Privatization, Structural Dependence, and the Problem of Legitimacy

: How should a normative evaluation of the merits and demerits of privatization proceed? In response to Joseph Heath’s approach to this question, I first argue that the difference between core and economic functions is not as relevant to establish the limits of privatization, as Heath suggests. I claim, second, that Heath problematically neglects the structural and aggregative effects of privatization. An instance of privatization that, if analyzed in isolation, is anodyne may no longer be so when seen as a further contribution to an already expansive process of privatization. Finally, I argue that Heath fails to consider the risks that pervasive privatization poses in terms of democratic legitimacy, by giving rise to a situation where, on the one hand, citizens are dominated by a privatized state and, on the other hand, the state itself is dominated by private actors.


I. INTRODUCTION
How should a normative evaluation of the merits and demerits of privatization proceed?In response to this question, Joseph Heath (2023) has advanced three main theses.First, what we may call an asymmetry thesis: "as a prelude to any detailed discussion of the merits of privatization", he argues, "it is essential to distinguish core functions of the state from what are often described as 'economic' activities" (2).While, as far as the coercive apparatus of the state is concerned, "it would not be surprising to discover that there are strong objections to privatization, stemming from basic philosophical principles" and, indeed, it would be "courting disaster to begin replacing elements of that apparatus with private organizations" (4), by contrast, within the realm of economic activities, there are cases of anodyne privatizations, understood as "privatizations that no reasonable person could object to" (2).
Second, and relatedly, Heath proposes a case-by-case thesis.Since, he argues, there are no global arguments for or against privatization, "each case must be evaluated on its merits" (17).The (de)merits of privatizing the management of highways may be very different from the (de)merits of privatizing the provision of childcare.Therefore, the only plausible way of assessing the justifiability of certain privatization decisions is to analyze the specific costs and benefits involved in each instance of privatization, assessed in its own terms.
But how should we think about the 'merits' in relation to which each case of privatization must be so judged?Heath's answer-his third thesis-is in line with mainstream economic theory: "state involvement in the economy should be guided primarily by the norm of efficiency, which is to say, the objective of correcting market failure" (36).When the market fails on both the demand and the supply side, then the state must act as both a purchaser and as a provider.But in most cases market failure occurs on one side only, and it is on that side only that state intervention is required.This may sound slightly confusing since, as Heath himself later shows, many types of privatization respond not to market but rather to government failures.For example, the privatization of highways often involves shifting from a system of tax-financing to a system of tolls (beneficiary-financing) to correct for the failure of the state, not of the market, to solve the collective action problem of congestion.Or, the outsourcing of state-financed service provision may be introduced in order to correct for the failure of the state to secure quality provision, because of scarce competition within the public sector.Further, markets and states can both fail, in which case, according to economic theory, nonprofits intervene (Hansmann 1980).For example, in the case of nursing homes, because of important information asymmetries due to the fact that the beneficiaries are themselves unable to assess and report on the quality of received services, states may turn to nonprofits, with the expectation that, because of nonprofits' non-distribution constraint, which prohibits the distribution of profits to managers, such organizations will lack an incentive to exploit their beneficiaries' vulnerability and ignorance.In any case, what matters is that, according to Heath, the merits of each case of privatization must be judged exclusively, or at least primarily in terms of efficiency.Let us call this the efficiency thesis.
In this paper, my aim is to question each one of the above theses.With regards to the asymmetry thesis, I will argue that, when it comes to privatization, the difference between core and economic functions may not be as relevant, let alone 'essential', as Heath suggests.Arguments against the privatization of core functions either distinctively apply to such functions alone, but they are generally untenable, or when more plausible, they would seem to extend, with similar force, to at least some economic activities as well.
As for the case-by-case thesis, I will contend that Heath's approach to privatization problematically neglects its structural and aggregative effects-effects that give us reasons to take seriously the overall scale of privatization, beyond specific instances of it, taken serially.An instance of privatization that, if analyzed in isolation, might well be anodyne may no longer be so when seen as a further contribution to an already expansive process of privatization.
Finally, in relation to the efficiency thesis, my view is that there is little reason to think that efficiency should be the supreme value when it comes to assessing the privatization of public goods.Heath is right that appeals to distributive equality and decommodification cannot often play the anti-privatization role that its defenders would like them to play (see also Cordelli 2020, chapter 1).However, this does not exclude the possibility that other values should play an important role in justifying limits to privatization, even when further privatization would be desirable in terms of efficiency.I will argue that democratic legitimacy is one such value.
Although I disagree with Heath's approach to the question of privatization overall, as it is summarized in the three theses above, I do not disagree with his main conclusions that, "someone who is a supporter of the welfare state might also support the privatization of certain state services in certain cases" (2) and that "examples of anodyne privatization would include […] the contracting out of public services in cases where relatively complete and enforceable contracts can be written (such as municipal garbage collection" (35).Indeed, unlike Heath, I interpret the recent philosophical literature on privatization, including its most critical contributors, as unanimously converging on these conclusions.For example, Aviay Dorfman and Alon Harel (2013), who present inherent objections to privatization, limit their objections to the management of prisons and military functions.They make clear that their objections do not extend to welfare and economic functions, leaving open the possibility that the privatization of at least some of these functions may well be anodyne.
Similarly, Debra Satz (2018) develops a pluralistic set of standards for assessing different cases of privatization, while affirming that "the case for or against privatization almost always depends critically on contextspecific details" (9).She precisely mentions the privatization of garbage collection as a generally unproblematic case.Even my own work on privatization (Cordelli 2020), which offers principle-based reasons to object to the privatization of some functions beyond putatively core ones, limits such reasons to either cases where, due to the unavoidable incompleteness of contracts or to the desirability of discretion, private actors are delegated quasi-legislative-not merely executive-decision-making powers, or to cases where privatization has reached a certain scale.I present no objection to isolated instances of "contracting out of public services in cases where relatively complete and enforceable contracts can be written" (35).Garbage collection may well be, in certain contexts, one of these instances.
I thus find Heath's conclusion largely unobjectionable ('largely' because, as I have mentioned, an instance of privatization that, if analyzed in isolation, might well be anodyne may no longer be so when seen as a further contribution to an already very expansive process of privatization).This is why, instead of primarily focusing on such a conclusion, I believe it is more fruitful to focus on Heath's overall approach to the question of privatization, as it is summarized in the three theses above.
Before I begin, a clarificatory note on the scope of my critique is in order.I suspect that much of the disagreement I have with the way Heath approaches the question of privatization derives from a more fundamental disagreement about the role of the state and its justification.Heath writes as if the state's fundamental rationale is instrumental, and public and private institutions are just interchangeable means for the fulfillment of independently defined ends.In other words, he appears to endorse what elsewhere I have called 'the interchangeability assumption'.The government's main business, in this view, is solving collective action problems, and supplying goods and services efficiently, rather than instituting and constituting just power relations among free and equal persons.This means, in turn, that what states do, precisely like what businesses do, can be almost exclusively assessed according to a consequentialist, efficiencyoriented logic.Elsewhere, I have argued at length against this instrumentalist conception of the state, and against the interchangeability assumption (see Cordelli 2020, chapter 2).However, for the purpose of the present paper, I will leave these more fundamental issues aside, in the hope CORDELLI / PRIVATIZATION, STRUCTURAL DEPENDENCE, AND THE PROBLEM OF LEGITIMACY ERASMUS JOURNAL FOR PHILOSOPHY AND ECONOMICS aa that even those who do not endorse my preferred, Kantian conception of the state could agree with the substance of my critique.

II. THE ASYMMETRY THESIS
Why draw a sharp line between core and economic activities?Why think that anodyne forms of privatization only pertain to states' 'economic activities', which do not constitute the coercive apparatus of the state and that, by contrast, it is "courting disaster to begin replacing elements of that apparatus with private organizations" (4)?On the one hand, Heath would seem to think that, in the case of core functions alone, there may be inherent reasons, based on "basic philosophical principles" (4) to object to privatization-reasons that are not contingent on state regulation or other contextual factors.On the other hand, he claims that: Most obviously, because of the dangers of corruption, there is considerable merit in compensation systems that neutralize pecuniary concerns for state employees who are directly involved in the exercise of coercive power.(4) Such systems are distinctively offered by the public sector, which can mitigate the risk of corruption by providing strong employment security, low financial incentives, as well as "strong shared values" (4).These remarks offer a more contingent argument-about both the importance and likelihood of avoiding corruption-against the privatization of core functions, which is supposed to not extend with equal force to economic activities.Both sets of arguments, I will now suggest, ultimately fail to support the claim that an analysis of privatization should draw a sharp line between core and economic activities.What objection, stemming from 'basic philosophical principles' could there be that would rule out the privatization of core functions, but not the one of economic activities?Since Heath does not defend any specific objection, I will consider the main arguments already available within the philosophical literature.One such objection is that the privatization of core functions amounts to an abdication of a legitimacy-grounding responsibility on the part of the state.As put by Peter W. Singer (2003, 226), in the case of the privatization of military functions "When a government delegates out part of its role in national security through the recruitment and maintenance of armed forces, it is abdicating an essential responsibility".Here by 'essential' Singer means a responsibility upon the adequate dischargement of which the duty to obey of citizens depends, the assumption being that a state's right to rule depends on its ability to protect its citizens' basic rights, through the provision of policing and national defense (see Pattison 2010, 436).Such arguments, however, would not be able to rule out the privatization of core functions as we know it, because such privatizations, as Heath rightly points out (3), do not generally entail full state withdrawal but rather the contracting out or outsourcing of prison management, military, and police functions, for which the state retains primary responsibility.A state could, at least in principle, perfectly fulfill its responsibility to protect its citizens' rights by delegating its fulfillment to a set of proxies, including regulated private actors, rather than by providing those functions directly.The objection is thus untenable.
A more plausible, principle-based objection would be that core functions cannot be legitimately performed by private actors, because the latter either lack the legitimate authority to exercise coercive powers or they cannot communicate condemnation for public wrongs 'in the name of' the public, which is to say, they cannot punish (3).A version of this argument is defended by Dorfman and Harel (2013).Whereas this objection arguably succeeds in ruling out the independent exercise of coercion or punishment by private agents, including all forms of vigilantism, it arguably cannot rule out cases, such as contracting out or outsourcing, where private actors are formally authorized by the state to exercise coercion or inflict punishment on its behalf, and where private actors act within the boundaries of their authorized mandate (or contract).After all, if I authorize an agent to perform an action on my behalf or to communicate a message in my name, and that agent acts within the boundaries of my authorized mandate, we can say that the agent has acted on my behalf or communicated the message in my name, and this is so even if the private agent has necessarily exercised some discretion in deciding how to act (see Cordelli 2020).
But now assume that private actors, even if formally authorized and regulated by the state, would still lack the capacity or moral standing to exercise certain powers or to make certain decisions in the name of the public, that is to say, in a genuinely representative capacity.Elsewhere I have defended a version of this position (Cordelli 2020, chapter 5).There I argue that to act as a representative of the public, an agent must be able to exclude certain reasons of a non-public nature from its deliberation and that private actors, even when acting under state authorization, fail to meet this condition.This argument, however, does not rule out the privatization of core functions only, for the simple reason that policing and war-making are not the only functions that ought to be exercised in a representative capacity-'in the name of all'.We may have very good reasons to want decisions on, say, the boundaries of freedom of speech within the digital public sphere to be made in the name of all, and to question on these grounds Twitter's standing to make such decisions in a representative capacity.For the same reasons we may also question Bill Gates' capacity to set global health policy priorities in the name of all, even if such decisions do not pertain to the state's monopoly over the exercise of coercive powers.
These cursory remarks are just meant to show that finding plausible principle-based objections to the privatization of core functions might be much more difficult than Heath suggests.If such objections can be found, they are likely to extend to some 'economic activities' as well.
But what about the argument from corruption?As we saw, Heath claims that the avoidance of corruption may be a strong, and perhaps sufficient, reason against the privatization of core functions.Here the concern with corruption can be understood in two different ways.First, there is the concern that the presence of pecuniary incentives will lead private actors to perform their functions in ways that undermine or corrupt the desirable or just exercise of those functions, or of the powers involved in that exercise.For example, military contractors have often been accused of fighting wars or conducting peacekeeping operations in ways that lead to either the protraction or the repetition of conflicts, rather than to their prompt resolution, because they benefit from that protraction or repetition (Pattison 2010, 446;Taylor 2015, 4).Second, there is the concern with undue influence in politics.For example, the private prison industry has been accused of lobbying for, and of being secretly involved in the drafting of, laws and policies that would increase rates of incarceration (Satz 2018).The risk here is of corrupting both the political process and the social goal of diminishing such rates.Now, if such concerns offer a sufficient or, in any case, a very strong reason against privatizing coercive functions, why is this reason not equally strong when it comes to other forms of privatization?Consider, as an example, the privatization of healthcare in the Italian region of Lombardy-the first region in the western part of the world to be hit by Covid 19.One reason why Lombardy was found completely unprepared in its response to the pandemic, compared to other Italian regions, in spite of being among the richest regions of Europe and of priding itself for having an excellent healthcare system, was largely a consequence of the privatization of such systems since the 1990s (see, for example, Sartor 2021).At the beginning of the pandemic, almost 50% of healthcare providers in Lombardy were privately owned and managed, although the state still acted as the main purchaser of their services.To make profits, such providers had progressively reduced the provision of preventative care and emergency services.These were the least profitable of all healthcare services, and yet the most needed not only to face pandemics but also to reduce the number of preventable illnesses across the population.Private providers had, instead, focused their investments on the provision of highly specialized, and much more lucrative, curative medicine.Although we would not generally describe the behavior of such providers as an "abuse of power" (4), the fact remains that their economic incentives corrupted, indeed perverted, the appropriate balance between preventive and curative healthcare, thereby undermining a rational and just provision of healthcare services, hence costing illnesses that could have been avoided and lives that should have been saved.Similar considerations extend to the problem of undue influence on politics.Obviously, not only entities which perform core functions can exercise such influence.Like private prisons can lobby for higher incarceration rates, so pharmaceutical companies who enjoy a monopoly or quasi-monopoly on life-saving vaccines can lobby against reforms to intellectual property that would allow the cheaper reproduction of those vaccines (see, for example, Corporate Europe Observatory 2021).Stakes for the wellbeing of citizens, in both the healthcare and the pharmaceutical cases, seem as high as in the case of abuses of power by prison managers, nor there is a reason to think that the problem of perverse incentives, and of corruption can be more easily avoided, through regulations, in the case of healthcare providers or pharmaceutical companies than in the one of prisons or military firms.Therefore, if corruption is the most obvious reason to limit the privatization of core functions, such reasons would seem to extend with equal force to at least some 'economic activities' as well.
Interestingly, however, Heath also claims that, at times, avoiding corruption provides a reason for privatization, not against it.In his words: The accumulation of significant resources and economic power within the state can be, in effect, corrupting, benefiting constituencies whose interests are poorly aligned with the general public interest.( 27) Cases may arise in which "it is easier to maintain that alignment through external contracting than through internal supervisory arrangement" (27).As an illustrative example, Heath mentions the case of elected representatives from coal producing regions in the US who would likely oppose any effort at climate change mitigation.This fact, Heath says, is generally seen a consequence of successful lobbying by the coal industry."But now imagine"-he adds-"if state governments owned the coal companies!This would make things worse because public managers are immune from lobbying restrictions imposed on private firms" (27).Fair enough.But why, then, does not this rationale suffice to also justify the privatization of core functions, such as policing or military functions?Indeed, without great effort, one could imagine elected representatives from very red states in the US to be similarly counted on to oppose any effort at diminishing criminalization and incarceration rates, and at ending ongoing war conflicts.This could be understood as a consequence of lobbying efforts from the private prison and military industry.But now, one could say, imagine if all prisons, as well as the military security industry, were publicly owned and managed!This could make things worse for it would provide prison managers and the military with an incentive to lobby for policies that would increase criminalization rates, or for the continuation of ongoing conflicts, without being subject to the same restrictions as private firms.Why not then privatize prisons and the military industry, so as to prevent empowering constituencies whose interests are poorly aligned with the general public interest?Like the corruption-based argument against privatization seems to know no sharp line between core and economic functions, similarly the corruption-based argument for privatization knows no such line.
The point of this section has been to show that principled arguments against the privatization of core functions either distinctively apply to such functions alone, but they are generally untenable, or when more plausible, they would seem to extend, with equal force, to at least some economic activities as well.Similarly, arguments for the privatization of economic activities may extend, with equal force, to core functions.One cannot, therefore, coherently keep a sharp line between core and economic functions, supporting the privatization of the latter (at least in some cases), while ruling out the privatization of the former.

III. THE CASE-BY-CASE THESIS
I now turn to the case-by-case thesis: the idea that a normative assessment of the justifiability of privatization decisions should proceed by analyzing the distinctive benefits and costs, in terms of efficiency, involved in particular instances of privatization, taken separately.Heath's case-bycase approach has the advantage of clearly showing how and why the benefits (for example, decongestion) and thus the reasons that may prima facie justify the privatization of certain goods, say, highways (that is, moving from a tax-financed system to a toll system) may be very different from the benefits (for example, increased competition), and thus the reasons that may prima facie justify the privatization of other goods, say, public service provision (for example, the outsourcing of healthcare or welfare provision).Similarly, the costs involved in some types of privatization (for example, the problem of contract incompleteness in cases of outsourcing) may be absent in other types of privatization (for example, the selling-transfer of ownership-of a public industry).
The problem with a case-by-case analysis, however, is that it both neglects and obscures the aggregative, structural, and dynamic effects of privatization, especially on (i) the overall balance of power between the public and the private, states and private corporations; and (ii) the relationship between citizens and their state.

III.I. The Balance of Power between the Public and the Private
The overall scale of privatization, above and beyond the particular kind of privatized functions and the specific types of privatization, affects both the level of dependency of the state on the private sector and its capacity to regulate and control the later.
Dependency arises because the more a government privatizes, the more it loses in-house capacities and know-how to perform a variety of functions directly, the more it becomes dependent on the private sector for the performance of essential tasks.In turn, the higher its dependence on the private sector, the more the state will be vulnerable to, and powerless in front of, pressures from such a sector-a sector populated by actors with a vast amount of economic resources and, in a context of international competition, with the power to threaten to bring their resources elsewhere.In the words of Henry Farrell (2018): As more aspects of the economy are privatized, it becomes easier for actors who might benefit from privatization to press the state to make CORDELLI / PRIVATIZATION, STRUCTURAL DEPENDENCE, AND THE PROBLEM OF LEGITIMACY ERASMUS JOURNAL FOR PHILOSOPHY AND ECONOMICS aa further concessions.As states become more reliant on the private sector for information and resources, they become more inclined to acquiesce to the demands of private actors.( 175) The problem of dependency is a problem of scale, for whereas the privatization of some, say, military functions or IT functions may not suffice to put business in a privileged position vis a vis the state, the privatization of most or all military and IT functions will.This is also because, the wider the loss of capacity and know-how on the part of the public sector, the more difficult it will become for the state to initiate a process of insourcing to avoid undue pressures.The problem of dependency is further worsened by the likelihood of a 'brain drain' from government to the private sector, which can generally afford to pay higher salaries (Freeman and Minow 2009;Verkuil 2007;Michaels 2017).
In sum, privatization can generate a form of structural dependency, which in turn changes the overall balance of power between the state and market actors, giving rise to a situation where the latter have the incentives, resources, and power to pressure and, ultimately to dominate, the former.Because the problem of dependency is conditional on the overall scale of privatization, it remains invisible from the perspective of a caseby-case approach.
One could, however, respond that the problem of dependency can in large part be resolved by means of regulation.But this response risks being circular, because the state's effective capacity to keep private actors under appropriate control and accountability standards itself depends on its level of dependence on the private sector-sector that can use its resources to impede needed regulations.Even if we leave this problem aside, there are further reasons to think that the scale of privatization can undermine a state's regulatory capacity.This is because, as Heath himself acknowledges (p.), if a government privatizes because it lacks the expertise to solve a certain problem, then, it will also likely lack the capacity to design contracts that specify its needs with a level of detail sufficient to protect itself from exploitation by private contractors.To which we may add that, if government lacks the capacity to directly perform certain functions or to do so efficiently, it will also likely lack sufficient capacity to coordinate, plan, oversee and regulate those to whom those functions are delegated, and to do so efficiently (Freeman and Minow 2009).This problem is made more acute by the transnational nature and international mobility of many private actors, which makes their tracking and monitoring particularly difficult.Take, as an example, the progressive expansion of contracting in the U.S. military during the first decade of the 21st century.The Department of Defense's contracting budget increased by more than 100% between 2000 and 2005.This expansion led to a situation in which public officials did not even know, and were not able to easily find out, how many private contractors were working for the military (Minow 2009, 16).Monitoring was rendered particularly difficult by the transnational and mobile nature of private security firms.
This problem of control is further intensified by both the likelihood of brain drain from the public to the private, and by the fact that, insofar as government agencies often privatize to save costs, they are then reasonably reticent to hire new personnel to supervise their private contractors.For example, while federal spending in the USA exponentially increased between 1960 and the first decade of the 21st century, much of which was directed to finance newly contracted services, the number of civilian workers remained pretty much unchanged (Dilulio 2015).Who was then supposed to monitor and supervise the also exponentially higher number of proxies (about 40% of the workforce employed by the U.S. government) who were required to manage that spending?This led to a situation in which, as Paul Verkuil (2009, 313) puts it, "if government does not have adequate personnel to oversee its outsourcing, it does not have adequate personnel to read the reports on outsourcing submitted by its private overseers."In other words, the more government outsources, the less capacity it is likely to retain to gather basic information about performance, costs, and outcomes, and thus to choose competent contractors, as well as to enforce contractual terms.
Finally, as law scholar Jon Michaels (2017, 131) explains, privatization often achieves its purposes by "swapping out heavily regulated and dutybound civil servants for less regulated (and therefore more pliable) private contractors".Indeed, many of the standards of accountability that apply to public actors do not extend to private contractors.Therefore, the more expansive the scope of privatized functions, the lower the degree of control that a democratic state, and thus also its citizens, retain on the performance of those functions.
The problem of control, like the one of dependency, is a problem of scale.It is a consequence not of particular instances of privatization, taken serially, but of the aggregate effects of the systematic use of outsourcing on the capacity of the administrative state to retain adequate CORDELLI / PRIVATIZATION, STRUCTURAL DEPENDENCE, AND THE PROBLEM OF LEGITIMACY ERASMUS JOURNAL FOR PHILOSOPHY AND ECONOMICS aa control over the private sector, when such a sector is hired to act as its agent.A case-by-case approach risks obscuring this problem as well.

III.II. The Relationship Between Citizens and their State
But what about the effects of privatization on citizens' relation to the state?Heath argues that "citizens formulate their opinions about the qualities of the state on the basis of their interactions with state-level bureaucrats" and that "to the extent that they develop a negative attitude, they tend to apply it to all its activities, optional or not" (37).Insofar as public sector organizations lack the right incentives to provide polite and responsive costumer service, Heath concludes that states have reasons to avoid providing specific, optional services, such as running airlines or hotels.Perhaps.But Heath does not consider the overall effects of privatization on citizens' attitudes towards the state.There are, indeed, important reasons to believe that the privatization of public services, when expansive in scope, has significant effects on citizens' civic vigilance and, more broadly, on their interest in politics.
One reason, and a direct and foreseeable consequence of the loss of control analyzed above, is that privatization makes abuses less detectable.As Jody Freeman and Martha Minow (2009) observe: Outsourcing impairs the visibility necessary to check for such abuses [like waste and fraud], because private companies control info about cost, performance, and other vital data that otherwise would be open to review by government agencies.(5) Now, the harder it becomes for citizens to find out what happens, the more they should be vigilant, but the less inclined to be vigilant they will be.For if citizens cannot see that something is wrong, they will take themselves to have little reason for vigilance.Further, the harder it is for them to find information about abuses, the more demanding their exercise of vigilance will become, the less inclined to be vigilant they will be.
But there is more.The privatization of public services, even when limited to their provision, occludes the role of government behind a myriad of market and charitable actors through which the government comes to provide needed benefits to its citizens.Privatization thus directly contributes to what Suzanne Mettler (2011) has called the 'submerged state'."The policies of the submerged state"-Mettler explains-"remain largely invisible […] Even when people stare directly at these policies, many perceive only a freely functioning market system at work" (5).This in turn "leav[es] citizens unaware of how power operates, unable to form meaningful opinions, and incapable, therefore of voicing their views accordingly" (5).
Insofar as, in their daily lives, citizens experience government mainly through the provision of goods and services, when people do not perceive their own government as the main provider of the benefits they need, they see little reason to care about their government and thus to actively participate in politics.1It follows that, in contexts where 'the face' of government is largely privatized, citizens' interest in politics, including the urge to ensure their government's accountability through continuous vigilance, tends to diminish (as the empirical literature on the submerged state confirms), and civic apathy to grow.Generalized across a polity, therefore, pervasive privatization "contribute[s] to civic disengagement, including ignorance of public affairs, disenchantment with government and political apathy" (Norris 2000, 309.See also Cordelli 2020, chapter 4).Privatization thus generates a problem of vigilance, beyond those of dependency and control.Like the latter, also the former is not a problem that can be captured by a 'case-by-case' analysis of privatization, for the outsourcing of some specific healthcare services to private providers here and there or the existence of a few private schools or childcare centers as supplements to public providers, may not create any problem of civic apathy.But the systematic outsourcing of those very same public services may.
The point of this section has been to argue that instances of privatization cannot be assessed 'case-by-case', without taking into account the aggregative effects of privatizations, as well as the scale of privatization already in place in a given society.A same decision to privatize healthcare services may have different implications for the overall balance of power between the public and the private sector, as well as for the relation of citizens with their state, in different societies, depending on what else and how much has been already privatized in those societies.This interestingly also means that cases of privatization that may count as perfectly anodyne, if analyzed in isolation, may become more problematic if seen as contributions to the further privatization of a system of provision that is already largely privatized.A structural or systemic analysis of privatization is thus needed, above and beyond a 'case-by-case' one.

IV. EFFICIENCY THESIS
If my, admittedly cursory, analysis of the three problems of dependency, control and vigilance is sound, then privatization, especially when expansive, would seem to generate a serious problem of democratic legitimacy.Avoiding such a problem, in turn, provides us with strong reasons to limit the scale of privatization, even when imposing these limits would come with some efficiency costs.
The problem of democratic legitimacy I have in mind has three dimensions.The first dimension follows from the problem of dependency and amounts to the domination of the state by private actors.Privatization provides private actors with incentives to pressure the state to enact policies that will continue the privatization process.As we saw, the more extensive the privatization of public functions, the more dependent the state on the private sector, the more powerless it becomes to reject those pressures, or to start processes of insourcing to reduce them.The result is both the risk of domination of the democratic process by private interests and the violation of a basic principle of political equality-some parties (that is, market actors) enjoying much more political power and opportunities for political influence than others.Both concerns, note, directly threaten the legitimacy of the democratic process, regardless of whether they lead to more or less efficient outcomes.
The second dimension follows from the problem of control and amounts to the domination of citizens by a privatized state.Here the problem is that if, (i) due to the pervasive problem of contract incompleteness, private contractors are often left with wide degrees of discretion on how to interpret the contract and to perform relevant functions, and if, as we saw, (ii) the more a state privatizes, the less likely it is to retain the effective capacity to exercise appropriate forms of control, monitoring and supervision over private contractors, then, the obvious result is that (iii) citizens become subject to unaccountable exercises of discretionary powers, that is to say, subject to the arbitrary discretion of private state contractors.Such discretion may concern decisions such as: the setting of eligibility criteria for the allocation of unemployment benefits or welfare services; the prioritization of claims to certain kinds of medical treatment over others; determinations concerning the kind of infringements for which inmates can be sanctioned within a prison; and so on and so forth (see Cordelli 2020).The problem with cases in which such discretion is exercised arbitrarily, because private state contractors remain insufficiently monitored or unaccountable, is not that the democratic state is CORDELLI / PRIVATIZATION, STRUCTURAL DEPENDENCE, AND THE PROBLEM OF LEGITIMACY VOLUME 16, ISSUE 2, WINTER 2023 aa dominated by private interests but rather that citizens are dominated by an internally privatized state that rules their lives arbitrarily, through its contractors.Once again, this concern is independent of whether private actors exercise their discretion in ways that benefit or rather undermine overall efficiency.
The third dimension follows from all the three previously analyzed problems-dependence, control, and vigilance-and amounts to the abdication of the necessary conditions for democratic self-government, through their progressive erosion (see Cordelli 2020, chapter 4).In a society where: (i) the political process is systematically dominated by private interests and thus the basic principle of political equality is violated; (ii) citizens lack efficacious control, through their institutions, over those (private) agents who are left with the responsibility to manage public funds or to provide essential public services to them; and (iii) the risk of civic apathy and indifference is widespread, then the basic conditions of democratic self-government are compromised.Yet such a society is, as we saw, a foreseeable consequence of the systematic privatization of public goods, functions, and services, precisely because of the problem of dependence, control, and vigilance that such privatization structurally produces.I say 'structurally' for many of the causes of such problems do not pertain to the subjective inclinations or corrupt intents of particular agents (see Young 2011).Rather, they derive from objective conditions, such as the presence of specific economic incentives; the overall balance of power between the public and the private; the transnational nature of many private actors, as well as the existence of international competition; the problem of brain drain, which follows from the fact that market actors can pay higher salaries; and so on and so forth.
Jointly taken, the three concerns just analyzed-the domination of the political process by private agents; the domination of citizens by a privatized state; and the erosion of the basic conditions of democratic self-government-amount to a very serious deficit of democratic legitimacy.We may call this 'the problem of democratic legitimacy'.The prevention of this problem provides strong reasons to limit the overall scale of privatization, even when more privatization would mean more efficiency.
Why, one may ask, should legitimacy have priority over efficiency?I cannot provide an exhaustive answer to this question here.I will limit myself to offer a hypothetical case that, albeit far-fetched, provides intuitive support to the priority of legitimacy (see also Stilz 2016).Imagine Germany tomorrow conquered Italy but only decided to take over the CORDELLI / PRIVATIZATION, STRUCTURAL DEPENDENCE, AND THE PROBLEM OF LEGITIMACY ERASMUS JOURNAL FOR PHILOSOPHY AND ECONOMICS aa management of the Italian welfare state, promising higher efficiency and quality of service, while leaving other domains of state action untouched.Imagine Germany delivered on its promise.Italians would still have a complaint, and I would say a rather strong one, against the German takeover, and this is so even if we assume such a takeover could happen without the use of violent means.The complaint would be grounded on the value of collective self-determination, and on the lack of legitimacy of the German government vis-à-vis Italians.Italian citizens could rightfully complain that, despite its efficiency gains, the takeover would deprive them of the right to collectively decide, as equals, on important decisions concerning their society, to control those decisions by holding accountable those empowered to make them, and would also subject them to a dominating, because alien and unaccountable, will.
Similarly, even if the systematic privatization of 'economic activities' could lead to more efficient outcomes, citizens may still reasonably complain that, by giving rise to the legitimacy problem, such privatizations would subject them to an arbitrary form of ruling and would deprive them of the basic conditions of collective self-government.
In sum, once we leave behind a case-by case approach and focus on the aggregative and structural effects of privatization, legitimacy concerns enter the scene.Once such concerns are considered, it becomes clear that there may be very good reasons to limit the overall scale of privatization, even in cases where more privatization, by assumption, would come with efficiency gains.Efficiency, it seems, is not the only, or even the primary value in light of which the 'merits' of privatization decisions should be assessed.Further, anodyne forms of privatizations may become more problematic, when seen as further contributions to an already expansive process of privatization.

V. CONCLUSION
In this paper I have argued, contrary to Heath, that a philosophical assessment of the justifiability of privatization decisions (i) need not start by drawing a sharp line between core functions and economic activities; (ii) should not exclusively proceed on a case-by-case basis, thereby neglecting the aggregative and systemic effects of privatization on the overall balance of power between the public and the private, and the relationship between citizens and their state; and (iii) should assess the (de)merits of privatization not only in terms of efficiency, but also in light of other values such as legitimacy.I have further suggested that the scale of CORDELLI / PRIVATIZATION, STRUCTURAL DEPENDENCE, AND THE PROBLEM OF LEGITIMACY VOLUME 16, ISSUE 2, WINTER 2023 aa privatization matters beyond the kind of privatized functions and the modalities of privatization.Instances of privatization that, if taken in isolation, may well be anodyne, become more problematic, normatively speaking, when seen as further contributions to an already very expansive process of privatization.