David Landau, “The Reality of Social Rights Enforcement,” Harvard International Law Journal (2012), excerpt

David Landau, “The Reality of Social Rights Enforcement,” Harvard International Law Journal 53 (2012), excerpt (footnotes omitted)

Introduction

For all practical purposes, the debate about whether to include social rights in constitutions is over. Social rights are rights of citizens to receive services such as food, health care, housing, and social security. The U.S. Constitution does not include any of these rights, and most American scholars have long taken a position against their inclusion. But the American position against social rights is an outlier; there is now a “near consensus” (outside of the United States) that countries should include such rights in their constitutions.¹ Moreover, there is an increasingly vibrant and varied jurisprudence on what these rights mean and how they should be enforced. Social rights are not mere paper rights; courts around the world are actively enforcing them.²

However, there is a basic disconnect between the theoretical claims being made about the enforcement of social rights and the empirical realities of their enforcement. In the theoretical literature, scholars equate a robust enforcement of social rights with the advancement of the prospects of marginalized groups--by ensuring that citizens have minimum levels of things like food and shelter, the courts will improve the lot of the poorest members of society. Yet much of social rights enforcement is aimed not at the poor, but instead at middle- and upper-class groups. When courts in the developing world prevent pension reforms or salary cuts that would affect civil servants, when they order the state to give an expensive medical treatment or pay a pension to a middle-class professional, or when they force the state to raise subsidies for homeownership, they are deciding cases that help mainstream rather than marginalized groups.

One aim of this Article is simply to marshal empirical evidence and explanations showing that most of the literature mischaracterizes what social rights enforcement is--courts can aggressively enforce these rights and yet do little to affect social transformation.³ … It suggests at least two major reasons for these trends. One important factor is the nature of the judiciary. Despite the extensive literature in constitutional theory on counter-majoritarianism, courts are actually pro-majoritarian actors in many circumstances. Indeed, they are often populist actors--they sometimes favor middle class groups with social rights like food and housing precisely in order to gain political support. Another reason for this trend is that courts are likely to choose certain remedies because of ideology and resource constraints, and these remedies are particularly ineffective at targeting lower class groups.

Thus, the second major aim of this Article is to reorient the very rich, but in my view misguided, debate about how social rights should be enforced. Scholars led by Mark Tushnet and Cass Sunstein have argued that “weak form” or dialogical enforcement of social rights, whereby courts point out political failures to fulfill these rights but generally leave the remedy to the discretion of the political branches, is the best way to balance a desire to enforce social rights and the legitimacy and capacity strains that such enforcement places on courts.⁴ The dialogue-based approach has not really been used outside of South Africa, and in that country it has not accomplished much. Systematic failures in both legislative and bureaucratic politics in developing countries make dialogic approaches unlikely to work in those countries--the intended recipient of the dialogue is unlikely to respond effectively.

Instead, courts have relied mainly on two models of social rights enforcement: (1) in an individualized model, courts give a single remedy to a single plaintiff for provision of a treatment, pension, or subsidy, but tend to deny systematic remedies that would affect larger groups; (2) in a negative injunction model, courts strike down benefit cuts or other laws that change the social benefits being given in the status quo. Courts focus on these two models because they look most like more traditional modes of judicial review. However, both models have a very pronounced tilt towards higher income groups; they are unlikely to do much for poorer citizens. Moreover, they appear to do little to improve bureaucratic performance.

All of this argues for remedial innovation, but toward stronger forms of review and judicial supervision, not weaker ones as argued by Tushnet, Sunstein, and others. Experience in both Colombia and India has shown that more aggressive, unconventional enforcement strategies--especially the judicious use of structural injunctions--can more effectively target social rights' interventions towards the poor. Moreover, these strategies may be more effective at strengthening civil society groups and at inducing important changes in the bureaucracy. The conclusion is not that structural injunctions are the right answer to all social rights problems; they will fail in many political contexts, and the resource costs that they will place on courts may be too high to pay in many circumstances. It is that there is a desperate need to innovate with aggressive remedies if social rights are to live up to their transformative promise….

I. The Existing Debate

A. The Debate on Inclusion of Social Rights in Constitutions

…Still, the debate about the appropriateness of these [economic and social] rights has continued to be important to scholars and constitutional designers. An important group of scholars, particularly in the United States, continues to defend the position that social rights have no place in a constitution. The major arguments here revolve around the undesirability of judicial enforceability of these rights, and focus on two prongs: first, that judges lack the democratic legitimacy to enforce these rights, and second, that they lack the institutional capacity.¹⁰ The argument begins by positing that social rights are a subspecies of positive rights, which entail the right to receive something from the state rather than merely requiring the state to leave one alone.¹¹ Enforcement of these rights might require, then, that the judge order the state to provide people with goods or services, which would raise the specter of “the courts running everything--raising taxes and deciding how the money should be spent.”¹² Judges lack the democratic legitimacy to carry out this kind of policymaking, and they lack the capacity to do so. Courts are unsuited to decide where to spend the state's limited resources, and they will have trouble giving precise content to vague rights of the sort of the right to food or housing.¹³ Michelman's observation that courts perform poorly when adjudicating “polycentric” issues is particularly applicable to socio-economic rights; they have an inherent “[r]aging indeterminacy.”¹⁴ The result of all this is that courts are unlikely, in practice, to enforce socio-economic rights: they will be unwilling to incur the wrath of the political branches or to fulfill undertakings located so far beyond their own capacity.¹⁵ Courts will auto-limit in order to avoid sanctions from other branches of government or from the public: “It is futile to rely on the judiciary to provide basic welfare for the disadvantaged, if the political branches are unwilling to do so.”¹⁶

These arguments have been attacked by a different group of scholars, who argue that social rights are actually not different from traditional, first-generation rights, and can and should be enforced by courts. The distinction, these scholars argue, may even be “meaningless.”¹⁷ These scholars argue that social rights have a negative dimension as well as a positive dimension: enforcement may often require that courts enjoin states from taking some action that threatens social rights (for example, industrial development that threatens the right to health, or forced evictions from slums that threaten the right to housing).¹⁸ They also argue that enforcement of civil and political rights may often require the spending of significant amounts of state resources--for example, the right to a fair trial requires the state to spend significant amounts of money.¹⁹

On the legitimacy and capacity points, these commentators note that the enforcement of traditional civil and political rights can also involve the court in complex remedies (the U.S. school desegregation and prison reform cases are examples), and that courts can (albeit perhaps awkwardly) develop the capacity to deal with these sorts of cases.²⁰ Finally, these scholars note that courts can undertake many types of social rights enforcement without provoking unduly complex issues of enforcement or policy line-drawing--in many cases, the court can provide an individualized remedy to a single plaintiff, which obviates the need to make a large-scale intervention in public policy.²¹

The truth in this debate is almost certainly somewhere in between. That is, the critics of the conventional view are right that social rights enforcement is not always and inevitably different from negative rights enforcement. By restricting themselves to certain kinds of cases and certain remedial techniques, courts can assimilate enforcement of social rights to enforcement of more traditional kinds of rights. But there is often a difference of degree. As Tushnet notes, “it is not that recognizing social and economic rights would have budgetary consequences, while recognizing other constitutional rights does not . . . . Protecting background private law rights and first- and second-generation constitutional rights is cheap, though not free. Protecting social and economic rights is expensive.”²² Moreover, while negative injunctions and individualized remedies could likely enforce some kinds of social rights, the enforcement of many kinds of rights are likely to require the creation of new programs.²³ These tasks are difficult for courts to perform, and they may refuse to perform them because of a perceived lack of capacity or legitimacy. As I explain in more detail below, the fact that social rights have some aspects that are more easily assimilated to traditional rights enforcement, and other facets that would require courts to undertake radical tasks, is important. It means that in practice, courts are likely to enforce social rights either by issuing negative injunctions or by giving individualized remedies to individual plaintiffs. Such methods of enforcement will be least likely to get courts into serious trouble.

B. The Debate on Enforcement and the South African Obsession

Most of the more recent work in the field has focused on the specific question of how social rights should be enforced rather than the older question of whether they should be included in constitutional texts in the first place. Some critics of social rights argued that if social rights were actually put into constitutional texts, courts would be unlikely to actually do anything with them.²⁴ Understanding their lack of democratic legitimacy and institutional capacity, courts would merely ignore these rights. Empirical experience has shown this observation to be false--courts have found a variety of approaches to enforce these rights. However, scholars have emphasized a clear tension between the desire to enforce socio-economic rights once they find their way into the text and the strains on both capacity and democratic legitimacy that courts may feel if they aggressively enforce them.

The theoretical debate, however, has focused almost entirely on a single country (and largely on a single case). In the famous Grootboom decision, the South African Constitutional Court held that the political branches in South Africa had violated the constitution by failing to develop a housing plan that would meet the immediate needs of the poorest people most in need of assistance, like the plaintiff.²⁵ But the Court refused to order an individualized remedy for the plaintiff, such as an order that the state provide her with housing--the constitution did not create a right to housing “immediately upon demand.”²⁶ Nor did the Court give the details of such a plan and require the political branches to adopt it or try to implement the plan itself. Instead, the Court merely stated that the political branches had the obligation to “devise and implement a coherent, coordinated programme” and that a “reasonable” part of the total housing budget had to be reserved for those in desperate, immediate need of housing.²⁷ The underlying concerns of the Court appeared to be the ones of the critics of social rights--the Court was concerned that it would lack the legitimacy and capacity to issue a stronger order.²⁸

A prominent group of American constitutionalists lauded the decision as a reconciliation of two imperatives previously thought mutually exclusive by most--the enforcement of the detailed social rights now found in most constitutions and the assurance that courts do not overstep their bounds of democratic legitimacy and capacity. Thus, Mark Tushnet wrote that the Court's work constituted a new kind of judicial review, “weak form review,” that allowed courts to judicially enforce these rights without involving them in complex public policy decisions or letting them run roughshod over the legislature.²⁹ In other words, the Court gave the right to housing some judicially enforceable content, but at the same time, gave “legislatures an extremely broad range of discretion about providing” the right.³⁰ In a similar vein, Cass Sunstein wrote that the Court had effectively “steer[ed] a middle course” between holding socio-economic rights non-justiciable and holding them to “create an absolute duty” to provide housing or food or health care for everyone who needs it.³¹ Instead, the Court had enforced the right to “promot[e] a certain kind of deliberation . . . as a result of directing political attention to interests that would otherwise be disregarded in ordinary political life.”³²

More recent work has critiqued the positions of Sunstein and Tushnet. A large group of both South African and American scholars has argued that weak-form enforcement, as exemplified by Grootboom, did not work--the legislature did not produce the plan that the Court requested, and the case did virtually nothing to actually advance the right to housing.³³ Many of these academics have argued that Grootboom had more or less the right idea but needed to be ratcheted up: the remedy needed to be made a little less “weak” in order to be effective.³⁴

In a series of recent articles, Brian Ray notes that the South African Constitutional Court has abandoned the Grootboom approach for another tactic that Ray calls engagement.³⁵ The core of the engagement remedy is that the Court orders the state to negotiate with the plaintiffs so that a satisfactory agreement can hopefully be reached. … Ray argues that engagement, which the Court has also used in subsequent cases, is an alternative to Grootboom that also manages the tension between the need to enforce these rights and the capacity and legitimacy problems that courts feel when they enforce them. Engagement “falls somewhat short of the call by the Constitutional Courts' critics for full-fledged judicial interpretation and enforcement, but the same features that make engagement something less than strong court enforcement also enhance its legitimacy.”³⁹ As with Grootboom, however, there are real questions about the general effectiveness of the engagement remedy, at least as it is currently used by the South African Constitutional Court--engagement has failed in several subsequent cases.⁴⁰

C. The Reality of Social Rights Enforcement

Critics of social rights argued that courts would probably respond to the constitutionalization of social rights by declining to enforce those rights. Tushnet, Sunstein, and Ray argued that courts should relieve the tension between enforcement of social rights and democracy/capacity issues by adopting the “weak form” or dialogical review used in Grootboom or an engagement remedy like the one used in City of Johannesburg.⁴¹ But neither approach seems to function well or to describe accurately the majority of social rights enforcement occurring around the world.⁴² The critics were wrong to suggest that social rights enforcement would not occur; in reality, courts have found a variety of ways to give content to these rights. But the South African solutions seem deeply bound up with the political situation and legal culture of that country and have not been used anywhere else.⁴³

In reality, courts have found other ways to manage the tension between the enforcement of social rights and the capacity and legitimacy costs perceived to go along with that enforcement. Many courts appear to rely upon an individualized enforcement model--when an individual plaintiff comes to the court asking for provision of some particular medicine or treatment, they grant relief to that individual plaintiff.⁴⁴ This model relieves the tension noted above by providing relief to only a single plaintiff, thus avoiding complex management issues and making it appear that the court is not intervening massively in public policy. Even though the aggregate affect of these decisions on the public budget can be very large, the individual decisions appear to be familiar court-like work; the court is simply deciding whether one plaintiff is entitled to a remedy against one defendant. A second way in which courts manage the tension is by issuing negative injunctions striking down a law and maintaining the status quo, rather than issuing positive orders forcing the state to provide a service.⁴⁵ Again, by doing this, the courts are assimilating social rights enforcement into the enforcement of traditional first-generation rights--it is issuing a merely negative remedy for the right.

Both of these remedies allow courts to carry out social rights enforcement relatively securely and without worrying that they will be seen as overreaching beyond the traditional tasks of courts. But there is a significant cost--both tools are heavily tilted toward middle class and upper income groups rather than poor plaintiffs. In other words, in much of the world social rights enforcement is vibrant, but accrues to the benefit of higher class groups rather than those social groups most in need.⁴⁶ With individualized enforcement, this occurs because individual middle class rather than poor plaintiffs are more likely to know their rights and to be able to navigate the expense and intricacies of the legal system.⁴⁷ In the negative injunction cases, it occurs because the state usually tries to cut middle class pension and health care benefits for civil servants and other middle class groups rather than those few services going to the very poor.⁴⁸ Put another way, the status quo gives the poor relatively little to protect through negative enforcement. To make matters worse, courts often manage the tension between social rights enforcement and democracy by engaging in judicial populism--issuing decisions that are calculated to raise the ire of the political branches but to gain strong support from the middle class groups. For example, courts often strike down austerity measures that limit middle class social benefits precisely because they will have the support of the median voter when they do so, which may insulate them from retaliation.⁴⁹

All of this suggests that the conventional literature misunderstands both the general nature of social rights enforcement and the tradeoff faced by courts. Most of the literature on social rights, whether in favor of or against enforcement, assumes that it is a counter-majoritarian exercise, and that the beneficiaries of its enforcement will be marginalized groups.⁵⁰ This does not appear to be true--social rights enforcement is essentially majoritarian in many cases, and the beneficiaries are middle and upper class groups rather than the marginalized. Moreover, the real tradeoff faced by courts when choosing remedies is more complex than simply a tradeoff between effective enforcement and legitimacy or capacity issues. Instead, there are three issues--the legitimacy/capacity cost to the court, the effectiveness of the intervention, and the question of which group benefits from the intervention.

Table 1: The Effects of Socio-Economic Rights Remedies

Table 1 summarizes these tradeoffs. Individualized enforcement may have a low legitimacy cost and does not strain the capacity of the court, but it primarily benefits upper income groups. Further, the evidence indicates that it does little to improve the performance of the bureaucracy in providing social services, and thus it may be relatively ineffective as well. Negative injunctions are effective (at least at maintaining the status quo) and may have only moderate capacity and legitimacy costs to the court (depending on their macroeconomic effect), but they again benefit primarily upper income groups. Weak-form enforcement or engagement appears to be targeted at lower income groups and to have low legitimacy and capacity costs for the court, but it also appears to be ineffective.⁵¹ A fourth approach, structural enforcement, is familiar from U.S. public law⁵² and occurs when a court issues broad orders aimed at reforming institutional practice over a long period of time.⁵³ This appears to hold some promise at targeting relief towards lower income groups, and may be able to do so effectively in some circumstances, but it obviously involves the court deeply in polycentric decisions and thus may put a significant strain on the legitimacy and capacity of the court….

III. Evidence from Other Countries

In this section I present evidence from other countries to support both hypotheses. In particular, I emphasize that remedies, both individualized enforcement and the negative injunction approach, appear to have a pronounced tilt towards upper-income groups. The individualized enforcement approach seems to do little to improve bureaucratic performance, while the negative injunction approach tends to involve the court in serious macroeconomic messes. Finally, a structural injunction approach, although relatively rare in comparative law, appears to have some promise both in targeting lower-income groups and in effecting positive changes in the bureaucracy, at least in certain political contexts.

A. Individualized Enforcement

The comparative evidence described below strongly supports the inferences, drawn from the Colombian data, that individualized enforcement of social rights tends to disproportionately benefit middle and upper class groups, and that its effect on bureaucratic effectiveness is ambiguous at best. Furthermore, individualized enforcement, especially of the rights to health and social security, appears to be very common in comparative constitutional law.²¹⁶ This is likely because individualized enforcement appears to be “court-like”: it involves courts in one-on-one disputes without seeming to involve them in complex policy disputes that are beyond their competence. The analysis here will focus on Brazil, which is a case that has been studied relatively extensively….

Case law appears to pay little attention either to resource limitations or to the economic position of the petitioner in these individual cases. For example, in a seminal 1997 case, a petitioner suffering from a rare degenerative disease requested stem-cell treatment in an American clinic; the treatment would cost $63,806.²²¹ The state raised the argument of resource limitations, but the Brazilian Federal Supreme Tribunal upheld the claim, stating, “[b]etween the protection of the inviolable rights to life and health . . . and the upholding . . . of a financial and secondary interest of the State, . . . ethical-juridical reasons compel the judge to only one possible solution: that which furthers the respect of life and human health.”²²² As Octavio Luiz Motta Ferraz notes, the basic rule of the Brazilian judiciary in these individualized cases is that “the right of the individual must always prevail, irrespective of its costs.”²²³ As in Colombia, the individual cases are detached from their systematic context.

In contrast, few collective claims have been filed, and these have usually been denied. Hoffmann and Bentes posit that the Brazilian judiciary is still steeped in a civil law tradition and thus unwilling to take on the obvious policymaking role implied by aggregate litigation.²²⁴ That is, they decide individualized claims from a “purely individual civil rights perspective” without giving much thought to economic or social impact; thus they tend to grant these claims.²²⁵ The arguments flip for collective cases, where arguments about social and economic impact are used to justify non-concession because judges are unwilling to appear to be making large-scale policy judgments.

There is no direct quantitative data on exactly who is filing these cases, but the accumulation of other relevant evidence means that “[i]t is not difficult to guess.”²²⁶ Work by Ferraz has found that the richest states file the bulk of the claims. His study found that of 4,343 suits filed between 2005 and 2009, the ten Brazilian states with the highest scores on the United Nations' Human Development Index generated ninety-three percent of the claims, while the ten states with the lowest scores generated only seven percent.²²⁷ Other research has shown that most claimants rely on private lawyers, and that most claims are for high cost medications and similar goods rather than basic health needs. Thus, the conclusion of observers has been that most of this litigation is being filed by middle-class petitioners rather than the poor.²²⁸ This is not surprising, given that Brazilian legal devices, unlike the Colombian tutela, are relatively complex and expensive, and, as Hoffmann and Bentes find, that the poor have a “general lack of rights consciousness and trust in the judiciary.”²²⁹

Finally, Hoffmann and Bentes find little evidence that this individualized jurisprudence has provoked positive effects on the executive bureaucracy. They find that the major policy decision to include HIV drugs on the lists of allowable medicines was unrelated to litigation, although they do find that some other medicines have been included on the list because of a critical mass of litigation.²³⁰ Still, they find that, as in Colombia, most of the real problems in the Brazilian health bureaucracy are problems of “implementation” rather than design; even when a drug or treatment is included on the list, patients often do not receive it without legal action.²³¹ In this sense, as in Colombia, the Brazilian courts have become a partial replacement for the bureaucracy rather than helping to improve bureaucratic action.

B. Negative Injunctions

Another very common tool has been enforcement of social rights by negative injunction, which prevents the government from withdrawing some existing benefit. This is closely related to the concept, found in international law and already discussed, of non-retrogression: reductions of existing social benefits will be subject to heightened scrutiny in order to determine their appropriateness.²³² The Colombian Court used this principle in its 2000 public sector salary case, in which it ordered all public sector employees to receive increases in salary at least equal to the rate of inflation. The negative injunction is very popular in comparative law as a means to enforce social rights, most likely because it also appears relatively court-like: the judiciary is not involved in making complex budgetary allocations or otherwise constructing policy, but instead merely prevents the state from putting some new policy into effect. In other words, enforcement of social rights by negative injunction makes these rights look more or less like other kinds of rights.

This kind of social rights enforcement is likely to have a strong tilt in favor of more affluent groups. Middle and upper class groups tend to have pensions, decent health care, and other subsidies. These benefits can be attractive targets for governments that urgently need to cut budget deficits and which may be under international (IMF, World Bank, etc.) pressure to do so. In contrast, the very poor do not have many benefits for the government to take away; therefore, states are less likely to cut these benefits during recessions and periods of structural adjustment. One of the most common types of social rights enforcement “for the benefit” of the poor illustrates this point; the poor may file injunctions against evictions from slums built on lands that they do not own, or from living on the street, in both India and South Africa.²³³ The jurisprudential logic in both countries is that there is a constitutional right to housing, and while the positive aspect of this right (the building of decent housing for all) cannot be realized immediately, the courts will at least enforce the negative aspect of the right by making it more difficult to evict poor tenants from their existing homes in ramshackle slums or on the streets. These cases may do something for the poor, but not much.

The more typical negative injunction case in comparative constitutional law benefits the middle class. A good example is the Hungarian Constitutional Court's activism on social benefits in the mid-1990s, when the Court struck down a series of important government measures. The Hungarian government, in the midst of a severe economic crisis and under pressure from international organizations like the IMF, attempted to cut many benefits from the social benefits system (pension, child supports, sick leave, etc.) and to move from a universal system toward a need-based system.²³⁴ The jurisprudential basis for these judgments was that people with existing benefits had property-like rights to those entitlements that the government could not take away lightly.²³⁵ While there is a debate about the appropriateness of these decisions,²³⁶ there is no doubt that these judgments benefited mainly upper-income groups, and that Sajó is correct in calling the Hungarian social rights “middle class entitlements.”²³⁷ Like the Colombian UPAC and salary cases, these decisions were also very popular with the public. A poll taken just after the decisions found that eighty-nine percent of the population had heard of them, and that eighty-four percent of those who voted for the ruling parties and ninety percent of those who did not favored the decisions.^238…

A final point is that this kind of jurisprudence, although seemingly court-like, tends to get judiciaries in big trouble. This happens because these cases tend to impinge on core macroeconomic policy decisions at precisely the moment in which governments are experiencing budgetary stress and need to undertake structural adjustments. These are often “populist” decisions and may be popular with the public, as the Hungarian and Colombian decisions show.²⁴² But they also anger presidents and legislatures, who may seek to attack or overhaul judiciaries as a result….

C. Structural Injunctions

Structural injunction-like devices have been rare in comparative constitutional law. Although various scholars have pointed out their theoretical utility in resolving difficult social rights problems, they remain for the most part the pipe dream of academics in other countries, a remedy that exists in journal articles but is almost never seen in reality.²⁴⁶ This has been especially true in South Africa, where the Court has aggressively taken on the challenge of defining constitutional rights but has been willing to give only very limited remedies for their violation.²⁴⁷ And, as the Colombian example shows, structural remedies are expensive, time-consuming, demand a tremendous amount of legal and political skill from the judiciary, and only appear to work well in certain political contexts.²⁴⁸ On the other hand, they have the potential to correct some of the biases seen in the other devices, and they may be especially promising for targeting lower income groups. The limited comparative experience that exists for these devices supports these hypotheses….

IV. Implications

The analysis presented above has significant policy implications for how domestic courts should enforce social rights and for how international bodies and organizations should think about enforcement of these rights and principles. The normative assumption on which I base this section is that it is desirable to improve targeting of enforcement towards lower-income social groups. (In my conclusion, I return to the idea that a relatively middle-class-based jurisprudence on social rights is probably inevitable, and I discuss the implications of that fact). I emphasize three points in this section. First, international policymakers, particularly those on the Committee on Economic, Social, and Cultural Rights, should emphasize and better define the minimum core and should deemphasize potentially dangerous concepts like non-retrogression. Second, the international dialogue between constitutionalists in different countries should emphasize remedies rather than rights alone, and a consideration of the U.S. experience (which amply demonstrates both the possibilities and limits of structural reform litigation) may be useful for these ends. Third, policymakers designing or reforming a judiciary should consider ways not only to preserve judicial independence while maintaining a link to the people, but also to rein in populist behavior by the judiciary….

V. Conclusion: Coming to Grips With a Middle-Class Social Rights Jurisprudence

Social rights enforcement has been vibrant in a number of countries. However, the patterns of enforcement show disturbing relationships. One claim in this piece is that there is a perverse relationship between choice of remedy, the likely set of beneficiaries, and the perceived (although perhaps not actual) strains on a court's capacity and democratic legitimacy. Empirically, court are most likely to enforce social rights by negative means (such as striking down a law) or via individualized rights enforcement, since these tools are closest to the tools courts use for everyday judicial review. But both are bad ways to enforce social rights claims--they have perverse distributive effects and do not appear to do anything to improve the performance of the bureaucracy. Even structural remedies are difficult to accomplish successfully. The Colombian and Indian examples show that they demand a lot of the court's resources and do not work in certain political contexts. But, at least sometimes, they can work. They should be part of the judicial toolkit, and scholars should start building theories for when and why these kinds of remedies are effective.

The broader point is that we need to reevaluate what social rights do; we must re-envision them as a largely middle-class phenomenon. As such, their enforcement is mostly majoritarian. While U.S. constitutional theory coined the phrase “countermajoritarian difficulty,” American scholars have long noted that this is an oversimplified view of what the Supreme Court does and that it essentially follows majority will in many circumstances.²⁹⁶ But the same vision has not penetrated much of comparative scholarship. Making a full evaluation of the fact that courts are majoritarian in many (perhaps most) circumstances is a task for another article. Middle-class centric judicial enforcement may not be an entirely bad thing, given the dearth of legitimacy that most state institutions have in developing countries and the low quality of their bureaucracies. This suggests that the core question attending this sort of enforcement is the rebound effect that judicial action has on bureaucrats, politicians, and civil society groups--does it strengthen civil society and improve bureaucratic performance? A big part of the answer, again, is likely to rely on remedial innovation: courts may need to intrude more on democratic institutions in order to improve them.