Equality of Opportunity in a European Social Market Economy

This article investigates what role the ideal of equality of opportunity should play in a European social market economy (ESME). After defining ‘social market economy’ and sketching different conceptions of equality of opportunity, it is argued that a social market economy must implement a substantive version of equality of opportunity. Subsequent sections assess how such a robust version needs adaptation in light of the EU’s special nature: first, it assesses the merits of a direct transnational application of interpersonal substantive equality. Second, it considers what the ideal requires in a ESME understood along internationalist lines: even on this account, labour mobility creates tensions between EU citizens’ claims to equal prospective chances in a fair cross‐border competition against each state’s prerogative of providing the highest level of education to its residents. The concluding section offers some suggestions how we might alleviate this tension between domestic equality of opportunity and national autonomy.


Introduction
In a recent contribution to the Journal of Common Market Studies Jürgen Habermas investigates whether it is permissible to diverge from the ideal of equal representation of each citizen in democratic procedures when adjusting such procedures to the special case of a supranational political order. Though he considers the ideal of equal representation of citizens essential for the legitimacy of any nation-state, Habermas offers a qualified defence for derogating from it in the European Union (2017, p. 179). In making his case, Habermas appeals to the special value of political equality amongst member states, which exists alongside the ideal of interpersonal political equality amongst European citizens. My argumentative strategy in this contribution runs parallel to Habermas' reflection, except that the ideal of equality on which I will focus reflects our considered judgment of the justifiability of a domestic economic order. For those subscribing domestically to an ideal of a social market economy (SME), the justifiability of such an order hinges on the realization of robust equality of opportunity: since some, but not others, end up occupying positions of authority, power, income and wealth, our societies must offer equally talented and motivated persons the same prospects of reaching such advantageous positions, irrespective of their social background.
If equality of opportunity is a sine qua non for a just market economy domestically, then how must we conceive of this requirement once we seek to realize SME at the European level? Probing the role of equality of opportunity in a plausible conception of a European Social Market Economy (ESME) is the task of this contribution. I argue that although the basic justificatory requirement remains intactjust as equal weight in representation remains part of the supranational democratic idealit must in certain respects be weighed against other values that compete with it at the supranational level.
The argument proceeds as follows: In the next section, I briefly offer my understanding of the notion of social market economy. Section II distinguishes different conceptions of equality of opportunity and discusses in detail the one that aligns most plausibly with SME. Specifically, I make the case that a substantive opportunity condition is one part of three-fold justificatory requirement of inequality in positions within such an economy. So if the EU aspires to become an ESME (as it claims in Article 3.3 TEU), then there is a prima facie case that it needs to satisfy robust equality of opportunity. Section III turns to the EU and describes how some laws and directives concerning equal treatment and nondiscrimination already point in the direction of the ideal. This might prompt somecall them transnationaliststo conclude that we can apply the demands of robust equality of opportunity directly to the EU as a whole and require its institutional rules to advance this ideal. Despite noting some notable features that point in this direction, I will resist this straightforward inference. Instead, I consider the issue in detail from an internationalist perspective and suggest that even on this perspective, there is a prima facie case for accommodating some robust opportunity measures at the supranational level (section IV). This may require drastic interference in each state's autonomous choice regarding the provision of primary and secondary education. Since internationalists presumably want to uphold each state's special prerogative to provide educational goods to citizens, this seems to create a dilemma (section V). I suggest that resolving it in favour of the state's special prerogative is possible, but generates a new and interesting argument for member states convergence. The final section concludes.

I. Social Market Economy and its Alternatives
The notion of social market economy (SME) is ambivalent. Whilst the German 'soziale Marktwirtschaft' from which it historically derives has (ordo-)liberal and anti-socialist connotations, 1 the English term, at least in its current usage, denotes a socio-economic order that places certain constraints on the working of the market and seeks to correct some market-generated inequalities through public policies. This distinguishes it from a free market economy in which no, or at any rate much fewer, constraints and correction mechanisms are in place. For example, Scharpf, in his seminal studies of why the EU is not and cannot become a social market economy, uses the term as a high-level classification to denote features that different varieties of (Continental-)European capitalism have in common. States to which the SME-label applies, according to Scharpf, fit either into the 'Christian-Democratic' or 'Social-Democratic' variety of capitalism and they are distinguished from 'Liberal Anglo-Saxon' capitalism by the fact that they put significantly more emphasis on 'ensur[ing] social equality, social security and the decommodification of labour ' (2010, p. 233). 2 Whilst Scharpf and other social scientists use the SME label to classify real-existing capitalist economies into different categories, the use to which I will put it here is gradual and normative: I will assume that SME is the ideal-type socio-economic organization that an institutionally-embedded market economy should have in terms of striking a balance between the values that free markets are ordinarily thought to enshrine, notably efficiency, freedom and personal responsibility ('free market values') and non-market values such as social equality, fairness and solidarity ('social values'). Since a complete account of this ideal-type SME would amount to a fully spelled out account of economic justicea task to which one surely cannot do justice in the space of a short contributionthe description here is limited to noting three minimal market-constraining conditions that SMEs have to make adequate room for the realization of the social values mentioned above.
First, an SME is a socio-economic order in which some domains of life are not subjected to the commodifying tendencies of the market: to use a phrase by Debra Satz (2010), in an SME 'some things should not be for sale'. One animating idea behind SME is that of society conceived as a 'community of equals'. Where markets threaten it through the commodification of important human needs, market-regulating institutions will impose limits on them (cf. Anderson, 1993). This corresponds to Scharpf's first and third characteristic mentioned above. Since the threat of commodification is not my primary concern here, I will not discuss this requirement in more detail.
Second, an SME will correct the outcomes of markets via redistributive taxation to contain disparities of income and wealth. There are several reasons for this, but again I must do with an impressionistic sketch: a first reason has to do with the fairness of distributive outcomes when a collective agent owes all under its authority equal treatment (Scanlon, 2018, p. 12). A second reason is to protect weaker members of society from poverty, domination and exploitation. Yet another is to ensure that inequalities between the majority and the richest segments of society do not become too extreme, for this may again threaten the idea of social equality and endanger the proper function of democratic institutions (Miller, 1995, p. 37). Third, and most important for our purposes, an SME will structure social institutions and regulate markets in such a way that the process by which markets create outcomes treats each citizen fairly. 3 This is the domain in which claims of fair equality of opportunity come into play. 4

II. Equality of Opportunity: Concept and Conceptions
In the most abstract terms, equality of opportunity is the ideal that characterizes an open socio-economic order that stands in contrast to one that is characterized by a fixed hierarchy of positions (Arneson, 2015, p. 1). Assuming that there will be different and differentially rewarding social positions, equality of opportunity mandates that the allocation of 2 The varieties of capitalism on which Scharpf draws are from Esping-Andersen, 1990. 3 I do not mean to suggest here that equality of opportunity is solely about proceduresmy suggestion is only that it is centrally about fair and equal chances rather than outcomes, given that it already assumes that differentially desirable positions of wealth and authority are in fact justifiable. I thank an anonymous referee for prompting me to clarify this. 4 In this contribution, I focus on the ideal of equality of opportunity. I do not thereby mean to imply that investigating the EU's implication on the other elements of SME, namely material inequality and commodification do not constitute equally valuable objects of assessment. individuals to social positions is 'determined by some form of competitive process, and all members (…) are eligible to compete on equal terms' (Arneson, 2015, p. 1). In such abstract terms, few will be opposed to this ideal. But to get to a more concrete definition of what equality of opportunity means, it is useful to distinguish between the concept of equality of opportunity and various conceptions of it. The distinction is famously drawn by Rawls in regard of different conceptions of the concept of justice (Rawls, 1999, p. 4). The concept of equality of opportunity can be differentiated along four dimensions (Lazenby, 2016, p. 67;Westen, 1985): 1. an account of the subjects of the principle, such as a distinctive group 2. an account of the agent(s) responsible for its realization, such as the state 3. an account of the object of application, for example, education, jobs, positions of authority, advantage etc. 4. a description of obstacles to the realization of equal opportunities, for example, social class, race, sex, all non-chosen features etc. (Lazenby, 2016, pp. 67-68).Within these dimensions that any account must have, we point out important differences between conceptions to get a clearer focus on the particular conception of equality of opportunity that I think SME must support.

Subjects
The kind of equality of opportunity that is entailed by SME considers all those subject to the relevant socio-economic order to be its subjects. So the relevant criterion is institutional: whether or not equality of opportunity has purchase amongst a set of persons for SME purposes depends on whether or not these persons are subject to the common market institutions of a particular socio-economic order. Now what exactly the necessary and sufficient features of such a socio-economic order are is a difficult question, but I will assume that it requires at least a common institutional framework that settles and enforces (either directly or indirectly) market-relevant rules such as those regarding: • Who may participate: the institution restricts or permits market access to external parties via barriers and tariffs; it designs minimal participant requirements internally ('qualified investor'); it creates categories of agents (such as limited liability companies) to facilitate trade; • What counts as ownership: the institution determines the nature and extent of property rights, including intangibles such as trademarks and patents; • What may be traded: it sets minimum standards (of health, of environmental hazard etc.) for the kinds of products that may be traded; it creates quality standards and other norms to improve information markets; • How trade may proceed: it determines what counts as permissible competition, it prohibits and sanctions market-undermining practices such as insider trading, market cornering, discrimination, collusion, cartel formation, etc. Such an institutionallygrounded idea of who counts as subject stands in distinction to prominent cosmopolitan theorists (Caney, 2001;Kollar, 2017;Moellendorf, 2009) that defend a global (non-relational) principle of equality of opportunity according to which this moral ideal has purchase whenever or wherever any kind of difference in life prospect for persons arises. So, to take an example offered by Derek Parfit, if the life prospects of native Americans and Europeans prior to the crossing of the Atlantic were different, then these differences might give rise to complaints of inequality of opportunity (Parfit, 1997, p. 206). I do not take a stance here on whether or not such a reading of the subject of equality of opportunity is also a plausible one, but simply register that this is not the kind of conception with which I am concerned in this contribution because I aim to investigate an institutionally-grounded conception that defenders of an SME should find convincing.

Responsible
Agents: Though various possible ideas of who has responsibility for guaranteeing equality of opportunity are possible, the idea of agents as far as SME is concerned follows the institutional path just described, that is ensuring equality of opportunity is realized lies ultimately with the market-regulating public authority that enforces the legal framework. Object: SME is primarily a claim about how a socio-economic order strikes a balance between market values and other social values. Thus, the positions of authority and advantage on which equality of opportunity as part of an account of a SME must take a stance might appear to be those that are generated through the market. But this would be too narrow. There are important positions of authority and advantage, for example in those public institutions indispensable to guarantee the orderly functioning of the market, that will also need to be considered. So although advantageous positions in the economic sphere make up the most important share of the positions that an account of equality of opportunity in an SME will cover, the argument is not restricted to these positions.

Obstacles:
So far I have distinguished between different conceptions of the ideal of equality of opportunity and suggested that SME would (a) take those subject to the common market to be its subject, (b) consider positions of authority and advantage generated through the existence of such a market to be the main objects of equality of opportunity. But surely, this does not yet offer a complete account. What is missing is a definition of what should count as obstacles to the realization of equal opportunity. This allows us to distinguish the account that will be defended from both narrower and broader ones. As we just saw, some accounts consider any factor over which the agent had 'no control' to be an obstacle to realizing the relevant object. Other accounts are significantly more permissive in terms of what can generate acceptable inequalities. If we take advantageous employment positions to be the quintessential object of inquiry, then 'merely formal' accounts of equality of opportunity see the ideal as already satisfied when nobody is prevented (by law, or through intimidation or coercion) from applying for these positions.
Merely formal accounts of the obstacles to equality of opportunity are accepted by a broad range of philosophical views, including most who favour a free market economy over SME, if only because the meritocratic idea underpinning formal equality of opportunity will likely lead to a more efficient allocation of 'human capital'. 5 However, I will argue below that the relevant accountto be one for a social market economymust go beyond merely formal considerations concerning the obstacles to advantageous positions. Let me now describe this account in more detail and explain why merely formal equality of opportunity is insufficient.
In a market economy there will almost by definition be some scarce positions of advantage: there will be managers, heads of regulatory bodies, highly-paid corporate lawyers, chief designers and engineers, business consultants and many more desirable offices. For these positions, there will be competition: some will end up in them, others will not. The simple thought, to be developed below, is that showing that substantive and not merely formal equality of opportunity obtained is part of the justification that society must be able to offer to those who lose out in competition. According to T.M. Scanlon, the justificatory burden that an institution that generates inequalities in positions generates must meet is threefold (2018, p. 70). I adopt these for the discussion of a market economy with such positions: 1. Institutional Justification. It must be shown that having a market economy that generates such inequalities in power and authority is justified, for example, because it is beneficial to everybody; because it maximizes liberty; or because (in the case of a SME) it is likely to improve the position of the worst of members of society. 2. Procedural fairness: the allocation of the scarce positions must follow a process that is at least 'sufficiently sensitive' to the justifying rationale for unequal positions. 6 Competition for these positions must be open to every person qualified for them; application procedures be transparent; discrimination avoided, etc. 3. Substantive Opportunity. It must be shown that the process of skills-acquisition as a result of which some win the procedurally fair competition for positions of advantage is also justifiable to each (2018, p. 70).
The third justificatory requirement is the one that distinguishes a 'substantive' account of equality of opportunity from merely 'formal' accounts, namely that it must be shown that the conditions through which the differences in suitability for an advantageous position arose are also justified. Why should we take this to be a requirement, and what exactly does it entail? Two arguments strike me as most powerful. The first one points to the complaint of unfairness that losers in competition under merely formal equality of opportunity would have.
The following case should be helpful to exemplify the complaint of unfairness: Imagine Smith is chosen over Jones as chief designer in a fair application procedure because Smith has special skills in trademark design that Jones lacks. But now suppose that we learn that Jones, who comes from a modest social background, never had the opportunity to receive the special training that Smith received because trademark design is only taught at an expensive private institute that Smith (who comes from an affluent background), but not Jones, could afford. And suppose further that we know, counterfactually, that if Jones had also received the special training, then he would now be definitely as good a candidate for the position. To most people it seems clear that although Smith 'won' the competition in a manner that impeccably observed procedural fairness, and although placing Smith in the position would satisfy institutional justification, this is not all that matters. Rather, social institutions should ensure that equally capable young people with unequal backgrounds and financial endowments should still have equal prospects of actually reaching positions of advantage. This is in fact more or less the formulation that Rawls chooses when he explains his version of fair equality of opportunity (1999, p. 63): 'Assuming there is a distribution of natural assets, those who are at the same level of talent and ability, and have the same willingness to use them, should have the same prospects of success regardless of their initial place in the social system'.
A second, 'democratic' argument for substantive opportunity looks beyond the fairness towards the individual application and considers the issue from the broader issue of what kinds of elites (groups of people filling positions of advantage) a socio-economic order should aspire to have. According to Elizabeth Anderson, 'a democratically qualified elite must be an elite that is integrated across all the major lines of social inequality and division that characterize it ' (2007, p. 597). This is so because effectively serving the democratic interest requires such things as knowledge of the interests and concerns of people from all social backgrounds, 'a disposition to serve those interests', and both technical and social competence of serving these interests, which includes respectful interaction with all sectors of society (2007, p. 596). This kind of consideration speaks in favour of going beyond formal equality of opportunity if doing so would raise the prospects of democratizing elites, that is making them more broadly representative of all sectors of society. And there is every reason to believe that this would be the case: decoupling skill level from socio-economic background would both enhance the diversity of class amongst elites, and, given the de facto correlation of wealth with other dimensions of diversity (such as minority status) increase overall diversity.
What would the realization of this ideal of equality of opportunity require? This is a difficult question, which even Rawls avoids discussing in full detail. But we can say at least something about some policies and practices that it would prohibit, and some of the institutional features that an ideal-type SME would have in place to get closer to its realization. First such a socio-economic order would seek to balance, insofar as this would be possible (and also insofar as this is tolerable in light of other values), some unfair advantages in skill-acquisition that those initially occupying 'more favourable positions in the social system' receive. Just to give one clear example, this requires free pre-university educational opportunities for each person that are of at least roughly equal quality. In other cases, it may demand providing those from disadvantaged backgrounds with subsidies to make up for advantages that others receive. And in some cases, it might require prohibiting ways of conferring advantage on the better off, for example superior educational opportunities that only the very rich can afford.

EU-wide substantive opportunity and transnationalism
The discussion thus far has concluded that an SME will contain a robust principle of equality of opportunity, inequalities in position of advantage and authority that arise in a market economy need to meet the threefold test described, including substantive opportunity. We are now in a position to ask where this leaves us in terms of adopting this ideal to a transnational European social market economy. There are, I believe, two different, and interestingly distinct ways of pursuing this question. The contribution first looks at one of these, which I label a transnational approach in this section. This account insists that the existence of the EU's supranational market-embedding institutions (that determine to a significant extent who can trade what under which conditions with whom) make it the case that, first, the subjects of the principle in an ESME are all EU citizens, and second, that the responsible agent, that is the appropriate addressee of some complaints about inequality that is generated as a result of the common market's existence, today is the EU itself. In other words, the inequalities in position generated anywhere in the union as a result of the common market need to satisfy a principle of substantive equality of opportunity, namely that all economic positions of authority and advantage in the European market economy must: (i) be open to all EU citizens; and (ii) EU citizens at the same level of talent and motivation should have equal prospects of ending up in these positions.As far as (i) is concerned, we can note, as an interpretative claim, that the EU has at least to some extent taken on the role of a public authority of ensuring formal equality of opportunity amongst EU citizens: Unlike any other supranational institution, it has created, through treaty provisions, directives and ECJ case law, an intricate system of rules that stipulates some dimensions along which each EU citizen can, as EU citizens, demand equal treatment from public authorities and private actors in certain domains across the union. These constantly 'widening and deepening' (Bell, 2011, p. 612;Moore, 1999, p. 102) entitlements to equal treatment are enforceable not only against each member state, but directly at the supranational level. Moreover, the union guarantees to each EU citizen a set of rights against discrimination in respect of some protected characteristics (Article 2 TEU; Council Directive 2000/43/EC; Council Directive 2000/78/EC; Ellis and Watson, 2012;Tryfonidou, 2014). Does this make a sufficiently strong case that the EU is as a matter of fact moving in the direction of realizing a transnational version of substantive equality of opportunity, both concerning agency and subjects? I think this would be a too hasty conclusion. After all, the point that the EU, partly as a matter of market access, partly as a matter of recognition of rights against non-discrimination, requires certain forms of equal treatment establishes, if at all, that there is a concern with formal equality of opportunity, such as fair procedures.
At any rate, the fundamental question is the normative one, namely how plausible the case for the opportunity principle -(ii) aboveturns out to be. It has the strongest intuitive appeal, I would argue, for those positions that would be absent if the EU did not exist, notably those positions of advantage that are necessary to guarantee the proper functioning of the common marketin other words its main agencies like the Commission, the ECB (for the eurozone), and so on. Surely equal access must be guaranteed to all talented EU citizens, and moreover, complaints about inequalities in access to these positions would appropriately be addressed to the EU. So both the subjects-and the agency-dimension are supranational. Moreover, it is plausible to believe that substantive opportunity might be a requirement for the pool of applicants applying to these positions of authority. Notice that for this special case of positions in the supranational bureaucracy, both the arguments from fair competition and the argument from democratic elite representation for substantive opportunity seem to have some force: if no student from Romania ever won a prestigious internship at the Commission or if no Greek economist were ever hired by the ECB, we would be concerned both because it would strike us as unfair that domestic educational systems have such an influence on people with equal talents and motivation under transnational competition, and because we would be worried about the structure of Eurocracy in terms of national representativeness. But these positions are few and rarethe Commission employs 32,000 peopleand so perhaps their significance should not be overstated. Moreover, as far as individual fairness is concerned, the fact that each would-be Eurocrat still has a national, fully insulated 'bureaucracy labour market' to fall back on (has thus still access to a rewarding employment in a similar occupation under conditions of robust equality of opportunity vis-à-vis all competitors in this market), arguably mitigates this kind of case.
Much more important is the question of whether the EU as an institution owes EU citizens a justification for unequal access to other positions, notably high-profile jobs anywhere in the common market. Clearly the best reason to think this stems from the fact that, unlike any other market-establishing institution beyond the state, the common market governed by the EU includes more or less complete mobility of labour, hence reshapes domestic opportunities and positions by opening up transnational competition for advantageous jobs. Moreover, and contrary to 'protected employment categories' like civil servants, there is for most positions of advantage in the market, continuous competition such that a top architect who competes in one country can compete in every member state for employment.
Of course, there is a question about how competitive the transnational EU labour market really is given certain natural entry barriers for foreigners, most notably language proficiency skills. Against this though, it can hardly be denied that in some domains, competition is more transnational the higher rewards attach to the position. 7 At least for some positions, competition is truly transnational. Surely, it would be odd if equality of opportunity somehow need not apply for the most desirable positions.
There is an important objection to this line of reasoning that immediately comes to mind: those positions that are genuinely competitive on an EU-wide level (think: CEO, top-engineer, investment banker, chief surgeon etc.) are normally positions that require a very high level of education, typically including university studies and perhaps even a doctorate. But, and here comes the objection, for the level of study and education that these positions require, there is already in place de facto some form of equal opportunity for the simple reason that the ECJ hasat least since the 1985 Gravier ruling (Case C-293/83 Gravier [1985]; Beerkens, 2008, p. 413;Barnard, 2012)enforced a strict regime of EU-wide equal access to institutions of higher learning.
Since all those competing for these high-flying transnational positions had equal opportunity to go and hone their skills at the same institutions (if they were admitted), there is no further requirement of substantive opportunity that needs to be satisfied. But the problem with this argument is that even if universities, including top (private) ones, are equally formally accessible for all EU citizens, this does not mean that the applicants may not have competed for these educational positions on very unequal terms because of their nationally provided pre-university education. 8 But avoiding these talent-and motivation-unrelated forms of (dis)advantage was exactly the requirement of substantive opportunity in the domestic SME context.
Though I have shown some sympathy for this effort of establishing transnational substantive opportunity, the argument strikes me as ultimately unpersuasive for the following reason: it simply assumes that because there is a common market, the relevant agent to whom complaints against differential access to positions of authority and advantage should be addressed is the supranational one. But this requires an argument. The best argument here would be something like a 'cantilever argument', that is, an argument that shows to the supporter of SME that his commitments to the state being in charge of substantive equality of opportunity in a domestic SME forces him to ipso facto accept that the EU must be in charge of European substantive equality of opportunity. But it is hard to come by such an argument because, although we noted that the EU does give rise to some transnational competition, there is still some purely internal competition within member states that gives rise to claims of equality of opportunity for which, presumably, each member state must retain ultimate responsibility. But this very fact establishes a crucial difference between the EU and each nation state. As a result of this, we cannot simply assume that the relevant claims by those who lose out in transnational competition must be addressed to the supranational institution.

IV. EU-wide Substantive Opportunity and Internationalism
A quite different way to think about equality of opportunity in the European context, one that is perhaps anyway more in tune with how the majority of EU citizens understand the EU's point and purpose, is the following internationalist stance: looking at its historical development, the EU's basic premise has been, and continues to be, something like 'Keynes at home, Smith abroad' (Gilpin, 1987, p. 355): the transnational market guarantees and best instantiates the free market values, and the social values that any SME must further are protected by each member state. In relation to our question of unequal access to jobs, this implies that the complaint of somebody who loses out as a result of transnational competition for jobsperhaps setting aside positions in the union's bureaucracyis still properly addressed to the member state of which he or she is a citizen (agency). Moreover, the subjects are still for each state those who compete for employment within its jurisdiction.
In what follows I want to adopt this internationalist picture and ask how the relevant member state could respond to the complaint of a group of persons that lose out within the member state in which he or she grew up and was trained against better-qualified job seekers from other EU member states, a phenomenon that will inevitably occur sometimes in an open transnational labour market. My suggestion is that even on this more constrained internationalist understanding of what equality of opportunity in an ESME amounts to, the ideal pushes us in the direction of adopting supranational measures to ensure substantive opportunity at the domestic level.
Let us look at how the state should respond to institutional justification, procedural fairness and substantive opportunity in such a case (recall section II): as far as institutional justification is concerned, the state would argue that permitting the particular practice that results in unequal positions to be filled by the most qualified person from across the EU is justified because it permits to making the worst off people domestically better off by sustaining higher levels of prosperity. In terms of satisfying the condition of procedural fairness against the complaint of having lost out, the state will point (rightly) to the fact that the practice of choosing the most qualified candidate from a transnational pool of candidates is the best method of realizing the institution's goal. Moreover, it does not impermissibly discriminate in so doing. Now the crux of the matter will again be whether the state can succeed in offering an adequate justification to the domestic citizen's complaint that the group of EU third country applicants who now predominantly are given positions of advantage in a fair process (let us assume) had better chances of developing their talents to become suitable candidates for the relevant positions, that is, whether the state satisfies substantive opportunity.
The following example will clarify the case I have in mind and point out a difficulty: suppose that initially, there are two fully separate markets in labour and manufactured goods, say, Sweden and Italy. Imagine two groups of students, one Swedish and one Italian, with equal talent and motivation each aspire to become high-flying executives at a car manufacturer. In the initial situation, there is an Italian labour market in which the Italian group of students has a certain chance of reaching their goal at a local car manufacturer, say Fiat, and there is a Swedish market in which the Swedish students have a chance of gaining such a position at Volvo. Let us assume that each state separately fulfils the requirements of robust equality of opportunity. In such a situation, as it arguably obtains across most nation states globally (setting aside the fact that no state actually satisfies the ideal of SME fully), it does not matter whether either of these two groups of students will receive certain advantages relative to the other. So long as each potential competitor in the domestic labour market receives equal opportunities to turn talent into skill, equality of opportunity is realized because a better education for the Swedish students relative to the Italians does not threaten equality of opportunity between the two.
But now suppose that we apply more realistic conditions of the common market, namely that there is a unified labour market at least for highly skilled engineers. Changing the assumptions in this way will have the effect that the Swedish and the Italian students as a group will begin to compete not merely for equivalent local jobs, but the various jobs for which they apply are now identical such that potential employers will have to evaluate each against each other. Since these positions of advantage are by definition scarce, how each of them fares impacts the chances of the others. Now suppose that Swedish students received a better primary, secondary and tertiary education and as a result of this fact are (at the same level of talent and motivation) the more desirable employees to have. In such a case all relevant positions (in both Italy and Sweden) would be filled by Swedes, including some in Italy that would have gone to Italian students had they had an equally fine education.
The question is thus whether the Italian automotive-executive-to-be has a complaint against her state. In one respect, the state seems to have done everything that the ideal of robust equality of opportunity requires: all the Italians at the same level of talent and motivation still have precisely the same prospects of gaining access to the advantageous positions. Against this though, the complainant would insist that in letting 'new' participants who have received better training (at same levels of talent and motivation) compete in the domestic labour market, their state as the agent responsible for ensuring equality of opportunity fails to ensure the conditions amongst all persons de facto competing for these scarce positions. Why should the fact that the group of people one loses out against received their unfair advantage abroad be of any import? Imagine all rich parents sent their children to a country with very expensive boarding school at a young age only to reinsert them into the competition for position once their education is done. If the level of education abroad was much better than the education that poorer children receive at home, then surely this would violate substantive opportunity. But what is the difference between this case and the case where parents are also abroad?
Perhaps the state might respond by pointing to two seemingly pertinent differences between these cases: whereas the only consequence for children with poorer background in the boarding school example is that it worsens their opportunities at equal levels of talent and motivation, the practice of permitting a transnational labour market in fact also expands the possible option set of those now losing out to better skilled-competition from abroad. Second, though it may be (under certain circumstances, if this were a widespread practice etc.) permissible to prevent rich parents from advantage-laundering through foreign education, it would be impermissible to withdraw from the common market because of the substantive costs that this would involve, includinglet us assumecosts to the worst off members of society.
But the problems with these responses by the state are easy to make out: as far as the first point is concerned, the example was designed exactly in such a way that all the scarce positions will go to those with the best training at the same level of talent and motivation, including the comparatively advantageous jobs in Sweden. Perhaps there is a 'nominal' or 'formal' increase in opportunities for Italians with a worse skillset to apply for scarce positions because there are more of them. But if there is now a large enough pool of better-educated applicants, then this 'nominal' increase is neither here nor there as far as substantive opportunity is concerned. As for the second point, the complainant would surely respond that aside from either keeping the transnational labour market or returning to a less efficient closed-border economy that worsens the position of the worst off, there is of course a third option: keeping the transnational market and equalizing educational and skill-acquisition opportunities across the union for all those with equal talent and motivation. Moreover, if a concern with efficiency and aggregate prosperity (for redistribution to the worse off) was part of the justification for keeping the common market, then keeping it whilst equalizing educational opportunity would likely even better ensure that goal.

V. An Internationalist Dilemma?
If the argument of the last section is correct, then it seems that internationalism regarding equality of opportunity in a ESME faces a trilemma of the following sort: of the three desirable outcomes of (a) a free transnational labour market, (b) individual state autonomy over the provision of educational goods, and (c) substantive domestic equality of opportunity, only two can be realized simultaneously. 9 Whilst transnationalist would simply drop (b) as they consider national autonomy less essential than the common market or substantive equality of opportunity, the internationalist attaches significant value to national educational autonomy.
It is important to point to a somewhat unusual feature of this particular autonomy-market-equality trilemma: what renders country A incapable of guaranteeing fair access to positions of advantage within its economy under conditions of cross-national labour mobility is not some other state B's action that we might find problematic anyway on some independent groundas we would in similar dilemmata when, for example, state B competitively lowers taxes to win the competition for portfolio income or when it keeps down health and safety standards to attract production sites. In this case by contrast, it is state B's superior provision of an independently valuable human good, namely education, that makes it harder for A to ensure equality of opportunity within its jurisdiction. This should count as a weighty reason against prohibiting states from providing this good as well as they can, meaning that either equality of opportunity or labour mobility would need to go. So should the Internationalist perhaps just give up an unrestricted endorsement of transnational labour mobility in the EU? I believe that she should not, or at any rate, she should not do so if the reason is to guarantee that its already advantaged members should be protected from (admittedly substantively unfair) competition from abroad at the expense of those worst off in the countryrecall that this was part of the justification for a common market in labour in ESME. 10 The final possibility, which I want to explore here briefly, is whether we can somehow moderate the apparent tension in the domain of providing substantive equality of opportunity. To do so, I want to go back to an analogy on which I relied earlier, namely the one between states providing superior educational goods to citizens in a transnationally open labour market and parents providing additional or superior educational goods to their children in society, thereby improving their children's position relative to those of children with less affluent backgrounds (Brighouse and Swift, 2009). We can make advancements in the former case, I hope, by (i) working out under what conditions a state may lack reasons to intervene by subsidizing the educationally disadvantaged in the latter, and (ii) noting a difference between the responsibility of the state and that of parents.
Regarding (i) consider the case of two equally affluent families, where the first set of parents decide to give some competitively beneficial advantage to their child, such as expensive educational training that makes the child more employable for positions of advantage in the economy, whilst the second set of parents decide to invest an equal amount in their child's spiritual development, such as in religious education, which, though intrinsically valuable, has no competitive benefit in terms of employability. It seems that in this case, the reason to subsidize the latter child is weaker than in cases where parents had no endowment and hence no choice for conferring advantage at all. Of course, the reasons to support the religiously educated child are just as strong as far as the competitive equal access to positions of advantage in the economy are at stake. But from the state's perspective, there is also a reason against this because in compensating the religious child's lack of competitiveness relative to the former (over and above the skills that is guaranteed to all), it would in fact start to subsidize the latter parents' choice for spiritual development.
But crucially, this anti-subsidy argument is only available where original endowments of parents are equal.
As far as (ii) is concerned, I think there is a fundamental difference between the two scenarios that concerns the relevant responsibility of the higher-level public agent. Whilst the state in the family case retains a default responsibility to ensure that children receive an equally good education, perhaps understanding the relevant higher-level role of the EU in the interstate case should be understood as one that needs to balance a concern with interpersonal equality in the domain of access to positions and a kind of interstate equality to realize different kinds of bundles in terms of pursing societal goals which includes, but is clearly not limited to, the provision of the good of education. The task for the EU, on this reading that seeks to respect state equality, is to make it the case that the choices by states are like those of the two sets of parents with equal endowments that reflectively choose the pursuit of different goals, rather than the case where some parents can make such choices whilst others cannot. On this understanding the EU, rather than trying to implement a supranational principle of substantive equality of opportunity, owes it to states to seek to make the choices that determine the level of provision of the good of primary and secondary education equally 'costly' in terms of the trade-offs that investing in better education requires. Thus understood, the EU should put its power behind increasing socio-economic convergence between states.

Conclusion
The ideal of equality of opportunity, like other political ideals, must come to grips with the complex nature of the European Union, namely its pouvoir constituent mixed in which both individuals as EU citizens and states (the legitimate representatives of citizens of collectives) interact. My core contribution to theorizing this ideal in the EU was threefold. First, I argued that a straightforward transnational application of interpersonal substantive equality of opportunity does not do justice to the special nature of the European institutional architecture: member states are still today the core agents responsible for the fulfilment of interpersonal equality. Second, investigating the role that equality of opportunity should play in a more internationalist ESME, it was pointed out that on such an account, under labour mobility there will be a tension between individual EU citizens' claims to have equal prospective chances of ending up in some desirable position of advantage in a fair competition and each state's prerogative of providing the highest feasible level of education.
Third, I made some preliminary remarks about how we might resolve this tension between domestic equality of opportunity and national autonomy over education: Since both of these are valuable goals, a European social market economy must recognize each as important. My proposal was to soften this tension via convergence amongst member states: if states had equal resource endowments to spend on education, we could justify the inequality that each state permits for its own citizens in terms of an autonomous trade-off between the supply of educational goods and the pursuit of other valuable goals. The interesting feature of this argument is that rather than deriving a justification for cross-nationally redistributive policies at the EU-level from premises about EU-wide interpersonal demands of economic justice that few internationalists would share, this argument is grounded in two core internationalist commitments.