The Economic Case for Markets as Mere Means

There is a potentially endless variety in the details of the design of public markets.Footnote ³ To keep the discussion tractable, I will focus on Ronald Daniels and Michael Trebilcock’s defense of voucher markets as an exemplary case of introducing markets in public services (throughout I will refer to ‘public markets’ and voucher markets synonymously).Footnote ⁴ Here I follow them in considering public services in general, but the reader should have in mind educational and health care services as primary examples.Footnote ⁵ Daniels and Trebilcock define vouchers as:

Daniels and Trebilcock argue for voucher markets on the basis of the theory of incomplete contracts.Footnote ⁷ Since public services are complex it is impossible to specify all details of the service in advance. Two principal–agent problems arise. First, the government (agent) may have insufficient incentives to execute the wishes of citizens/users (principal) and the latter may have insufficient information to prevent the government from abusing its position. Second, when contracting with specific providers – whether public or private – the same problems of incentives and information may arise between government (now principal) and providers (agents). Daniels and Trebilcock use this analytical perspective to compare three modes of social service delivery.Footnote ⁸ In each case, they try to diagnose the potential to overcome these principal–agent problems.

First, pure public provision scores worst. In the relation between citizens and government, control is diffused amongst many citizens and, therefore, ineffective. While they do not elaborate on this, Daniels and Trebilcock are deeply pessimistic about the functioning of the electoral market. In the relation between government and public sector workers, there are also insufficient incentives for efficient production, because these workers ‘have no claim on the residuals arising from efficient “production”’Footnote ⁹ and because they are not subject to any competitive pressures. To the contrary, they may have an incentive for inefficient production so as to maximize future budgets for their organizations.

Second, supply-side subsidies refer to the introduction of markets in the form of ‘contracting out’. Governments contract with private suppliers for a service over a specified amount of time. This promises better results in the government–provider relationship, since it introduces competitive pressures and the opportunity for providers to keep residuals from the service. Both potentially enhance efficiency. However, the citizen–government relationship remains problematic. While governments have an incentive to reduce costs (since they can claim the residuals), consumers usually have an interest in increasing quality, and government will fail to represent that interest absent effective control by citizens.Footnote ¹⁰

Third, demand-side subsidies (vouchers) remedy this problem.Footnote ¹¹ The key to their superiority is that government falls out of the equation: consumers bargain directly with providers. The latter have incentives to satisfy consumer demands (given competition and a claim to residuals) and the former have an incentive to press for cost-effective delivery (given that they are responsible for costs exceeding the voucher). Voucher markets, then, basically reduce the double principal–agent problem to a single one, which can be effectively managed, at least if favorable market conditions obtain.Footnote ¹² For voucher markets face two series of challenges: issues about the design of vouchers and about market failures.

Daniels and Trebilcock distinguish five design issues. First, entitlements to vouchers can be granted either on a universal or on a means-tested basis. Second, the value of the voucher needs to be determined. The voucher can give a right to unlimited services (under a certain description of the service) or be fixed or capped. Moreover, co-payments from consumers may be made obligatory or not. Third, extra (‘top-up’) payments above the value of the voucher can be allowed or prohibited. Such payments may create a much larger market but obviously lead to inequalities in access for consumers. Fourth, on the supply side, providers can be allowed or prohibited to refuse unattractive customers (‘cream-skimming’). The freedom to do business here may lead to the exclusion of some citizens (such as patients with prior conditions). Finally, governments must decide on what basis providers can qualify to act in the market. For example, governments may choose to include or exclude religious schools, but also require a certain size, management structure, etc.Footnote ¹³

A second series of challenges is about market failures. Daniels and Trebilcock discuss two often-recurring market failures. First, given the benefits they expect from competition, it is incredibly important to have effective competition and prevent monopolies from arising, so that consumers have a real choice between several providers. Second, mechanisms need to be established to remedy information failures. Consumers need to have good information about the quality of services; governments may need to step in with mandatory disclosure requirements or other measures.Footnote ¹⁴ This list of markets needs to be extended.Footnote ¹⁵ In many public markets consumers are vulnerable or weak agents (think of patients, children, addicts, etc.). In these cases they may not be able to compare offers made by providers and make effective consumer choices. Finally, there may be negative externalities from consumer choices. For example, one critic of public markets argues that many working-class parents judge it more important that children are happy at school (compared to the achievement focus of middle-class parents). This would lead to suboptimal benefits from the education system on a national scale.Footnote ¹⁶

These challenges to the creation of voucher markets have led to debates amongst economists about the opportunities to meet these conditions. Skeptical economists argue that the promised benefits will not materialize because the conditions for perfect competition cannot be met. Much of this debate is empirical, and at the level of specific services. The empirical evidence seems inconclusive; economists differ in their assessment of how strong the obstacles are.Footnote ¹⁷ However, the issues are not merely empirical. The term ‘design challenges’ has an inappropriately technocratic ring. As I will argue later, these issues raise thoroughly political questions about power and authority, equity and exclusion. But first we inquire how liberal theorists have related to this debate.

Pragmatic and Principled Liberals

Pragmatic liberals are mainly concerned with the end of universal public provision. They consider public finance necessary for reasons of distributive justice: because otherwise citizens who cannot pay for these services out-of-pocket will have no access. But when public funding is guaranteed, they are pragmatic about who provides these services. Mathias Risse, in a response to Julian Le Grand’s advocacy of quasi-markets, provides a good example:Footnote ¹⁸

This position is agnostic about the extent to which pragmatic liberals will be supportive of voucher markets: this is ‘left to efficiency considerations’. Therefore, liberal egalitarians – like their economic counterparts – may come down on either side of the divide between optimists and pessimists about the realization of the conditions for well-functioning markets. This is considered to be an empirical, not a philosophical issue. The pragmatic liberal is led down the economic path of technocratic tinkering with market conditions to see if a satisfying structure for these markets can be designed.

Note that pragmatic liberals’ concern for efficiency considerations means that optimal satisfaction of consumer preferences is their only relevant criterion at the level of means. Consumer choice is valued in so far as it enhances competition between providers, and thus leads to a better price/quality balance. There are also authors who stress the gain in freedom of choice as a separate normative ground for favoring markets, either because freedom of choice is intrinsically valuable or because it enhances well-being.Footnote ²⁰ However, other authors refuse to go down this road. For example, Risse explicitly claims it follows from Rawlsian egalitarianism that liberals should treat citizen-users as empowered choosers in some respects, but in other respects deny them choices within the public system so as to maintain that system as a whole.Footnote ²¹

Overall, then, the core conviction of the pragmatic liberal is that if (better: to the extent that) marketization leads to more efficient public services than direct public provision, then it can be endorsed by egalitarian liberalism. The means of public markets then deliver the best-possible realization of the liberal end of universal public services. Ends and means are treated as strictly separate matters.Footnote ²²

Now let’s turn to the principled liberals. They believe pragmatic liberals are mistaken to leave the question of means to economic theory. They deliberately collapse ends and means, by arguing that (at least some) public services are inherently public, and therefore cannot be left to the market. Different arguments are used to support this position – what follows is an attempt at reconstruction.

Common to many principled liberal arguments is an ontological starting point. Some goods can only be enjoyed as public goods, i.e. when government provides them uniformly to all citizens. They cannot, by their nature, be marketized. Some speak of ‘communal goods’,Footnote ²³ others of ‘solidarity goods’,Footnote ²⁴ ‘inherently public goods’Footnote ²⁵ or ‘irreducibly social goods’.Footnote ²⁶ This ‘impossibility theorem’ may seem a non-starter, since we can witness marketized provision of most of these goods in many places. However, appearances are deceptive. Principled liberals argue that once an inherently public good is delivered privately its nature changes. What we then observe is the provision of a surrogate with different characteristics.

This claim of a changing nature may be plausible or not. It may be plausible for some examples, especially for goods with a public goods-structure in the economic sense (non-rivalness, non-excludability). For many other examples, especially for divisible goods that can be delivered on an individualized basis (like health care and education), the ontological flavor of the argument may be misleading. But whatever of its plausibility, we can always ask the question: why should we have the public version of the good instead of the market version? To maintain the claim that marketized provision changes the nature of the good does not help, for we cannot the escape the question: so what? Why not go for the marketized version instead?Footnote ²⁷ To answer that question, the principled liberal argument needs to turn to normative grounds. And in fact, under the guise of ontological claims, most authors do argue that the public version is more desirable than the market version. Roughly, two types of arguments are prominent.

First, some claim that people have a preference for the public version. Fred Hirsch, in his famous attack on the ‘new commodity fetishism’, argued that people may intrinsically value the characteristics that come with public provision, such as the commonness of public places, the fact that a place is accessible to all without charge. In the market these people are badly served: ‘Their bad luck is that they derive utility from environmental characteristics that are outside of the capacity of the market to provide’.Footnote ²⁸ A similar argument is made by Michael Walzer in his discussion of the limits to the sphere of money and commodities.Footnote ²⁹ These arguments are problematic, because they may not be shared by other members of the community, who do not have these preferences and hence do not object to marketization. From their point of view, they must contribute to something they do not want; they could complain that a government catering to people on these grounds is being perfectionist. Elsewhere I have discussed this dilemma and argued that the stand-off between these two groups legitimates a political solution.Footnote ³⁰ Here let it suffice to say that the preference-argument is severely weakened by the perfectionist critique.

The second argument is that the public good in question is functionally necessary for its role in constituting a political community (or: social solidarity, cohesion, citizenship, social capital). Sometimes this is expressed in talking about the expressive role of having goods provided publicly.Footnote ³¹ One prominent example has been security provision by the police, which is seen as fulfilling this role of a ‘social glue’.Footnote ³² To prevent this argument from ultimately being grounded in a holist position (i.e. that the community is intrinsically valuable), it needs to take the further step that community is necessary for something which ultimately refers to the value of individuals. Liberal nationalists explain this in terms of a need for citizens to share a sense of collective destiny, if they are to accept state redistribution for public goods programs.Footnote ³³ The latter goods are in turn necessary for citizens to lead autonomous lives, so that the final basis of the argument is normatively individualistic.

There are several problems with this type of argument. One is that the argument seems circular: public goods provision is necessary to maintain support for a political community that in turn provides public goods. I hesitate to call this a vice, however, because we could just as well think this is a relation of mutual reinforcement (virtuous circle) between community and public goods. A more serious problem is that the argument cannot explain which goods should fulfill this function. Jonathan Wolff admits this when he says: ‘I am coming around to the view that, a few key elements aside, it may not matter so much what is in the market sphere and what is not, as long as the non-market sector is significant in size’.Footnote ³⁴ This is fine as far as it goes, but it means the argument cannot help us to decide whether to marketize a public service or not. Finally, the argument cannot explain why public financing is not enough as an expression of solidarity. Ultimately, it might be an empirical question how many goods need to be provided publicly to sustain a political community (this might lead to different conclusions for different cultures). This does not invalidate the argument completely, but it does not provide much help for our question either.

Both principled liberal arguments, then, have some value, but I also think both have important limitations. In the remainder of this article, I will present an alternative, which I hope is ultimately more fruitful in thinking about public markets.Footnote ³⁵

Revisiting Ends: Social Rights

The theory that I will propose in the remainder of this article diverges from the theories discussed so far in two crucial respects. It relies on a rights-based normative justification for public services: these services satisfy basic social rights to which every human being is entitled. In addition, it defends a crucial role for democratic decision making in public services. In this section I introduce these points and their connection to the marketization of public services. I will do so by arguing for three claims. First, citizens’ social rights provide a principled liberal justification in favor of public markets. Second, the same liberal framework also generates important objections to public markets. Third, a process of democratic decision making is needed to balance these considerations in concrete cases.

Revisiting Means: A Differentiation of Markets

Some may think that there is one major weakness in the liberal framework introduced in the previous section: it presupposes that it is possible to have democratic control over private providers. But, so the objection goes, once services are marketized, public control will prove exceedingly difficult, and powerful producers will be able to capture these markets and use them in their own interest. Thus, we face a hard choice: either to keep these services firmly in public hands (violating the value pluralism case for meaningful choice) or to accept full-blooded marketization (relinquishing control over them in democratically desired directions). The ‘markets as a mere means’ option, by wanting to have it both ways, is politically naïve. Liberal design will play into the hands of a neoliberal scheme in which public services are organized as a means for private rent-seeking. Let’s call this the ‘Neoliberalism Objection.’

Let us first reconstruct the objection theoretically. It is based, first of all, on the view that the exercise of power in markets is the norm, not the exception. Neoclassical economists tend to see power as a specific phenomenon (market power) arising only in contexts of market failure (monopolistic or oligopolistic markets). By contrast, the Neoliberalism Objection presupposes that power is an unavoidable part of all markets: the rules of the market and the ensuing market structure give power to some over others. This presumption can be defended by recourse to the analysis of power in markets given by Brian Barry. He showed that the widely accepted liberal claim that consumers can have power over firms depends on rejecting the picture of a kind of state of nature in which consumers and firms contract with each other from a baseline situation without any transactions (on that picture they do not exert power over each other but only make offers which may be mutually beneficial). When taking current levels of sales as a baseline, consumers can exert power over firms by threatening to make them worse off, i.e. by not buying their products any longer. Similarly, employers may exert power over employees, and capitalists over politicians.Footnote ⁵¹ On the basis of this conception of power, analyses can be made of the power balance between contracting parties by showing how the rules of the game alter the ability of each party to make credible threats to the others.

The Neoliberalism Objection uses this theoretical starting point for an analysis of public markets. In these markets, there are three main types of actors: consumers, producers, and the state.Footnote ⁵² The objection holds that within public markets producers will be able to exert power over both consumers and the state. This leads to a form of neoliberal governance. Liberal governance of society presumes a rather strict separation between public and private responsibilities. Neoliberal governance blurs these boundaries. The state is captured by special interests that use the machinery of government to further their private interests. In public markets, producers lobby governments for favorable rules and exploit open contracts with governments, so as to make excess profits to the detriment of accessibility and quality for consumers. Some of the market strategies mentioned earlier within an economic framework (such as cream-skimming or exploiting information failures) need to be understood in this light. Neoliberal governance in effect is a return to feudalism, in which the boundaries between private and public sphere are deliberately violated, so that political functions can be organized so as to favor targeted private interests.Footnote ⁵³ Once markets are opened, this degeneration from a liberal to a neoliberal governance structure automatically arises.

Having thus reconstructed the Neoliberalism Objection, we now respond to it. My response will be in two parts, partly affirmative, partly critical.

First, the theory defended here concurs with the objection in that it is critical of developments toward markets in which producers control public services and successfully lobby political bodies only to reap maximum profits. Producer-dominated markets embody a view of public services as a nexus of private contracts only; both between producers and consumers and between producers and government. This is a libertarian view, and, as Samuel Freeman has convincingly argued, libertarianism is antithetical to a liberal-democratic view of political authority. He too explains this with recourse to the analogy with feudalism: ‘Like feudalism, libertarianism conceives of justified political power as based in a network of private contracts. It rejects the idea, essential to liberalism, that political power is a public power, to be impartially exercised for the common good.’Footnote ⁵⁴

I propose that the notion of the common good mentioned by Freeman should be operationalized as the reconciliation of a set of potentially clashing liberal interests: on the one hand the liberal interest in markets for public services, grounded in the value of autonomous choice, and on the other hand the liberal interests which may serve as exceptions or restrictions to markets. Finding the common good means democratically balancing these interests which together make up the framework developed in the third section. From this perspective, I would agree with the Neoliberalism Objection that the detrimental effects of marketization on access and quality that occur where producers dominate both consumers and governments, are morally problematic. Liberalism requires a balancing between user and citizen interests in public markets. Producer interests in profit maximization are instrumental to these and should not predominate.

However, where I would diverge from the objection, is in its view that such a development toward producer dominance is a necessary consequence of introducing public markets. In effect, what the critics are claiming is that things go wrong as soon as one introduces public markets, and thus deliberately attempts to give leeway to private actors and price signals (the market as means), while at the same time wanting to keep the system under public control (the public ends). By contrast, I would argue that the dominance of producers in public markets is a possible, but not necessary effect. Much depends on the institutional set-up of such markets and this in turn depends on the balance of forces within politics. This somewhat more optimistic view of institutional opportunities for maintaining democratically controlled public markets is based on a two-pronged critique of the Neoliberal Objection, both conceptual and empirical.

On the conceptual front, the objection presupposes an overly sterile distinction between liberal and neoliberal governance. It presupposes an ideal-typical image of (good) liberal governance within which there is a clear demarcation between the private and the public, and (bad) neoliberal governance, where the boundary between public and private responsibilities is blurred. But it is philosophically controversial, to say the least, to presuppose that in social life a clear private/public distinction can be drawn.Footnote ⁵⁵ In the present context, drawing a strong demarcation risks distorting the picture both of what happens on the public and on the private side. On the public side, a system of purely public production is not necessarily less vulnerable to private rent-seeking. In such a system private interests can also creep in. Actors within public systems remain, after all, private individuals. Even if many of them develop a loyalty to their publicly defined roles, it remains possible that some of them abuse their roles for private gain. On the private side, the ideal-typical image presupposes that ‘normal’ markets for standard consumption goods are completely detached from public influence. But that too is misleading. Markets are always (also) publicly constructed, if only because governments and courts have to recognize and enforce claims about private property and contracting before any market can get off the ground.

There is thus no alternative to looking at public markets with an open eye. They may or may not undermine publicly set ends, depending on the institutional set-up. This brings us to the empirical point. Comparative empirical research by political scientist Jane Gingrich shows how political actors in different countries try to marketize public services according to their own ideological commitments. In her framework ‘the Left’ and ‘the Right’ in politics have preferences for different types of markets, with different balances of power between government, users, and producers. The outcomes of the political processes vary between countries and between markets, with producer-dominated markets as only one possible outcome.Footnote ⁵⁶ Other researchers have stressed, too, that politicians over the last decades have used specific forms of marketization for ideological and strategic reasons.Footnote ⁵⁷ This kind of research confirms that we need to look at the institutional details of marketization. Rules on market access, quality control, profit constraints, price regulation, etc. differ widely. There is no indication of a general tendency for public markets to escape democratic control.

Bringing together the conceptual and empirical point, the objection to public markets needs to be rejected. It starts from the legitimate (liberal) wish to safeguard public interests in public markets. It also starts from a credible picture of markets as always embodying power relations between actors. However, it overstates the dangers of private provision under public control, compared to available alternatives which are always also vulnerable to private rent-seeking. We need to move from ‘the market’ to a much more differentiated view of markets, the institutional frameworks in which they are embedded, and the different types of effects on the power relation between actors. Different actors will have the upper hand within public markets, depending on which rules of the game are adopted.

All of this does not diminish the fact that the Neoliberalism Objection points to a very important danger for public markets. It is extremely valuable as a warning signal. This is one more reason to be concerned with the quality of democratic decision making in this area.

From Ends to Means: Democratic Decision Making

The argument in favor of the need for democratic decision making was made earlier. As we saw, even when public service delivery has been marketized, there are still many decisions to be taken. Many of these decisions remain politically controversial and cannot be fully decided at the highest level, in national legal frameworks, for these leave many details unspecified. Concrete decision making at lower levels, in municipalities, in public organizations, and even in the professional execution of these services will often bear on the same sets of politically charged issues. This problem is framed in the economic approach: one of incomplete contracting between principals and agents. The solution within this approach is to remove government as intermediary principal/agent between consumers and providers and have a direct bargaining process between users and providers. The democratic approach defended here recognizes the same problem: the inherently incomplete, open nature of the decision-making process in these complex services. However, it presents a different solution. Government should not be withdrawn from the triangle with consumers and providers, since decisions at all levels are unavoidably politically charged. Democratic decision making is the best way to deal with such an unavoidably political predicament.

The philosophical background of this approach is a different way of thinking about ends and means. As we have seen, principled liberals often collapse ends and means: avoiding or allowing markets becomes itself a matter of ends. Pragmatic liberals distinguish ends and means very sharply and once public ends have been set they relegate thinking about means (i.e. the desirability of markets) to efficiency considerations. The perspective defended here conceptualizes ends and means as a continuum. ‘Ends’ refer to abstract moral considerations – about social rights, inequality, autonomy, etc. – which provide relevant considerations when evaluating the legitimacy of public markets. These have been identified in the third section above. When we descend to lower-level issues of institutional means, these moral considerations do not disappear; they rather re-appear transformed into very concrete policy issues and even professional decisions on the work floor. The moral dispute is carried over to lower levels.Footnote ⁵⁸

For example, at the highest level of abstraction, different theories of distributive justice are on offer, as an input into the question when inequalities on public markets are unjust. Some will defend a sufficientarian view, others a luck egalitarian view, still other a Rawlsian view, etc. But these views will also affect one’s position on lower-level policy issues in health care or education systems. Today, there may be a controversy about whether physicians should be allowed to give priority to treatments of employed persons over unemployed ones (even if this would benefit the latter as well). Tomorrow, it might be whether a hospital can refuse treatment of uninsured persons. Next week, whether it is acceptable to close down an emergency facility, which will mean that some citizens will have to travel further to hospital in an emergency than others. Next month, whether a physician should prescribe a cheaper drug to her patients than she used to, in order to save costs for her organization.

If we accept – as most theorists and citizens do – that the most abstract decisions about public services should be decided democratically, then the same should be true for many of these less abstract decisions specifying the more abstract ones. It is much too crude to see governments as setting the ends and market agents (public service providers and/or citizens in their role as users) as setting the means.Footnote ⁵⁹ The division of labor on which the edifice of public markets is built (public responsibility for finance, market execution of provision) needs revision. Public responsibility extends to details of provision, since the devil is in the details. This is not meant to deny that there will remain higher and lower levels of decision making. It would be absurd to make every doctor’s decision a matter of national parliament. Rather, the democratic approach calls for appropriate decision-making mechanisms at every level, since it recognizes that at every level in public service delivery the moral interests identified in the third section are bound to re-appear. Hence at every level democratic mechanisms are needed to balance these interests in a legitimate manner when they conflict.

This leaves wide open the exact forms this democratic engagement can take. What does favoring democratic decision making in public services mean concretely? A convincing answer to this question cannot be given in the abstract. We need to do justice to the specifics of different public services as well as different levels of decision making. Obviously, democratic control over teachers does not require that a teacher’s single decision to let his pupils practice math today instead of geography should be susceptible to majority-voting by his pupils (or their parents). Or that every parent in the country should partake in a national decision to allow experiments with parental co-payments at public schools. The teacher’s decision is part of an ongoing working relationship, and he needs to be accountable to his pupils and their parents in a more general manner. The decision on co-payments arguably needs to be debated by stakeholders (including representatives of parents), but ultimately determined by vote of parliament. I cannot here deliver fine-grained discussions of democracy for all public services and decision making at all relevant levels. Instead, I want to mention three pertinent issues that general democratic theory suggests would need resolution in order to develop a theory and practice of democratic decision making for public services.Footnote ⁶⁰

First, here as elsewhere in the debate about citizen participation, we face a choice between more representative and more direct mechanisms of democratic decision making. I mention one example of this tension. In the field of organization science, there is an increasingly widespread recognition that mechanisms of ‘exit’ need to be supplemented by (or even replaced with) mechanisms of ‘voice.’Footnote ⁶¹ However, many have argued that those individual consumers who participate actively, either by changing providers (exit) or by using complaint mechanisms or participating on patient assemblies and the like (voice), are not necessarily representative of the whole group of consumers. In response, some have suggested that we need to turn to mechanisms of delegated decision making, which would mimic representation in formal representative organs. Here, for example, patient organizations could make up for the unrepresentativeness of individualized voice mechanisms.Footnote ⁶² This proposal remains confined to representation of users. But even those who do not (at present) use a service have a stake in its delivery, if only because they contribute to it by paying taxes, and consume other services which compete for the same budgets. Therefore, we also need to think about adding representatives of the public at large to the democratic arena – but only at those levels where their interests are at stake.

Second, a theory of democratic decision making for public services needs to be situated with respect to the opposition between theories of aggregative and deliberative democracy.Footnote ⁶³ Deliberation has been seen by many as superior because it promises a more informed and reasoned democratic process. However, an important problem has been that citizens are not equal in their power, resources, and capacities to influence decision making.Footnote ⁶⁴ This puts a heavy burden on the organization of fair and inclusive deliberative processes. This problem is especially acute for public services. As we have seen earlier, many users of public services are in weak and vulnerable positions. They (partially) lack autonomy, which impinges on their capacities for choosing between services. These same deficits recur as a problem for participating in deliberations, although this will obviously differ from case to case (a victim from a car accident will lack the information about his injuries, but chronically ill patients may be very well informed about their illness). Similarly, we need to address other well-known problems of deliberative democracy, such as those relating to the procedures of deliberation and how these may (subtly) exclude or marginalize the voices of certain groups.Footnote ⁶⁵

Third, within general democratic theory a classical problem is how to decide the proper scope of the authority of a democratic community: who should be subject to its decisions, and how far does this authority extend into private life? In the context of public services this translates into the problem of the relation between democratic and market-based mechanisms. Traditional justifications for democracy often appeal to the fact that the authority of the state cannot be evaded (except by emigration, which is often prohibitively difficult). Because of its coercive power, citizens need to be able to participate in making the decisions that will eventually affect them. By contrast, private associations, since they are voluntary (exit is possible) are not subject to such a heavy justificatory burden. Given this dichotomy, increasing the scope for individual choice for public services would seem to suggest that we can do with a lower democratic burden. If a Catholic can pick a Catholic school, the state need not set up decision making at other (say, secular) schools to ensure that Catholic interests are represented there as well. Pupils of Catholic schools may be free on Catholic holidays, while other schools may make different choices. These and many other decisions would not require the democratic involvement of every citizen (whether directly or represented by government).

The argument in this article shows that this kind of reasoning is only partly correct. For example, if Catholics are an important part of the population in a given society, we could imagine that they have a legitimate interest that secular schools at least provide basic information about Catholicism, in order for their pupils to be able to participate in social life in that society. In general, part of what happens within the walls of a school will be legitimately parochial and this can be justified by the market-based mechanism of choice and exit. Another part of it, however, will be legitimately of concern to every citizen and this part needs to be justified by democratic mechanisms. The scope-question then is where to draw the line between these two parts. Which issues can be left to private actors, which ones need to be resolved collectively? That is itself a matter that needs democratic resolution.

All of these questions are opened up once we recognize that public services need not be uniformly provided by the state, nor completely left to the market. It is only in that configuration – markets as means to public purposes – that liberalism’s general commitments to democracy in political life and to markets in economic life come together on one and the same battle field. This requires a continuous democratic engagement with these public markets, both to ensure individual choice and to restrain such choices where necessary to protect effective social rights for everyone to enjoy these services.