the approach recommended in this part of the chapter accepts that consumers do sometimes support their liberty being restricted. This is because in some situations, even if consumers know what actions are in their best interests, they understand that they are either unable or unwilling to take these actions without the state's assistance. If there is a domain in which it can be shown, either by direct evidence or reasonable supposition, that most consumers would support having their freedom restricted in some way, this will strengthen arguments that some kind of paternalistic regulation might be legitimate.

7.4.6 The degree to which consumers are dealing with large quantities of complex information they are unable to process

Consumers do not have the cognitive powers to process large amounts of complex information and make meaningful comparisons with alternative products in a realistic time frame.[48] This information-processing weakness has already been discussed above under the "magnitude of potential harm" factor (at [7.4.1]). It was put forward as one category of situation where consumers might cause themselves serious harm if legislative safety standards were not in place. Decisions about the safety of airlines, automobiles, buildings and medicine were used to illustrate areas of decision-making where consumers do not necessarily have the cognitive abilities or expertise to accurately calculate risk and make welfare enhancing decisions. The consequences of a poor decision in these cases could potentially lead to serious harm.

It is also worth considering this "information-processing" factor independently of the "degree of potential harm" factor. Not all cases where consumers must deal with large amounts of complex information is there the potential for major consumer harm such as permanent disability or death. Poor decisions in areas such as credit contracts or standard-form contracts might cause only minor financial distress.

Even when the degree of harm is relatively small, a paternalistic intervention might still be justified. This is because the cognitive weakness in processing information provides, in itself, a response to arguments against paternalism. Other cognitive weaknesses, such as over-confidence or a lack of willpower, require a policy maker to point to serious potential harm in order to establish a justification for intervention. In these situations if the harm is minor, governments are usually reluctant to interfere with informed consumer choice. The choice might be in some way distorted by the cognitive weakness, but it is nevertheless an informed choice.

In contrast, the information-processing weakness, by its very nature, provides an additional argument for paternalistic intervention, independent of arguments about the magnitude of harm. In cases where information is lengthy and complex, the consumer is unable to make an informed choice. This factor challenges anti-paternalist arguments about both liberty and welfare. Preserving consumer freedom of choice does not seem intrinsically worthwhile in cases where information-processing weaknesses mean that consumers are not so much making a poor choice as making no choice at all. Moreover, it makes no sense to say that consumer freedom of choice is the best method of maximising consumer welfare in cases where the consumer is not actively making a conscious, fully-informed choice. Even if suppliers are legally required to provide full information, if the information is too lengthy and complex it is unrealistic to expect consumers to take it into account when making purchasing decisions. In these cases the government might be in a better position than the consumer to make a welfare-enhancing decision on the consumer's behalf. The government has the potential to devote time and resources to processing the relevant information, calculating risks and determining acceptable minimum standards.

Standard-form consumer contracts are a good example.[49] In these contracts the details of the parties' rights and duties can be lengthy and complex. Even if these terms are written in plain English consumers do not have the computer-like information processing skills that enable them to quickly assess and compare these terms with competitors' terms. Most consumers do not even begin to try to examine these terms. There would not be many people in today's world who have not clicked "I agree" in a software licensing agreement despite having failed to scroll down the box of terms to read them. Typically consumers read only the terms that describe the price and the broad nature of the product. These are the terms that are of sufficient importance that consumers will read them and allow them to influence their decision to purchase.

Since most consumers are not making decisions based on unread terms there is no incentive for suppliers to compete on the basis of these terms.[50] The fact that most consumers do not bother to read most of the terms of a standard-form contract does not necessarily mean that they should be labelled as "irrational". Some economic theorists actually describe consumers' behaviour as "rational ignorance",[51] or as examples of "bounded rationality".[52] In other words, it would be irrational for a consumer to undertake the time-consuming and arduous task of reading, understanding, and assessing these terms and then performing a complex comparative analysis with competitors' terms. Moreover, most of the unread terms deal with risks that are unlikely to eventuate. For example, they specify what will happen if either of the parties defaults, or if the supplier wishes to terminate the contract or change the terms. Consumers are unwilling to devote the time and effort required to read and understand these terms.

The consequence of consumers not turning their minds to the desirability of each term of the contract is that the standard economic theory of free-market efficiency no longer works. The supplier has little incentive to ensure these unexamined terms are fair. This concern over the consumers' failure to negotiate non-core terms in standard form contracts has led Australia and the United Kingdom to introduce paternalistic legislation to ban unfair terms in these contracts.[53] New Zealand has not yet enacted unfair terms legislation.[54]

7.4.7 The degree to which the problem is affecting children, young adults or other potentially disadvantaged groups

As has been discussed above, the latest findings in behavioural economics suggest that consumers consistently suffer from cognitive biases and weaknesses that appear to reduce their ability to make welfare-enhancing decisions. These weaknesses and biases are likely to be more pronounced in children and young adults who may not have developed the maturity to accurately assess the costs and benefits of various purchasing decisions.[55]Cognitive weaknesses might also be more prevalent in certain groups of particularly vulnerable or disadvantaged consumers such as the elderly, illiterate, mentally impaired or sick.[56] If a paternalistic intervention is targeted at young people or other vulnerable types of consumers, then the anti-paternalistic argument that consumers should be left alone because they are rational welfare-maximising beings is potentially less persuasive.

Youth is a particularly pertinent characteristic when considered in combination with the "addiction" factor. Young people might choose to start consuming an addictive substance before they have reached sufficient maturity to make a fully informed and rational choice. By the time they mature into adult consumers they could have a serious addiction issue which impacts on their cognitive abilities to make rational choices that maximise their own welfare.[57] Minimum purchasing ages for alcohol and restrictions on tobacco advertising aimed at children are examples of paternalistic interventions that regulators justify by reference to the particular vulnerability of young consumers. An example from the United States is the paternalistic intervention introduced in 2009 to limit the availability of credit cards to under-21 year olds.[58]

7.4.8 The degree to which there are additional non-paternalistic reasons for enacting the law

In some situations of "poor" decision-making by consumers, it is only the consumers themselves that suffer harm. Other forms of self-defeating consumer behaviour have a negative effect on not just the consumer but also the rest of society. For example, the consumption of alcohol has the potential to cause direct harm to others including violence to innocent people, damage to property, car crashes and health and social service costs. The financial cost of alcohol consumption to the New Zealand economy has been estimated to be between $1.4 billion and $4 billion per year.[59] The broader social cost of alcohol consumption has been estimated at as much as $16 billion.[60] Similarly, other consumer problems such as smoking, gambling, obesity and over-indebtedness cause harm to people other than the consumer.

In these cases the government might be attempting to both reduce consumer harm (paternalism) and protect other people from being harmed by the consumer behaviour (non-paternalism). Having additional non-paternalistic goals can strengthen the case for an intervention that also has paternalistic goals. In each case, however, it is vital to acknowledge the goals of the intervention in order to determine whether the intervention is judged to have been a success. For example, assume that the objective of taxing tobacco is only to internalise the costs that smoking has for society. This is a non-paternalistic goal which aims to generate sufficient revenue from tax to cover the costs to society of cigarette smoking, such as extra expenditure by the public health system. Research in New Zealand shows that, in the case of tobacco taxation, the revenue generated by tobacco tax more than pays for the costs of tobacco use on society.[61] Therefore, it is a success in terms of achieving the non-paternalistic goal of covering the costs that the consumer behaviour has on other people. If, however, the government also has the paternalistic objective of changing consumer behaviour in order to reduce consumer harm, then any assessment of the success of the taxation policy needs to calculate the degree to which the taxation has decreased tobacco consumption.

7.4.9 The probability of non-legal responses such as education or support programmes, failing to provide solutions to the problem within an acceptable time frame

One key question that policy makers should ask themselves when determining the legitimacy of legal paternalism, is how likely it is that non-legal responses, such as education or support programmes, will solve the consumer problem in an acceptable time frame. The more probable it is that these non-legal measures will fail to fully solve the problem within a reasonable time, the stronger the justification for legal intervention (either on its own or in combination with the non-legal approaches). The amount of time that is considered reasonable will depend on matters such as the seriousness and irreversibility of the consumer harm.

In some situations, if given adequate support and education, consumers can learn from their mistakes and begin to make better choices.[62] Competition between suppliers would then respond by providing consumers with the products that improve consumer well-being. In general, consumers are unlikely to learn well in situations where they are making infrequent mistakes and the feedback is delayed and ambiguous. They are more likely to learn when the negative outcome is prompt, unambiguous and painful. So, for example, making the mistake of buying a terrible tasting sandwich is the type of mistake where consumers learn fast and the market responds accordingly without any intervention. However, much of the negative feedback for other "poor" choices such as smoking, abuse of alcohol or poor eating will not be experienced by the consumer until years after initial consumption. These behaviours are less susceptible to correction by quick, effective learning.

In some cases government engagement in an education campaign or support programme might be enough to guide consumers in the right direction and thus render regulation unnecessary. In many cases, problem behaviour is exacerbated or caused by social factors such as unemployment, poverty or ill-health. In New Zealand, the Māori population is overrepresented in these socioeconomic and health statistics.[63] Not surprisingly this means they are often overrepresented in consumer problems such as smoking and gambling. For example, currently 45 per cent of Māori smoke compared to 22 per cent of non-Māori.[64] Investigating the deep-rooted social and cultural factors that underpin "poor" decision-making should be a part of finding appropriate policy and legislative solutions.[65]

7.5 Examples and Analysis

Consideration of the various factors discussed in Part 4 can help policy makers determine whether a paternalistic legal intervention might be justified and, if so, where on the continuum of soft paternalism (nudges) to hard paternalism (coercion) such an intervention should lie.

The key concept at this stage of the decision-making framework is to find enough good reasons for intervention to outweigh concerns about interfering with consumer liberty. If there are not enough reasons to legitimise any interference then there is no point continuing to look at how to design effective regulation to alter consumer behaviour. If there is no legitimate reason for any paternalistic intervention, then effectiveness is irrelevant. The first question must always be whether there are sufficient justifications for reducing consumer freedom in order to improve consumer well-being. Higher levels of the factors discussed above point to a stronger justification for paternalistic legal intervention. Of course, people will disagree about what levels are needed to tip the scales against liberty and might also disagree on the comparative weight of the various factors. The important thing is that this difficult moral question is addressed.

Below is an attempt to show how different combinations of results might influence the decision as to whether a paternalistic approach, either hard or soft, is justifiable. The discussion is divided into three categories of situations where consumers seem to be making poor choices. Of course, in reality, there is a continuum of scenarios with a variety of complex middle-ground scenarios, rather than three distinct categories. Nevertheless it is helpful to consider the following broad groupings:

(1) circumstances where none, or almost none, of the factors point to a potential justification for legal paternalism;

(2) circumstances where all or almost all of the factors point to a potential justification for legal paternalism; and

(3) circumstances with mixed results.

7.5.1 Circumstances where none or almost none of the factors point to a potential justification for legal paternalism

There are many instances where consumers make mistakes, but none of the factors point to a reason for paternalistic intervention. For example, consumers often make the irrational, impulsive decision to purchase more vegetables and fruit at the supermarket than they will use during the week. The result is a lot of good food gets thrown away and consumers waste their money. Most people would agree that this is a problem but would not consider it legitimate to introduce a paternalistic regulation to alter consumer behaviour. The magnitude of potential harm is relatively low: the harm has a degree of reversibility; addiction is irrelevant; it is not a child-related problem; consumers are not dealing with complex large quantities of information; there are no obvious additional non-paternalistic objectives to support an intervention; and most consumers would probably not approve of any kind of restriction of their liberty in this sphere. In a scenario with low levels of most or all of the factors, the argument that consumers should have freedom to make their own mistakes and the opportunity to eventually learn from them is compelling. Even if, theoretically, some sort of regulation could be developed to prevent or discourage consumers from making these kinds of mistakes, the intervention is likely to be viewed as unjustified paternalism.

This kind of example shows that the call to view behavioural economics as some kind of all-round justification for paternalistic interventions in free markets is misguided. The mere fact that we can now confidently say that consumers universally make mistakes and misperceive benefits and costs does not automatically mean that the government is justified in intervening to discourage us from making these mistakes.

7.5.2 Circumstances where all, or almost all, of the factors point to a potential justification for legal paternalism

There are some consumer problems where all, or almost all, of the factors present a potential argument for a paternalistic legal intervention. If consumers are making choices in a free market that appear to have a relatively large chance of causing extreme and irreversible harm and consumers generally support some kind of regulation, then it is easier to make a case in favour of legal paternalism. If information-processing limitations are also an issue then this further strengthens the case for intervention. In these kinds of circumstances, a hard paternalistic measure might be justifiable. For example, safety regulations for airlines, automobile and medicines are forms of hard paternalism where consumer choice is removed and replaced by government decisions about acceptable safety standards. In these cases, consumers are generally happy to be relieved of the burden of freedom of choice and it is widely accepted that the government has the competence to improve consumer welfare.

7.5.3 Circumstances with mixed results

In other situations where consumers appear to be making decisions that are not in their best interests an assessment of the factors will produce more ambiguous results. Perhaps some factors point to a possible justification for intervention and other factors do not or perhaps the factors are all assessed at moderate levels. There are two distinct types of poor decision-making where results might be mixed. The nature of any grounds for a paternalistic intervention will be different in each case.

(a) Trade-off decisions in respect of short-term benefits and long-term costs

The first scenario is where consumers make a trade-off between short-term benefits and long-term costs (for example, smoking, obesity, poor retirement saving, gambling, borrowing too much and the like). These areas involve problems of self-control and a tendency to overvalue the present in comparison to the future.[66] The problem with intervening in these areas is that there is likely to be a huge variation in the degree of support that consumers would give to restrictions on their freedom. There may also be disparity among consumers as to whether they even see themselves as making a mistake. Many who are engaged in these seemingly self-defeating behaviours will want to be left alone to make their own decisions about their welfare.

Without a high level of consumer consent, hard paternalism is difficult to justify in these cases. One scenario in which hard paternalism might be justified is where the harm caused to consumers is extreme. For example, the severity of problems that can result from heroin and other hard drug consumption provides a possible reason for banning the sale of these products.[67] Similarly, the potential to be financially crippled by debt because of poor borrowing decisions might justify the requirement that creditors engage in responsible lending.

In other cases involving self-control problems, the harm caused might be less extreme and/or occur only when the product is overused. Hard paternalism is much more difficult to justify in these cases. If the harm is reasonably serious it might be possible to justify a coercive intervention that is specifically targeted at vulnerable consumers where addiction or youth is a consideration. Banning the sale of cigarettes and alcohol to under-18 year olds are examples of targeted hard paternalism.

It is more difficult to justify a generalised, non-targeted paternalistic intervention in response to a self-control problem if the harm is less serious and depends on usage. This is where the debate about freedom versus protection is likely to get most heated. These are the situations for which advocates of paternalism are more likely to suggest soft paternalistic measures that nudge consumers, rather than force them to make better choices. Examples include banning alcohol advertising, plain packaging for cigarettes and opt-out government savings schemes. Libertarian paternalists argue that these soft paternalistic measures are acceptable because they respect consumer autonomy.[68] They attempt to help consumers make welfare-enhancing choices but do not entirely eliminate freedom of choice. Of course, the paternalism will make it more difficult for consumers to make the "poor" choice, but Thaler and Sunstein argue that "[i]n its most cautious forms, libertarian paternalism imposes trivial costs on those who seek to depart from the planner's preferred option."[69]

The anti-paternalists respond that paternalism cannot be libertarian. It is only more or less intrusive.[70] They also claim that it is impossible for the policy maker to know that consumers would change their behaviour if they had complete information, unlimited cognitive abilities and no lack of self-control.[71] Further, individual consumer preferences and biases will differ between individuals and across time and this makes policy decisions about welfare complex and problematic.[72] Consequently, they argue, it is safer to leave consumers to make their own decisions about their own welfare.

Other scholars insist that when harm reaches a certain level it is facile to deny the welfare implications. For example, O'Donoghue and Rabin argue that:[73]

Economists will and should be ignored if we continue to insist that it is axiomatic that constantly trading stocks or accumulating consumer debt or becoming a heroin addict must be optimal for the people doing these things merely because they have chosen to do it.

The policy debate in this area must explicitly address the question of whether there are some situations where we can safely assume that consumers do make inferior decisions about their own welfare. The factors discussed in part four are crucial to this debate. Consumers who have self-control problems will not always want to be protected; it is therefore the other factors, especially the ones relating to the seriousness of consumer harm, which will be most relevant.

(b) Poor decision-making due to other cognitive weaknesses

Lack of self-control and "present bias" are not the only problems faced by consumers. Behavioural economics also suggests that consumers make "poor" decisions because of other cognitive weaknesses such as an inability to quickly digest complex lengthy information, a propensity to make impulsive decisions under pressure or over-optimism about the chances of a contract defaulting. These weaknesses can lead to consumer problems with mixed results on the factors. The seriousness of the harm might be at the lower end but, unlike the trade-off/willpower problems, the level of consumer support for paternalistic protection in these cases is likely to be much higher. Consumer behaviour that falls into this category includes entering standard form contracts with "unfair" terms, failing to negotiate a cooling-off period in a door-to-door sales contract or purchasing potentially defective goods or services without a guarantee as to quality.

The argument in favour of hard paternalism in these cases rests less on the severity of harm than the fact that consumers are genuinely making mistakes and are happy for the government to help them out. Cooling-off periods, prohibitions on unfair terms and prohibitions on contracting out of statutory guarantees as to quality are all examples of hard paternalism that might be justified in this way.

7.6 The relationship between welfare and preference

Beliefs about the relationship between "consumer welfare" and "consumer preference" will colour any debate about the legitimacy of paternalism. The goal of the paternalistic intervention is to maximise "consumer welfare"; but defining it is not an easy task. One of the challenges for policy makers is to determine to what degree consumer welfare overlaps with consumer preferences in any given area of consumer behaviour. There are three possible approaches.

(1) First is the neo-classical approach which argues that welfare should always be measured by reference to consumer choice.[74] So, if a consumer reveals a preference by making a choice, then this choice will maximise his or her welfare. Consequently, this approach does not advocate paternalism as a method of welfare enhancement. Where a consumer problem exhibits high levels on the factors discussed above this approach becomes less convincing and the second and third approaches seem to align more with reality.

(2) The second approach is advocated by libertarian paternalists.[75] It aligns welfare with "true" or "deep" consumer preferences that consumers find difficult to manifest without assistance. This approach emphasises the possibility that in some situations consumers make choices that decrease their welfare and that in an imagined world of perfect information, unlimited cognitive abilities and no lack of willpower, the consumer would engage in the behaviour that the paternalism is promoting. Anti-paternalists are extremely dubious of this attempt to discover "true" preferences.[76]

(3) The third approach is to make no reference to preference when assessing welfare. This method measures welfare by reference to predictions of experienced well-being rather than any conception of consumer preference.[77] Under this approach, welfare-maximising behaviour will usually, but not necessarily, be in line with consumer preferences (either revealed or stated). The main point is to nudge or force consumers toward what the government thinks is the best choice for the consumer, not what the consumer appears to prefer.

If policy makers are to justify a paternalistic intervention they will need to adopt either the second or third approach.

7.7 Effectiveness and unintended consequences

The strongest of all the arguments against the interference of the public with purely personal conduct is that, when it does interfere, the odds are that it interferes wrongly, and in the wrong place.[78]

If an examination of a consumer problem suggests that there are sufficient factors present to indicate that a paternalistic goal is legitimate, the next step is to consider issues of effectiveness and unintended consequences. The question at this stage moves beyond the philosophical one of whether or not it is legitimate to interfere with consumer liberty to a more practical question of whether the proposed regulation is likely to achieve its goals and whether there is a possibility of negative unintended consequences. The topics of regulatory effectiveness and unintended consequences are large ones and are touched on in other parts of the New Zealand Law Foundation's Regulatory Reform Project. This chapter does not intend to examine these issues in detail. Instead, it presents some introductory ideas about the importance of effectiveness and unintended consequences to decisions about the desirability of paternalistic intervention to protect consumers.

7.7.1 Effectiveness

There is no point enacting a paternalistic law that is unlikely to effectively achieve its objective of altering consumer behaviour. In a sense, "ineffectiveness" can be seen as the most obvious type of unintended consequence.

Many hard paternalistic measures, such as bans, are likely to be effective in achieving their goal of altering consumer behaviour because they severely restrict or remove consumer freedom of choice. If the sale of inexpensive, unsafe flammable children's nightwear is banned, for example, then consumers will no longer engage in the behaviour of purchasing this nightwear. If interest rates are capped then consumers can no longer purchase credit with an excessive interest rate. In other cases, even a ban on the sale of a product cannot completely eradicate the unwanted behaviour. For example, banning the sale of cocaine might stop some consumers from using the drug, but is unable to remove the risk that others will purchase it on the black market.

It is particularly difficult to determine whether soft paternalistic measures (where consumers retain the ability to make the "poor" choice) will effectively alter consumer behaviour. In these cases it is hard to predict if the chosen nudge, such as higher taxes, opt-out savings schemes, disclosure or advertising bans will influence consumer behaviour. Sometimes people respond to a paternalistic consumer regulation in ways that counteract the intended effect of the regulation. Research in the United States, for example, has shown that paternalistic regulations aimed at reducing obesity, such as calorific labelling and banning the sale of junk food in schools have had little or no effect on obesity levels.[80] Other research has shown that putting severe warnings on cigarette packets might just make a smoker feel guilty and reduce the pleasure from smoking but not actually change the person's behaviour.[81] It has also been suggested that taxing addictive substances that are used by a disproportionate number of low income consumers (such as tobacco and alcohol) might just raise money from the poor and not in fact reduce the consumption of the product.[82]

7.7.2 Unintended consequences

[O]ur ability to predict the full effects of governmental actions – much less the synergetic effects of hundreds of thousands of simultaneous government actions – is very limited. Far too often there are unanticipated results and costs, despite the most careful efforts of government officials.[83]

It is vital that policy makers assess whether a paternalistic intervention might have unintended negative consequences that outweigh any beneficial effects. In some cases, while the regulation might effectively reduce the targeted consumer behaviour, it might also unintentionally promote other harmful consumer behaviour. For example, one United States study has shown that a tax increase on beer was followed by a rise in the teenage consumption of cannabis.[84] Another study showed that while cigarette taxes reduce smoking, they are also associated with higher rates of obesity.[85] Opponents of interest rate caps claim that while caps will successfully prevent consumers from taking on debt with excessive interest rates, they might also cause low income consumers to be excluded from any affordable credit provision.[86]

When considering a non-targeted paternalistic intervention, the policy maker also needs to take into account any unintended but unavoidable costs on the rational consumer who does not need protecting. For example, alcohol taxation negatively impacts on responsible drinkers and heavy-handed credit regulation might prevent much desirable borrowing. The philosophy of asymmetric paternalism suggests that a paternalistic regulation is only acceptable if it creates large benefits for those who make errors while inflicting little or no harm on those who are rational.[88]

Some paternalistic regulations might also lead to unintended opportunity costs. For example, paternalistic tobacco regulation that successfully reduces smoking will reduce the opportunity for financial gain via tobacco taxes. This will be partly, but not entirely, made up for by a reduction in public health spending on smoking-related illnesses. Current New Zealand anti-gambling regulations provide for an interesting example of unintended opportunity costs of paternalistic regulation. The government have been negotiating a deal with Sky City casino that is currently being reviewed by the Auditor-General. That deal, before review, would repeal aspects of the anti-gambling legislation in exchange for Sky City's agreement to build a convention centre that will supposedly bring large economic benefits to New Zealanders (including hundreds of new jobs).[89] The retention of the paternalistic regulation will remove the opportunity to receive these economic benefits.

7.7.3 Using overseas regulatory models and research findings

Assessing the likelihood of a regulation being effective and the risks of unintended consequences is a difficult and potentially costly exercise. New Zealand is a small country with limited resources and it will sometimes be too costly for us to carry out the relevant research. Without adequate empirical knowledge there is the chance that the paternalistic intervention will be sub-optimal. One cost-saving option is to take advantage of relevant overseas research and to study the success or otherwise of overseas regulatory schemes. If an overseas regulatory approach has been shown to generate the desired results then this is encouraging. Moreover, importing tested overseas legislative models is less expensive than formulating New Zealand legislation from scratch.

Policy makers should proceed with caution when applying overseas research on regulatory issues to the New Zealand context however. Some overseas findings will not apply to the specifics of the New Zealand situation. For example, international studies have shown that alcohol consumption lowers when prices are higher.[90] Taxation of alcohol is therefore recognised internationally as an effective tool in reducing alcohol-related harm within the drinking population. This international research has led New Zealand economists to the logical conclusion that raising the excise tax on alcohol in New Zealand will contribute to reducing alcohol-related harm to our young people;[91] but a recent study of Australian and New Zealand students suggests that alcohol consumption by Australian and New Zealand young people might be relatively unaffected by price increases.[92] This study indicated that an increase in price of alcohol by as much as 25 per cent will not significantly reduce consumption. The conclusion was that taxation would have to be very high to result in a decrease in alcohol consumption among New Zealand and Australian youth. One possible explanation for this is that cultural norms around drinking might play a stronger part in New Zealand and Australia than in other parts of the world. The study relied on asking young people to imagine whether they would continue to purchase various alcoholic beverages at differing prices. The results might not translate into the behaviour of young people when faced with a real life price increase. Nevertheless, the study shows the potential dangers of applying the findings from overseas studies to decisions about New Zealand regulation.

In some cases, New Zealand's geographical isolation might provide the opportunity for effective paternalistic regulation that might not work as successfully in a less isolated country. For example, when New Zealand regulators introduce measures designed to reduce consumer problems such as gambling, smoking or alcohol consumption they do not have to contend with the possibility of consumers crossing nearby borders to continue their self-defeating behaviour in other countries or states. New Zealand's geographical isolation also places it in an extraordinarily good position to develop effective paternalistic regulation for our national online auctions because the transactions occur between New Zealanders rather than between consumers in different countries or different states as occurs on eBay in Europe or the United States.[93]

7.7.4 Evidence, experimentation and review

In some cases it will not be possible to predict with certainty that a paternalistic measure will effectively change consumer behaviour. Further, it is impossible to guarantee that there will be no unintended negative consequences from a paternalistic intervention. There might be no relevant overseas data or the available data might not translate well into a New Zealand context.

The New Zealand Government does not have large resources to devote to gathering empirical data for the purpose of evaluating regulatory effectiveness. However, even with unlimited resources it can be extraordinarily difficult to obtain clear evidence about the likely efficacy of a proposed regulation.[94] Unlike a drug trial, where efficacy can be determined by a randomised trial with a control group, it is challenging to develop useful trials in the context of regulation. Cause and effect are difficult to establish because regulations operate within complex cultural and sociological settings. In some cases consumers might only begin to change their behaviour in the face of a multi-faceted regulatory regime such as the one in place for tobacco.[95]

Faced with uncertainty, the regulator will need to decide whether the reasons for intervention are strong enough to justify experimental regulation. This decision will need to take into account that status quo bias, a key insight from behavioural economics, applies to government officials as much as consumers. In other words, once a regulation has been introduced it might be difficult to remove or change it. If experimental regulation is introduced it is crucial that a process is established for monitoring and reviewing the regulation in order to assess whether it has achieved the desired effects and whether there have been any unexpected negative outcomes.

7.8 Conclusion

The questions as to whether and when it is legitimate to create paternalistic consumer laws are of fundamental importance to New Zealand society and yet are easily neglected in the policy making process. This chapter has attempted to create an analytical framework for deciding when paternalistic interventions to protect New Zealand consumers might be appropriate. Paternalistic interventions are inherently controversial and good policy making requires that they are debated in full and supported by robust justifications.

The chapter recommends the consideration of multiple factors including: the magnitude, probability and irreversibility of potential consumer harm; the degree to which addiction is affecting consumer choice; the degree to which consumers want to be protected; whether the problem is affecting children or young adults; the degree to which consumers are overwhelmed by complex and excessive information; the degree to which there are additional non-paternalistic reasons for enacting the law; and the improbability of non-legal responses providing solutions in a reasonable time frame. This multifactorial approach acknowledges the complexity of the problem. In cases where an assessment of these factors points to a proposed paternalistic intervention being justifiable, policy makers should then consider the possibility of unintended negative consequences and the likelihood of the regulation effectively achieving its goal.

The suggested framework is not intended to easily produce definitive answers since quantification of the factors is difficult and ultimately a value judgment about the importance of liberty is required. Nevertheless, it provides a useful method for approaching the issue of paternalism in consumer law. It assists policy makers to address the relevant factors and to provide a well-developed justification for any proposed paternalistic intervention. Inevitably, any debate on this topic will be lively and differing points of view will result in different conclusions about the desirability of a particular paternalistic proposal. The decision as to whether the balance lies in favour of intervention will depend partly on the political views of the policy makers. Some will put more weight on the inherent value of freedom of choice and personal responsibility and others will emphasise harm minimisation and consumer welfare. There are no simple answers. Nevertheless, it is important that the policy debate focuses on the factors identified in this chapter rather than degenerate into a quarrel between those using the term "paternalistic" in a purely pejorative manner and those who think the term "consumer protection" provides a fully developed justification for any legal intervention in the marketplace.