Rethinking European Competition Law: From a Consumer Welfare to a Capability Approach

European competition law is predominantly focused on maximizing consumer welfare. This overarching purpose (which is supported by economic theory) leaves little place for safeguarding non-economic values, such as sustainability. This makes it difficult to allow cooperation between companies to contribute to such non-economic goals. In this article we explore whether it is possible to establish a different normative framework, in which such goals can be taken into account and can be balanced against the economic goal of consumer welfare. To answer this question, we take four steps. First, we discuss current EU competition law and the difficulty of fitting non-economic goals into the dominant interpretation of that law. Second, we propose a different normative framework, based on the capability approach advanced by philosopher Martha Nussbaum and economist Amartya Sen. Third, we argue that there are good principled reasons to incorporate non-economic goals into competition law. Fourth, we apply both the capability approach and the consumer welfare approach to three (illustrative) cases in which non-economic goals are at stake. Overall, we argue that the capability framework, although not without difficulties of its own, may provide a more legitimate theory for the interpretation of European competition law.


Introduction
To ensure the efficient functioning of markets, European competition law prohibits undertakings from entering into anti-competitive agreements and abusing a dominant position on the market. 1 The standard justification for European competition rules lies in the welfare gains of having competitive markets for consumers. Coordination of behaviour between undertakings drives up prices, which has a negative effect on consumer welfare. Competition law therefore serves to prevent a series of economic benefits (for companies) in order to realize another series of economic benefits (for consumers). However, companies sometimes enter into agreements to further a non-economic goal, such as rendering production facilities more ecologically friendly or strengthening social cohesion in inner-city areas. They may have different motivations for doing so, but in these cases the question arises: should European competition law take these non-economic interests into account? If so, how can this be done?
Since the 1990s the European Commission has increasingly adopted a consumer welfare approach to the interpretation of European competition law. In this paradigm, it is difficult to give non-economic interests their due, as we will show by discussing three recent cases in which economic and non-economic goals clash (Section 2). While the consumer welfare approach has been criticized, it is a challenge to come up with a constructive alternative. In this article, we propose the capability approach as an alternative framework for interpreting competition law. This approach was developed by economist Amartya Sen and philosopher Martha Nussbaum as an alternative to utilitarian approaches in welfare economics and theories of justice. We will introduce the main features of the capability approach and propose how it can be applied to competition law, focusing on two key issues: the identification and the weighing of capabilities in a competition case (Section 3). Third, we argue that there are good principled reasons to incorporate non-economic goals into competition law, both in terms of legal interpretation of the relevant EU texts, moral arguments about the appropriate division of labour between private companies and political bodies, and in terms of political considerations about the legitimacy of decision-making by competition authorities (Section 4). Fourth, we compare how both the capability approach and the consumer welfare approach examples below will illustrate -it is generally understood that non-economic interests are benefits that are either not directly included in the consumer welfare standard, or are not (reliably) economically calculable.
In contrast, a broad consumer welfare standard is somewhat more open to including non-economic interests. Such a standard is used for example by the Dutch competition authority in specific cases (as we will see below). It incorporates (easily) calculable non-economic benefits that are not directly related to the product in question but that do accrue to the consumers of these products. Thus it includes these non-economic benefits indirectly, to the extent that they pertain to the same group of consumers as the ones suffering price increases. The example of the Energy Agreement, below, will illustrate such an approach in practice. However, it is unclear just how strongly these benefits need to relate to the same group of consumers, in the same market. Clearly the further this is stretched, the more controversial the interpretation of the consumer welfare standard becomes.
Although not used in competition law, we can imagine the use of a third welfare standard: an inclusive welfare standard (the label is ours). 7 In contrast to the consumer welfare approaches, such an approach takes non-economic interests directly into account, and does not require a link to the same group of consumers. It does so by using a broader conception of 'welfare' and adding (measured and quantified) non-economic benefits to the assessment mix. Such a standard is based on the idea that anything can be expressed in terms of preferences and thus be quantified and compared, 8 by asking about consumers' preferences and using a willingness-to-pay quantification in monetary terms. 9 This does not answer the question of whether one should do so. 10 Although this standard does include non-economic interests, it does so by using a market-based concept (subjective preference satisfaction) as its starting point. We applaud this standard's principled decision to include non-economic interests (see Section 4). Nonetheless, it is our contention that to take full account of the value of public interests when assessing agreements between companies taking responsibility for public interests, we would need to go beyond such an inclusive welfare standard and introduce a non-welfarist standard, such as the one offered by the capability approach. But this is running ahead of the argument.
We will now first illustrate the difficulties of using a consumer welfare approach (narrow or broad), by introducing several competition law cases. In each of these cases non-economic interests present a problem for competition law, and the question is if and how to take them into account. These illustrations will come back in Section 5, where we will further discuss the application of the different standards.
A first illustrative case relates to the Dutch Energy Agreement (Energieakkoord), an agreement on a roadmap towards more sustainable energy in the Netherlands in 2020. Parties to the agreement include energy producers, distributors, the Government, and advisory bodies to the Government (Social Economic Council). They agreed to the (accelerated) closing down of five coal-fired power plants, which leads to less supply of electricity and to lower emissions of noxious gases and particulate matter. It is on this element of the agreement -in isolation -that the Dutch competition authority (Autoriteit Consument en Markt, ACM) provided an 'informal opinion' (a non-binding preliminary assessment), applying (also) European competition law. The ACM labelled the closing of coal-fired power plants a restrictive production agreement in violation of competition law. 11 Animal welfare cases provide a second class of illustrations of non-economic interests. For example, under ongoing consideration is an agreement in the Netherlands relating to intensive pig farming (the Verbond van Den Bosch). 12 Parties to the agreement include the provincial government, farmers, animal-fodder producers, veterinarians, and supermarkets. This agreement includes lowering the use of antibiotics in animal husbandry, which results in less resistance to antibiotics in humans and provides for greater animal welfare. But it might also lead to higher consumer prices. In this case there are both human health (i.e. consumer welfare) and animal welfare benefits, so that a broad consumer approach would give consideration to some of the benefits (i.e. those relating to humans) of the agreement. But we can easily imagine cases where there are only animal welfare benefits. For example, think of protecting the habitat of a species of birds not even on the brink of extinction. Suppose that dairy farmers, using their grassland for grazing cattle, in collaboration with nature conservation societies, come to an agreement to protect the black-tailed godwit (a protected species). Although not specifically beautiful, it makes an impressive melancholy sound, especially in flocks; but it nor its eggs are generally used as food. The effect of such an agreement might be higher prices of dairy products, but also the flourishing of the black-tailed godwit. 13 How to decide such a case? A third illustration concerns efforts to combat binge drinking. This phenomenon has become more widespread since the price for alcoholic drinks has decreased relative to other consumptions. Supermarkets in the UK and Ireland have been urged to take responsibility to counteract binge drinking, because of its health risks for young people, but also because it entails healthcare costs and police effort. However, these supermarkets are 'trapped in a downwards spiral of price-centric competition'. 14 Therefore an important chain of supermarkets in the UK -Tesco -proposed to reconsider the pricing of alcohol, but only if done together with its competitors. Such a pricing agreement would clearly go against competition law, at least if the possible benefits (avoided healthcare costs, lower policing costs, and better inner-city atmosphere and liveability) are not taken into the assessment. Legislative action was then considered, by setting a minimum price for alcoholic beverages. 15 In all of these cases it proves difficult to take non-economic interests into account in a systematic way if a consumer welfare standard is used (see also Section 4). It is for this reason that we will now propose an alternative approach.

The capability approach as alternative to resourcist and utilitarian approaches
The capability approach was originally developed by economist Amartya Sen and philosopher Martha Nussbaum. The capability approach is a flexible framework, which has been used in a variety of contexts. Sen's aim was to propose an alternative view of welfare, compared to standard welfare economics, which is based on the moral philosophy of utilitarianism. 16 Nussbaum's aim was to propose an alternative to utilitarian and resourcist theories in the area of theorizing about distributive justice. 17 The basic idea of the approach is that we should start from the fact that individuals have 'capabilities to function' in a specific way. A function or functioning is a 'being or doing' of a person. Eating, sleeping, driving, walking, being healthy, being educated etc. are examples of functionings. A capability is an ability or opportunity to function in a certain way. I have the capability to be well-nourished if I have the opportunity (given my income and access to food markets) to consume a nutritious diet. I may decide not to use my capability; I may choose not to eat today but to fast. But, as Sen has often argued, having such a personal choice is important. A fasting person, but not a starving person, has the opportunity to reach a sufficient nutritional intake. He has the freedom to Utrecht Law Review | Volume 12 | Issue 1 (January) 2016 choose how to function. We can then evaluate the level of welfare or the justice of a society by looking at how well it succeeds in guaranteeing each citizen a set of basic capabilities. This perspective differs from its competitors, resourcism and utilitarianism. Resources are inputs of any person's capability set. We need food to have the capability to eat, we need healthcare services to have the capability to be healthy. Resourcist theories define welfare or justice by looking at each person's possession of a basket of resources. On a macro-level, evaluating countries' performance by looking at the average resource bundle per person (GDP per capita) is an example of a resourcist approach. However, resources do not translate into capabilities in a straightforward manner. Two persons, one healthy and the other disabled, may need to receive different resource bundles to reach the same set of capabilities, given differential personal, social and environmental conversion factors. Therefore it is important to evaluate economic processes by looking at the actual set of functionings individuals are able to realize (i.e. their capability set), instead of merely at their resource bundle. 18 At the other end of the process, individuals derive a certain utility or well-being from their functionings. However, here too there is no one-to-one correspondence between a capability set and one's utility level. Two persons with the same capability set can experience different levels of well-being. If social policy focuses on utility levels, then it will be held hostage by persons who need champaign and caviar to reach the same level of well-being as their neighbours who are content to have fish and chips; or it can miss the urgency of poverty where people have adjusted their expectations and are happy to live on a low level of functionings achievement. The capability approach avoids these opposite pitfalls, ignores both resource inputs and utility outputs and steers our evaluations to the capabilities people actually have.
Sen and Nussbaum have also stressed that the capability approach can be seen as a human rights approach. People should have rights to basic human capabilities. 19 States are made responsible for the capability level of their citizens. This provides an important connection to the field of law, including competition law.

The capability approach in the context of competition law
For the connection between law and the capability approach to work, two important issues need to be resolved. First, which capabilities are the basic capabilities that people should have a right to? Sen and Nussbaum have given different answers, where Nussbaum has formulated a list of ten central capabilities, 20 and Sen thinks this selection process is best left to democratic processes. 21 Second, how to weigh them against each other and make trade-offs? These two questions, as to selection and weighing of capabilities are still very much open to interpretation and further development in the capability approach. They are also key to the application of the approach to competition law. 22 As a necessary first step, all capability interests of the parties that have a stake in a specific competition dispute would need to be identified. To tailor the approach to the competition-law context we propose to conceptualize these interests as consisting of three distinct types of capability sets: market-constituting capabilities, consumptive capabilities and third-party capabilities.
First, market-constituting capabilities consist of the capabilities to hold property and to contract, i.e. the ability to engage in voluntary exchanges to obtain goods and services. These capabilities are constitutive Utrecht Law Review | Volume 12 | Issue 1 (January) 2016 for the institution of the market; markets could not exist without them. 23 The capability approach can and arguably should recognize the two basic market freedoms -to hold property and to contract -as important capabilities that are valuable for citizens to possess. 24 These capabilities express the value of freedom of choice in the economic context. 25 This does not mean that the capability approach is fanatically pro-market, because as we will see the approach will also recognize other valuable capabilities, which will need to be weighed against these market-constituting capabilities. Nor is the approach anti-market. The capability approach will assess the legitimacy of specific markets on a case-by-case basis. 26 In a competition case, market-constituting capabilities have a certain weight because they express the producers' freedoms to conclude agreements with each other as they see fit. These freedoms may obviously come into conflict with the two other sets of capabilities, which we will now address.
Second, consumptive capabilities refer to a set of widely different basic capabilities that are essential to human flourishing: capabilities for health, education, nourishment, housing etc. These we call consumptive capabilities because each of these capabilities in turn requires (amongst other things) consumers' ability to purchase those goods and services which are necessary to realize them. The market is instrumental to the production and exchange of these goods: it may, or may not, be the best institution to achieve them. From a capability perspective, given the scarcity of resources, the most efficient system of providing these goods should be chosen, so that the total set of capabilities can be realized at least up to a threshold level. 27 In competition cases, this means that a consumers' interest in consumptive capabilities (say, to heating houses and cooking food) translates into an interest in getting market goods (say, gas and electricity) to be delivered to them at the lowest possible price; given the fact that this will leave them with as much budget as possible to satisfy their other basic capabilities (later in this section this claim will be qualified).
Third, we use the label of third-party capabilities to refer to capabilities of others than consumers or producers. Here non-economic interests enter the capability approach. In the Dutch Energy Agreement, for example, these are capabilities of future generations which stand to be harmed by damage to the climate if CO 2 emissions are not reduced in time. 28 In the case about intensive pig farming, animal capabilities are at stake. 29 In the case of binge drinking we need to deal with capabilities for health of (inter alia) young people, as well as capabilities for security of inner-city inhabitants. These examples are not exhaustive: there is a wide variety of possible third-party capabilities. Now, we do not claim that it is uncontroversial that these third-party capabilities should be recognized. But we do argue that competition law should make room to discuss and possibly incorporate non-economic interests (Section 4) and show how this can be done by conceptualizing these interests in terms of capabilities (Section 5).
The second step in any application of the capability approach to competition law will be to weigh the capability interests at stake. There are three theoretical issues here.
A first issue is how much weight every capability should be awarded. Capabilities cannot be weighed using market prices because such prices rely on each individual's purchasing power and willingness to pay. Imagine two different cases. In case X the animal capabilities of a group of chimpanzees must be weighed against consumers' capability to purchase life-saving healthcare medication, while in case Y the same animal capabilities must be weighed against the consumptive capability to purchase a new brand of cosmetics. Quite apart from the size of the respective groups (suppose the numbers of animals and consumers are the same for X and Y), the value of the consumptive capability will be different in a different product market. This is an important methodological difference with an economic approach, which simply counts preferences and Utrecht Law Review | Volume 12 | Issue 1 (January) 2016 is blind to the differential value of different capabilities. This means we need alternative ways to establish weights. In the end, whatever methodology we choose, the actual weighing process needs to be done in an institutional context that is democratically legitimated (see Section 4).
A second issue is the distributive principle guiding such a weighing exercise. The analogy with economic reasoning suggests maximizing the total level of capabilities. Let us assume that in a competition case there are four parties, A & B (producers) and C & D (consumers). We will assume that if we allow an agreement between A & B to stand this raises their capability levels to 15, while C & D's capability levels fall to 10. In the situation where the agreement is prohibited as a violation of competition law, A & B will experience a capability level of 5, and C & D of 25. The second situation, then, realizes a higher total capability level (30), compared to the first situation (25). However, as Nussbaum proposes, each basic capability should be realized up to a threshold level. If the threshold were put at, say, 8, then the second situation is problematic because it allows the capability level of A & B to fall below the threshold. Using a threshold principle means opting for the first situation would be the right choice, because there all parties' capabilities are above the threshold. This principle would be important in competition cases as well, as we will show, and holds that it is more important for each person to have a sufficient level of every basic capability, than to have a maximal aggregate level. For example, the value of the capability to contract is that it allows freedom of choice. The capability to contract is more meaningfully realized to the extent that consumers are able to choose from a larger variety of contract partners. Other things being equal, a choice is more meaningful when there are more alternatives to choose from ('your money or your life!' is not an attractive choice situation to be in). However, at some point, there are diminishing returns. To be able to choose from 50 brands of washing powder arguably is not much more valuable than to be able to choose from 30 brands.
Other basic capabilities exhibit a similar structure of diminishing returns. Most pertinent in the competition context is that thinking in terms of thresholds is antithetical to the economic objective of maximizing consumptive capabilities (and the accompanying effort of competition authorities to ensure the lowest possible price on the market). When the consumers' budgets are sufficient to satisfy all their basic consumptive capabilities then any further protection of consumptive capabilities only works to bring them even higher above the threshold than they already are. 30 Such increases above the threshold are less valuable than increases to bring them (or anyone else) up to the threshold level. When a price-increasing producer agreement serves to protect third-party capabilities (which we imagine have not yet reached the threshold) and thereby increases prices so that it decreases consumer capabilities (but still keeps them above the threshold), then the first may have more weight in the overall judgment than the latter.
A final issue is whether the weighing act should take a quantitative form (let us call this type of weighing 'calculation'), based on the model of an economic cost-benefit analysis, or a qualitative form (let us call this type of weighing 'balancing'). Some authors argue that capabilities can be measured. 31 If that is the case, then there are no principled objections to a quasi-economic weighing of capabilities. Other authors, however, follow Nussbaum in stressing the incommensurability of capabilities: each of them is of separate value and these values cannot be compared in terms of a common metric. 32 In the absence of commensurability, we can make choices, but they take the form of a discursive style of reasoning in which different types of interests are roughly balanced against each other. This type of balancing is also familiar in legal literature. 33 Our line on this is that we should proceed with care when we try to quantify non-economic interests. It may be possible to quantify where we mainly need to balance price-related consumptive capabilities to price-related market-constituting capabilities. But when non-economic interests come into the picture, quantifying would often present a distorted picture of the nature or magnitude of the interests at stake. 30 This may be the case for citizens in rich and sufficiently egalitarian countries. Any criticism that a price increase makes some groups unable to satisfy their basic capabilities can also be resolved by general redistribution -it does not necessitate prohibiting the price increase.

The argument for inclusion of non-economic interests in competition law
The previous section has argued that if we want to include non-economic interests into competition law assessments, the capability approach provides a good normative framework for doing so. But should we include these interests in the first place?
As the cases mentioned in Section 2 illustrate, undertakings, sometimes co-jointly with pressure groups and/or government bodies, may decide to take responsibility for a public interest and join forces in improving (parts of) their processes. Of course, such initiatives generally will not be against the self-interest of the companies concerned either, resulting in good public relations, a well-known and outstanding reputation or profit in a new market. 34 In competition law this type of cooperation is at odds with the competition law paradigm of individual agents, seeking self-interested profit maximization. In the discussion on the goals of competition law and standards to be used in assessing competition cases, arguments have been made against incorporating non-economic interests in a competition law assessment. We will review the three main arguments, which we have labelled the legal, moral and political argument. 35 We will reject each of them and provide reasons for including non-economic interests in competition law.
First, the legal argument holds that the relevant legal texts do not support inclusion of non-economic interests in a competition law assessment (though often this argument is enmeshed in the moral argument, presented next). 36 In the European context these texts are the Treaty on the Functioning of the European Union (TFEU), secondary legislation and policy documents of the Commission. The reason to focus on consumer welfare is, firstly, because the exception of Article 101(3) TFEU includes the requirement of a fair share of benefits accruing to consumers. Secondly, because in secondary legislation and accompanying guidance the Commission focuses on consumer welfare, 37 and because the Commission is an important actor in shaping and enforcing competition law, these latter documents carry legal and interpretational weight. 38 These arguments in our view are not convincing. On the contrary, the legal texts at least leave room for, but more importantly require, a broader interpretation. A coherent reading of the Treaty provisions, including its general goals, 39 and its integration clauses, 40 does not lead to any single goal (and therefore: standard) of competition law. Instead, several, sometimes conflicting, goals are possible. The Court's judgments support this view. 41 Therefore there are good legal reasons to include non-economic interests in competition law. 42 Second, there is the moral argument that private parties should concentrate on profit maximization and not try to realize social ends that lie outside the market. This argument presupposes a particular division of labour between the private and the public realm. Private actors concentrate on their private objectives, and in doing so ensure the optimal functioning of the market mechanism. Public actors legislate in favour of public objectives, and in doing so ensure a level playing field of publicly defined constraints under which market agents pursue their own interests. This argument, then, rests on two related grounds. On the one hand, it is concerned with the optimal economic functioning of the market, in terms of consumer welfare. Utrecht Law Review | Volume 12 | Issue 1 (January) 2016 Collusive agreements between private actors are a threat to this and competition law aims to restore the market's functioning by prohibiting them where they distort competition. On the other hand, the argument is concerned with the coherence of different parts of the legal system. Competition law has a different function from other areas of law. Whereas other areas may constrain market actors in order to realize noneconomic objectives, competition law constrains market actors to make the market function on economic terms. Thus, competition law should not be overburdened and made incoherent by also having to deal with these non-economic objectives. 43 The reason for calling this a 'moral' argument is that the upshot of this division of labour between private and public actors (and between different parts of the legal system) is that market actors cannot assume their moral responsibility in the same way in which this is possible for private persons operating outside a market context. A group of producers who believe it is their moral responsibility to tackle a social problem -as a group -is unable to do so because they are declared part of a system which is only geared towards one goal, consumer welfare. 44 In the field of business ethics, this position -that firms do not have a moral responsibility -is very controversial, to say the least. Most authors there argue that firms do have at least some moral responsibilities, although of course the exact ground on which this is argued as well as the scope of these obligations are under much debate. 45 The capability approach has been used to this extent as well. 46 One important reason for assuming moral responsibility is that legislators are often too slow or at too far a distance from practice to do be able to effectively regulate social problems. 47 On a global level it is also possible that there effectively is no government that can regulate. 48 The emergence of movements for corporate social responsibility in the last decades is a sign that cooperation and initiatives of private actors are necessary in an era of declining public capacity to tackle problems effectively. To prohibit agreements to reach non-economic goals would mean to drain private actors' responsibility for them.
Perhaps the real reason behind the reluctance to grant moral responsibility to private actors lies elsewhere: with the consequences it would have for the authorities having to evaluate these private actions when being called upon to settle disputes in the context of competition law. This brings us to the third and political argument: that a weighing of these interests which each have a very different character -a market interest against a non-market interest -should not belong to the legitimate authority of administrative agencies. 49 This legitimacy argument is serious: it holds that it is exclusively up to democratically elected parliaments to pass legislation that would close down coal-fired power plants, prevent binge drinking by providing a minimum price, protect animal welfare by providing detailed regulations, or impose on farmers obligations to protect the black-tailed godwit.
We disagree. 50 First, it is no less controversial to give exclusive recognition to economic interests in situations where non-economic interests are also at stake, than to acknowledge both sets of interests. In our view, therefore, it is the legal decisions by competition authorities that exclude non-economic interests which lack legitimacy. Once brought forward by one of the parties in the dispute, it is arbitrary to exclude non-economic interests to the advantage of the economic interests at stake. In political philosophy, it is Utrecht Law Review | Volume 12 | Issue 1 (January) 2016 widely recognized that only a democratic process which includes all interests at stake can gain legitimacy. 51 From this perspective, the political argument must be turned on its head. Moreover, if the weighing of an administrative agency dissatisfies groups of citizens and politicians, then nothing prevents law-making as a way of correcting those judgments and making regulations which express a different weighing of the interests for future cases. The explicit weighing of these interests by a competition authority will serve as a trigger for law-making authorities, who can still have the last word on the matter if they want to.
Second, we should not underestimate the potential for democratic elements to be added to the decisionmaking procedure at the administrative agency itself. For example, the public at large could participate through consultations. But more promising is to include elements of deliberative democracy, by involving a representative group of the public in the decision-making procedure where non-economic interests are raised. In this forum of debate, where all points of view are discussed and weighed on their merits, without the power base of the discussant being the defining factor of acceptance, 52 a balancing of interests would be at stake. 53 This means calling upon the public not as consumers, but also as citizens. 54 Judicial appeal of such a decision would ultimately place in the hands of the judiciary the question of the correct interpretation of the legal provision, its application to the facts of the case, and review of the procedures established to safeguard legitimacy of the decision-making process.
In conclusion there are legal, moral and political arguments for including non-economic interests in a competition law assessment. This brings us to the next and final step in our argument: applying the capability approach to the cases introduced above.

Applying the capability approach
It is one thing to present general arguments -as we have done in the previous section -stating that noneconomic goals should be taken into account in competition law. It is quite another thing to show how this can be done. Therefore in this section we will show how the capability approach can be used in competition cases, using the three cases presented in Section 2: the Dutch Energy Agreement, the animal welfare case and the binge-drinking case. We will start by discussing these cases under the currently dominant consumer welfare approach, which as we saw does not incorporate non-economic goals (5.1). Then we will show how the capability approach can be applied, by making use of the framework introduced in Section 3 (5.2). Finally, we will show the differences between the capability approach and its economic cousin, i.e. the inclusive welfare approach (5.3).

Applying the consumer welfare approach
Above we distinguished two versions of the consumer welfare approach. Under the narrow consumer welfare standard, only direct effects on the consumers of the product in question can be considered, especially the price effect on consumer welfare. Under this standard all three agreements of our example cases would be prohibited. The Dutch Energy Agreement concerns an agreement between energy companies and other stakeholders to move towards sustainable energy production in the Netherlands. A specific agreement to close down noxious coal-fired power plants was included in this overall agreement. Economic theory suggests that taking production out of the market will lead to a price increase (because of the decrease in supply). 55 Similarly, in the animal welfare cases higher production costs lead to an increase in the price of pork and an increase in the price of dairy products as a result of agreeing to protect the black-tailed godwit. On the impact of this difference also Lewinsohn-Zamir, supra note 10. 55 Although the parties to the Agreement argue that an increase in sustainable energy production and supply -also resulting from the Agreement (but not a direct result of the closing of the coal-fired power plants) -will balance this effect on supply. This was not taken into account by the competition authority in its preliminary opinion.
Utrecht Law Review | Volume 12 | Issue 1 (January) 2016 will also increase prices. All these agreements therefore lead to lower consumer welfare. Under the narrow standard, nothing counterbalances this effect, and therefore these agreements violate competition law.
In contrast, a broad consumer welfare approach allows for weighing these negative effects on consumer welfare against quantifiable benefits of these agreements for consumers. For example, the avoided healthcare costs of consumers of energy, pork (by ingesting less antibiotics), or alcohol (by drinking less) are calculated and taken into account. These benefits apply to the same group of consumers who suffer from price increases, albeit that the benefits occur on different markets. This approach was actually applied by the Dutch competition agency ACM in the Energy Agreement case. All consumers are affected by an increased price, but all also benefit from avoided healthcare costs of lower emissions of noxious gases. However, the ACM limited its acknowledgment of the benefits of avoided healthcare costs to the lowering of the emission of NO x , SO x and particulate matter, and -controversially -excluded the lowering of CO 2 emissions. 56 The compounds NO x and SO x play a role in creating harmful particulate matter, smog and acid rain and have an effect on health (especially on the respiratory system), on visibility, on corrosion (of metals and statues), and on the environment and the ecosystem. A lowering of these emissions has both an immediate local effect on health and an immediate and long-term effect on the environment, locally and globally. As for particulate matter, these very small particles contain solids or liquids that can enter the lungs and thus result in serious health problems, as well as leading to environmental damage. Thus a lowering of particulate matter has a direct local effect on health, as well as a more non-local effect on the environment. The ACM calculated the healthcare benefits of these three types of emissions and weighed them against the welfare loss caused by the agreement (its conclusion was that the latter weighed heavier; thus the agreement was still declared to be in breach of competition law).
The energy case shows how the broad consumer welfare approach can take into account non-economic goals, such as health, but only in an indirect way: by calculating avoided healthcare costs. Similarly, it is possible to present the advantages of preventing binge drinking as a matter of avoided healthcare costs of the youths involved, just as it is possible -though the chain of causality is stretched -to include the notion that using less antibiotics to raise pigs leads to lower human healthcare costs. However, there is a difference in valuing health in itself, as a capability (as we will show below) and valuing the avoidance of (quantified) healthcare costs. The limits of the broad consumer welfare approach become even more apparent in the non-consideration by the ACM of the long-term ecological effects of the Energy Agreement. Greater resilience of the ecosystem, for example, cannot be accounted for in terms of avoided healthcare costs. Furthermore, a consumer welfare standard in competition law generally has difficulty with long-term effects and favours short-term consumer benefits. 57 The consumer welfare standard also falls short when we consider the animal welfare case as there is no consumer welfare effect from increasing an animal's living space or protecting the black-tailed godwit. In the latter case one would need to stretch the notion of consumers to include a market for leisure, where consumers benefit from being able to hear and see different birds when hiking or biking, as these benefits do not necessarily accrue to the same consumers as those paying the higher price (on the dairy consumption market). All of this requires heavy conceptual gymnastics from the consumer welfare approach. Where benefits do not accrue in any straightforward manner to consumers, this standard therefore fails to integrate non-economic goals into a competition law framework. 56 Lowering CO 2 emissions would normally result in lower healthcare costs (locally), and abate the greenhouse effect (globally and in the longer term). But because CO 2 emissions are regulated by the ETS, the ACM argued that any reduction of CO 2 emissions from the closed plants would not necessarily lead to reduction of total CO 2 emissions: the ETS is a system of trade in CO 2 emissions, which caps total EU emissions. Plants emitting CO 2 have an 'allowance' to do so, which, if not used (or not up to its maximum), can be traded. From the (publicly available) text of the Agreement it is unclear whether the parties intended, or are able to take their resulting allowance off the market completely or to trade this under the ETS. Thus, there may be no effect at all. Of course, if the parties to the Agreement were to state that the CO 2 emissions were to be removed from the ETS altogether, as has been suggested as a possibility, the resulting avoided healthcare costs would also have to be taken into account. 57 See e.g. Cseres, supra note 3, p. 167.

Applying the capability approach
A capability approach would start by taking into account the three groups of capabilities mentioned in Section 3. First, consider the market-constituting capabilities for property and to contract. The capability for property of the energy production companies relates to their ownership of coal-fired power plants and other production assets. 58 This capability is in itself not limited by competition law. By contrast, the capability to contract is directly limited by competition law: concluding agreements restricting competition is prohibited. Prohibiting an agreement to diminish production capacity means limiting the energy producers' capability to contract. Interestingly, the market structure of this case means that this has a negative impact on the capability for property as well. For in this case there is a 'first mover disadvantage': a change to sustainable production by an individual producer is not feasible, because such a production mode is more expensive, leading to higher prices, so that consumers will switch to the producer offering a lower price for a less sustainable product. It is therefore unfeasible for one individual producer to take the lead. 59 If the alternative route of collective agreements is prohibited by competition law, then the capability for property is limited as well. More precisely: no capability is left to use one's property to produce in a sustainable manner. In the same vein, the market-constituting capabilities of producers of pork and dairy products, and those of supermarkets fighting binge drinking are at stake in the other cases.
Both the consumer welfare approach and the inclusive welfare approach fail to take into account these capabilities of producers. The capability approach is broader in that it values the freedoms (opportunities) of market participants both instrumentally and intrinsically: both for the consumption capabilities that market agents provide to each other, and for their abilities to enjoy and exchange their property. 60 This intrinsic valuation of producers' market freedom may be treated with suspicion by many in the competition context, given its historically primary aim of countering producer power to enhance consumer welfare. This historically grown focus is understandable, but in cases where non-economic interests are involved, we see its problematic implications. Producer capabilities cannot be taken into account, even where undertakings aim to protect a widely-valued public interest. Of course, this does not mean that market-constituting capabilities will be the only valuable thing on the table (such as in libertarian approaches which exclusively rely on the value of market freedoms) when using a capability approach. But they do need to be recognized and included as valuable capabilities, and then balanced against the other two classes of capabilities.
Second are consumptive capabilities. In the Energy Agreement case, these are all capabilities that need the supply of electricity: cooking, warming, cooling, washing, lighting etc. 61 As a result of the Energy Agreement these capabilities are negatively impacted, as the goods necessary to realize them become more expensive. Less total consumption is possible, leading either to less consumption of energy or of other goods. Similarly, in the animal welfare and binge-drinking cases, there are clear consumer capabilities at stake, which are negatively affected by any price increases as a consequence of producer agreements. These consumer interests are also included in the assessments of the consumer welfare and the inclusive welfare approach. The difference with the capability approach is that the inclusive welfare approach evaluates consumptive capabilities differently once a threshold is reached (see Section 4). As long as basic/core capabilities can still be met (are still on a level above this threshold), individuals have sufficient capabilities. An increase of the price of energy, which does not result in a situation where basic capabilities cannot be met above the threshold, is therefore less problematic. If there are weighty (third-party) capabilities on the other side of the balance, this might lead one to accept a decrease in consumer capabilities due to price increases. The consumer welfare approaches obviously would not agree with this: they do not take the remaining budget of consumers into account when thinking about the value of price increases. But the capability approach can 58 The capability for property also concerns the allowances of emissions, but this seems to us a derived capability that is limited by competition law: it is not competition law itself that prohibits individual companies to remove their allowances from the ETS. Jointly deciding not to trade the resulting allowances back into the ETS would make a stronger case for the Energy Agreement from all perspectives, however. 59 On the issue of this collective action problem as a basis for cooperation in responsible business conduct see D. discriminate between situations above and below a threshold, and take into account that a given price increase for a generally affluent population is not the same thing as the same price increase for a poorer population. Finally, third-party capabilities. As mentioned earlier, in this category we find a wide variety of capabilities. In the cases under discussion in this article we find four examples of third-party capabilities: capabilities for health, capabilities of future generations, animal capabilities and citizen capabilities.
First, in most of the cases discussed the capabilities for health of consumers are at stake. In the Energy Agreement case health is negatively affected through emissions, in the animal welfare case through the use of antibiotics, and in the binge-drinking case by drinking too much. In these cases, a broad consumer welfare approach could take health into account by looking at avoided healthcare costs (as explained above). However, valuing such a capability is very different from valuing the avoidance of (quantified) healthcare costs. Being of good health is a basic capability to function, which has to be taken into account when applying a capability approach. This capability is affected by lower emissions immediately leading to greater health. Furthermore, as indicated in Section 3, although the set of basic (minimum) capabilities necessary for individuals to flourish may be the same for everyone, the 'amount' of resources needed to reach a threshold level of the capability for health may differ between persons (interpersonal variability in converting resources into functionings). A reduction of emissions would immediately lead, for individuals with respiratory problems, to a higher level of functioning, even if for others the resulting improvements in health are negligible in this sense. The capability approach can take these differences into account.
Second, the capability approach may acknowledge capabilities of future generations. 62 In the Energy Agreement case, this would do justice to the fact that lowering emissions has important effects which go beyond any currently existing base of consumers. A reduction of emissions results in a (better chance of) flourishing of future generations, because of the expected long-term environmental effects. The ETS, of course, restricts the effect of lowering CO 2 emissions, which have a clearly long-term effect on the environment, unless the resulting allowances of CO 2 emissions are not traded back in the system: a capability approach would not change the fact that the case of the producers would be much stronger if this were the case. 63 Third, in the animal welfare cases, the capability approach might introduce the notion of animal capabilities. Then the capabilities to flourish of the animals involved will have to be balanced against the capabilities of humans, in their roles as producers, consumers and third parties affected through health impacts. There is great debate regarding the extent to which a capability approach should take animal capabilities into account. Nussbaum has argued that we should, 64 but others have criticized her. 65 Some have argued that we should look at ecosystem capabilities instead. 66 We cannot conclude this argument here, but argue that the framework of the capability approach is at least open to the possibillity for including animal capabilities.
Fourth, in the binge-drinking case there are some citizen capabilities at issue: capabilities of those living in the areas of consumption who are negatively affected by binge-drinking scenes. These might have to be specified as 'capabilities to live in a peaceful neighbourhood'. This is a rather individualist interpretation of the problem, which focuses on citizens being able to live their own lives without being disturbed by noise in the middle of the night, or by excessive pollution in the street. But the case may also be given a more collectivist twist by framing citizens' worries about binge drinking as a concern over social cohesion, or as paternalist concerns with respect to the value of the leisure activities of youngsters. Again, we will not take a stance on these issues here, but argue that the capability approach gives reason to coherently discuss these elements in a competition law case.

Comparison to an inclusive welfare approach
The discussion thus far has aimed at showing how a capability approach is superior to a (narrow or broad) consumer welfare approach in dealing with non-economic interests. This leaves us with another alternative to consumer welfare approaches, however: the inclusive welfare approach.
With respect to all of the types of cases discussed in this paper, an inclusive welfare approach could also take these non-economic interests into account: it generally uses a willingness-to-pay method. For example, current consumers may be asked how much they are willing to pay (for their electricity today) for a better environment for future generations. Thus, in the Energy Agreement case, consumers would be asked whether they would be willing to pay the higher costs of energy -amounting to 3-4 euros per year per household over 5 years -for a (switch to a) sustainable energy production (knowing that these costs are higher than the avoided costs of healthcare). 67 A willingness-to-pay survey could also ask consumers how much they are willing to pay for using less antibiotics on pig farms and greater animal welfare through providing more living space. It is even possible to use such a method for 'non-use' (for example keeping the rain forest healthy, knowing one will never travel there), in the case of protecting the black-tailed godwit. If such an inclusive welfare approach can take these interests into account in a satisfactory manner, then it might be judged a better alternative to consumer welfare approaches than the capability approach.
We do not want to overemphasize our disagreement with the inclusive welfare approach as it does also take non-economic interests into account. Nonetheless, there remain important differences with the capability approach to competition law. A general point is the one which has been important in developing the capability approach as an alternative to preference-based utilitarianism: that preferences may be distorted by problematic processes of adaptation and socialization. As a consequence, the answers in willingness-to-pay surveys have a problematic normative status. If in these surveys no one attaches value to future generations, animals or the health of parts of the population, then these interests do not count. Thus, the protection of these groups is dependent on the contingent value attached to them in these surveys. This does not do justice to the interests involved. These are two other important theoretical differences: the inclusive welfare approach will not take into account the (non-welfarist) value of market-constituting capabilities (liberties) and the threshold structure of consumptive capabilities, at least as long as individuals do not include them in their preferences.
To all of this an advocate of the inclusive welfare approach might retort that there is no other basis for making social choices. This is what an advocate of the capability approach would deny. 68 Social choices are legitimately made, in the end, in a democratic process. In that process, of course, citizen preferences play a crucial role. But the context of a democratic process is different from the context of a willingness-to-pay survey. The most important difference is that the democratic process is a means to form preferences on the basis of reasons, not to aggregate pre-existing preferences. All the substantive conclusions that we have argued for here, then, should be seen as reasons that may or may not convince citizens in a democratic process. Willingness-to-pay surveys do not allow such a process of reasoning to take place, and thus do not allow the subjects that are asked to report their preferences to change them on the basis of new information. 69 To the extent that an inclusive welfare approach is used by competition authorities to shortcut such a process, this is deeply problematic from a democratic point of view.

Conclusion
In this article we have argued that there are good reasons to include non-economic interests in a competition law assessment, and we have shown the shortcomings of the dominant consumer welfare approach in this respect. We have proposed the capability approach as an alternative and have explored what an 67 It might also make a difference to the consumers whether or not the allowance resulting from the CO 2 emission reduction would be removed from ETS completely. Even if this were not the case, the consumers might still be willing to pay the higher price, either as a political signal against the ETS and as a minor price to pay for uncertain, but possible, longer-term environmental benefits. Utrecht Law Review | Volume 12 | Issue 1 (January) 2016 application of this approach to competition cases might entail. In all of the cases discussed above, when applying a capability approach a balancing act is still necessary to come to a conclusion on whether or not to allow the agreement at issue under a competition law assessment. Our aim has been to show how the capability approach leads to a different result, by introducing three types of relevant capabilities. This shows that the capability approach (i) values market freedom, i.e. the capabilities to own property and to contract directly; (ii) recognizes consumers' capabilities but limits their value when consumer budgets are above specified thresholds; (iii) can incorporate a wide range of third-party capabilities. As a result, where consumer capabilities are already securely above a certain threshold, they may be outweighed by the value of the capabilities of future generations, the capability to health of current generations, or even by animal capabilities. Whether this is the case, is for producers to argue in the public sphere when they defend their agreements; for competition authorities to assess; and for parliaments to legislate upon when they are unsatisfied with the conclusions reached by competition authorities.