Research note. Judicial review of executive policymaking in advanced democracies: beyond rights review

Research note by Eduardo Jordão and Susan Rose-Ackerman

This note summarizes our article “Judicial review of executive policymaking in advanced democracies: beyond rights review”, published in the Administrative Law Review, volume 66 (2014). The article considers the cases of the United States, Italy, France, and Canada.

Our point of departure is the claim that administrative law ought to further three distinct goals: (i) protecting rights that may be violated by administrative action (rights); (ii) guaranteeing that administrative action will be democratically responsive (democracy); and (iii) guaranteeing that administrative action will be well informed and technically sound (competence). However, efforts to foster any one of these goals can end up compromising the others. In particular, when courts review administrative decisions, the challenge of balancing these goals is very clear. The more the law tries to protect the rights of the citizens by giving courts the power to review the substance of administrative decisions, the more likely it will be that the other goals are jeopardized. Judicial review of regulatory agency choices, for example, may end up with important regulatory decisions being taken by the courts– institutions that have low democratic credentials and very limited expertise in technical issues.

We ask: What is the best way for courts to balance the competing goals of administrative law when they review executive policymaking in advanced democracies? Courts tend to gravitate toward the protection of rights. We argue, however, that courts can also play a constructive role in furthering both democracy and competence. But what is the best way to balance these goals given that administrative decisions are, at least to some extent, based on technical and political considerations?

We argue that courts: (i) should subject the substance of agencies’ decisions only to a weak reasonableness test; and (ii) should concentrate on the administrative process, notably by enforcing a widespread duty to give reasons and by assuring generous rights of participation. Our argument has two parts. We begin in Part I with substantive review. For decisions involving policy or technical matters, we argue in favor of weak, reasonableness review as an appropriate way to balance competing goals. In addition to allowing oversight that protects rights, this standard of review gives some leeway to agencies to accomplish their allocated missions. Stringent judicial review of agency actions could lead the courts to usurp the policymaking competence of the agencies. As a rule, agencies are both more technically competent and more democratically accountable than courts. Deferential review of substance prevents courts from making decisions beyond their own areas of competence.

However, even if the judiciary does not provide aggressive judicial review of the substance of regulatory policy, that does not imply that courts should simply decline to rule on challenges to regulatory actions. Instead, they can concentrate on process and check to be sure that policy is made in an accountable, transparent, and responsive manner that draws on necessary expertise. Procedural review can be a partial substitute for substantive review and can promote both democratic legitimacy and regulatory competence. Through procedural review, courts can give sufficient leeway to the technical and political choices of administrative bodies and at the same time assure that the decisions are both transparent and well-informed. We focus on two different procedural aspects of administrative decision-making: the duty to give reasons and rights of participation.

The duty to give reasons is traditionally seen as a means to permit future review of the legality of administrative action. The idea is that only through the analysis of the reasons given by the agencies would it be possible to know whether agency action is consistent both with the legal text and with individual rights. In addition to the protection of individual rights, we argue that compliance with the duty to give reasons also serves broader goals of transparency and political legitimacy. The obligation to explain decisions is a consequence of the public administration’s policymaking activities. If an agency makes substantive choices that affect the use of public resources and the behavior of regulated entities, citizens should have access to the reasons for those choices.

Likewise, the right of participation usually translates into a right to be heard, which is linked to the protection of individual rights. For us, however, participation in rulemaking processes has a broader justification. It is a source of democratic legitimacy that allows citizens to influence the adoption of policies that reflect their beliefs and interests. Broad participation rights require administrators to accept input and data from those concerned with agency action. Agencies can lower the cost of outside participation by their own efforts of outreach. Such actions can help mitigate the information asymmetry that arises if the regulated industry dominates the consultation process. In ensuring a better balance, rights of participation can help make administrative decisions both more democratic and better informed—enhancing competence.

We illustrate our claim through a critical review of the law in Canada, the United States, Italy, and France.

In our opinion, the most interesting case is that of Canada. There, judicial review of administrative adjudication has evolved from very intrusive to quite deferential and nuanced. Until 1979, Canadian courts reviewed administrative adjudications very aggressively. The Supreme Court of Canada then ended the era of intrusive review, criticizing the excesses and creating a deferential standard of review for “patent unreasonableness.” This orientation persists to the present. Courts seldom apply the correctness standard to strike down administrative actions, and the few cases where they do are ones that raise general questions of law, human rights, constitutional issues, or jurisdictional concerns. Of particular interest to our inquiry is the Court’s contextual approach to determining the standard of review. Canadian administrative law has developed a workable two-stage structure. First, the Court determines the standard of review that it will apply. Second, it decides the case using that standard. The analytical framework used for the first stage was first called the “pragmatic and functional” approach. Since the Supreme Court decision in Dunsmuir, the reformed test is called “standard of review analysis.” This analytical framework is relevant especially because, under the prongs of the test, courts weigh substantial and institutional aspects of the agency’s decision. The framework requires the courts to assess which institution is better constituted to decide the issue under review. This pragmatic and functional approach allows the courts to strike a balance among the three-fold aspects of legitimacy: rights, democratic responsiveness, and competence. If competence or democratic responsiveness is critical to the decision under review, courts will usually defer to the agencies’ choices because of their higher political accountability or technical expertise. If, however, the protection of rights is particularly salient, the courts will intervene.

However, although Canadian Law presents a very interesting case of weak substantive review, it is not as good a model for procedural review. Procedural requirements are uncommon in Canadian federal legislation that covers rulemaking and adjudication. Lacking legal texts, the courts approach each type of agency action differently. For rulemaking, the courts totally refuse to review procedural aspects. The leading case here is Inuit Tapirisat. For adjudication, the courts use a contextualized test that adjusts procedural rights to the circumstances of the case. The leading case is Nicholson. There is no federal administrative procedure act that establishes the procedure to be followed for administrative adjudications. However, courts have long enforced a “duty of procedural fairness.” Such a duty leads to procedural rights only when an individual right might have been violated. It is not a means to further competence or responsiveness. A recent decision of the Canadian Supreme Court, nevertheless, may be taking a step to extend the duty to give reasons. In Dunsmuir, the Canadian Supreme Court elaborated on the content of its new standard of reasonableness that applies to the substantive review of administrative decisions. The Court requires that the decision be “justif[ied], transparen[t] and intelligib[le]”. Under this new regime, reasonableness is more related to the quality of the agency’s reasoning than to the outcome, that is, it is more linked to process than to substance. In the context of the new reasonableness test, justifying a decision in a transparent and intelligible way is linked to legal validity, irrespective of any violation of individual rights.

The United States also presents an interesting case. Contrary to the Canadian case, however, review of process is its distinctive feature In American Law, officially, judicial review of substance is very deferential – but in practice, that is not always the case. In reviewing agencies’ statutory constructions, courts claim to apply deferential doctrines, such as Chevron or Skidmore, but many decisions are based on quite strong judicial readings of statutory language. In reviewing other aspects of administrative action, the usual standard is the Administrative Procedure Act [APA]’s “arbitrary and capricious” baseline, which is also usually understood as a very lenient review. The practice, however, shows a much more mixed picture. The APA standard of arbitrary and capricious has been depicted by the Supreme Court as “searching and careful” (in Overton Park), but also as allowing only an “extremely narrow and deferential review” (in US Postal Serv. v. Gregory). The actual impact of the Chevron doctrine is much debated. It is among the most heavily cited cases of all time, and some empirical studies show that the percentage of administrative decisions that the federal courts affirmed rose after Chevron. Others, however, have found that the Supreme Court continues to impose its own interpretations of the law on agencies, often without even citing Chevron. Even when courts nominally apply Chevron, the deferential orientation is not clear. Under step one of Chevron, courts are supposed to assess whether “Congress has directly spoken to the precise question at issue.” In practice, courts use the so-called “traditional tools of statutory construction” to decide this issue, and sometimes they reach out to find and resolve ambiguity against agency interpretations. Courts frequently conclude that a given text is not ambiguous, but only after a rather long analysis of the purpose of the statute or its legislative history. In other words, sometimes courts seem to use the traditional tools of construction in step one to regain the powers of statutory construction that they lost with Chevron.

When administrative procedures are under challenge, though, American law presents a strong case. In the United States, both participation rights and the duty to give reasons are central to the notice-and-comment provisions of the APA. The focus is on political legitimacy and competence, not the protection of individual rights. Reason-giving is relevant to judicial review of agencies’ statutory constructions under step two of Chevron. In Chevron, the Court stressed that the Environmental Protection Agency had carefully explained why it adopted a specific notion of stationary sources of air pollution and hence the court deferred to the agency’s permissible interpretation of the legal text. The requirement of permissibility has a procedural character. Under the weaker standards of deference of Skidmore and Mead, reason giving is also crucial. A given interpretation of the law will probably be deemed persuasive under Skidmore if it is well enough explained to convince the court. A given interpretation will be given deference under Mead it it was adopted in well-specified and more established procedures, such as notice-and-comment rulemaking. There are examples of instances where substance and process intertwine in a way that is consistent with our own views of the proper role of the courts. Courts will apply a deferential review of the substance so long as the agency has carried out a process that invites broad participation, canvasses the relevant technical material, and ends up with a reasoned policy judgment.

The third case we presented was that of Italy. In Italian case law Substantive review has a very curious and inconsistent history. It moved from intensive review to deference and then back again. This back and forth illustrates the tensions of reviewing the decisions of the public administration in recent times. Traditionally, Italian courts work within a binary framework, giving limited review to discretionary decisions and stronger review to non-discretionary decisions. Eventually, going against this tradition, courts developed the concept of “technical discretion” (discrezionalità tecnica), for instances where the administrative authorities interpret ambiguous or debatable technical legislative terms. This was criticized in the Italian legal literature and courts eventually decided that technical discretion was not really discretion and should thus be subject to stringent review. This was in 1999. Two years later, facing the need to review complex decisions of the antitrust agency and lacking the resources and the technical knowledge to do so, the Consiglio di Stato came up with another concept of “complex technical assessments”, only to subject them to “weak review”. This again was much criticized, on the grounds that it would hinder “effective review”. Then came a new change of direction in 2004. Arguing that its reference to “weak review” had been misinterpreted, the Consiglio di Stato abandoned the concept and all the comparative institutional analysis that it had used to create it. As we can see, during two different periods, judges challenged the traditional binary orientation. Before 1999, using the notion of “technical discretion” courts applied limited review to cases where the terms interpreted were “technical” and “debatable.” Between 2002 and 2004, courts deferred to “complex technical assessments”, claiming they involved both complex technical expertise and policymaking that was better done by the agencies. These two periods brought Italian law closer to American and Canadian practice. For a short period, Italian courts explicitly admitted that the construction of ambiguous legislative terms can give rise to policymaking and, hence, to limited review. The courts did not properly justify the abandonment of this deferential position after 2004. The Consiglio di Stato introduced its new approach by claiming that it was just an explanation of its previous case law—in fact, it represented a complete change of direction. The Consiglio di Stato presented no justification for its new stringent review. The judges provided no theoretical explanation to why it was now possible to review aspects of the decision that were beforehand deemed to include policymaking. They just started to ignore the institutional difficulties of review that they had themselves brought up some years before. The hesitations and inconsistencies that still persist in Italian law illustrate the difficulties of applying stringent review to the complex decisions of specialized administrative bodies. The judicial movement towards more stringent review was aimed at “more effective” protection for the rights of citizens. However, such review can lead judges to interfere with the administration’s democratic legitimacy and technical competence.

The procedural rules of administrative action in Italian Law are also interesting because they go in the exact same direction that we claim to be desirable. In Italian law, courts have evolved from a traditional orientation that linked procedural safeguards to the protection of rights (the so-called funzione garantistica della motivazione). They now acknowledge that in some cases these mechanisms can enhance both the democratic accountability and the efficiency of public administration. The traditional orientation is clear in the 1990 Italian Administrative Procedure Act, which codified judicial practice and introduced a duty to give reasons. The Act explicitly exempts “normative acts and those of general application”, just because they usually do not cause direct violations of individual rights – reasons would therefore be unnecessary. In recent years, courts and the legal literature have begun to understand procedural mechanisms in a more nuanced manner. In addition to the funzione garantistica, they now stress that both the duty to give reasons and rights of participation enhance administrative accountability and competence. This trend is clearer and more effective in the review of independent agencies, where courts have held that procedural guarantees compensate for the agencies’ lack of legitimacy and help them to produce better regulations. Specific statutes require reason-giving for the normative and generic acts issued by independent authorities in areas such as energy and gas, telecommunications, and financial regulation. But administrative courts have then decided that the duty to give reasons is widespread for regulatory decisions – even in the absence of a specific statutory provision. The judges argue that the combination of normative powers and independence from the central government makes agency policymaking democratically problematic. Rather than condemn the practice, they, instead, require that agencies publicly justify their actions. Reason-giving is a partial compensation for the so-called deficit of legitimacy that affects independent agencies. In the domain of regulatory agencies, both the duty to give reasons and the rights of participation are further reinforced by the obligation to perform a regulatory impact assessment before promulgating a regulation. The adoption of a regulatory impact assessment obligation is linked to the need to promote good regulation. Regulatory quality is linked to participation. Hence, procedural rules are enforced in Italian Law in the exact same way that we claim to be desirable: they are tools to foster the goals of competence and democratic responsiveness, and not only tools to protect rights.

The French one is farthest from satisfying our own view of effective judicial review of administrative policymaking actions. France has had a consistent pattern of non-deferential review of substance over its recent legal history. Both structural and substantive factors contribute to this “judicialization of the administration.” Access to the administrative courts is quite easy. Rules governing standing, third-party interventions, and jurisdiction are generous. The liberal standing rules remain unscathed even after many reforms aimed at mitigating the overload of administrative cases. Judicial review has become progressively less deferential (under the standard of contrôle normal), and some authors talk about the decline, or the death, of the deferential standard of review (contrôle restreint). This system, however, is not well-adapted to the review of the actions of the modern regulatory welfare state. Independent regulatory agencies fit awkwardly into the French legal structure. The French courts have not created any new concepts in their review of technical, complex decisions taken by regulatory agencies and ministries. They apply the same concepts that they developed in the review of non- specialized administrative actions. Theoretically, French administrative courts apply a deferential standard of review to highly technical or politically sensitive cases. In practice, however, they are clearly less prone to find such instances when compared to their Canadian and American counterparts.

A good example of this tendency to overlook technical complexities is the case law on mergers. The Conseil d’État has reviewed for correctness the identification of the relevant market and the evaluation of anticompetitive effects, assessed the very existence of a merger, and established the criteria under which making an “exception for a failing firm” could be accepted. Likewise, the political component of many regulatory decisions is not officially recognized in courts. This is particularly clear in the domain of statutory construction. Canadian and American courts defer systematically to agencies’ interpretation of ambiguous terms in legislation. French courts, however, usually claim legitimacy to interpret these ambiguous terms. They view such concepts as “legal” because they are in the statutory text; therefore, the courts can interpret them. French courts define the realm of law broadly, allowing for far-reaching review. Thus, in cases where Canadian and American courts would acknowledge that agencies are engaged in policymaking and should be left alone, French courts tend to view agencies as making legal decisions that are therefore reviewable. In general, French jurists are comparatively less likely to accept claims that administrative authorities are best suited to deal with technical issues. As a consequence, both democratic legitimacy and deference to expertise play a comparatively less pronounced role in French judicial review of administrative action.

When review does occur, the Conseil d’État’s impact on government and agency policymaking is problematic for several reasons. First, the Conseil d’État’s decisions have traditionally been very concise. Conclusions are often left unexplained beyond references to legal texts. Courts make crucial decisions concerning the regulation of important areas like telecommunications or energy in a few short paragraphs. Second, in most cases the Conseil d’État does not discuss opposing arguments or explain why its reasons are better than the contrary reasons. The lack of dissent compounds this problem along with the secrecy of the deliberations. Third, if the Conseil d’État characterizes an issue as merely a legal question, this tends to hide the fact that there is room for alternative interpretations. If the legislation is ambiguous, the courts tend to choose an option and claim that it is the only legally possible solution. Political choices are presented as legal impositions. Fourth, open-ended “legal” concepts like “proportionality” and “general interest” empower courts to make their own balance of interests when they decide cases that are essentially political in nature.

Procedural review in France is also deceptive, if one accepts our normative claims. Like Canada, France has no general, legally enforceable procedures for the promulgation of secondary legislation. These procedures are not subject to any generic participation or reason-giving requirements. The only procedural mandate is the requirement that the Conseil d’État review draft decrees and ordinances. In France the courts have been reluctant to adopt a general duty to give reasons, adopting the requirement only in a certain number of decisions, mostly connected to the violation of rights. The Conseil Constitutionnel has held that legally enforceable participation rights exist for environmental issues, but both the Conseil d’État and the Conseil Constitutionnel have consistently refused to recognize a general obligation to provide reasons for administrative decisions in the absence of a statutory or constitutional provision. If applicable, the obligation is tied to the protection of rights and especially to adjudications where the outcome for the individual is particularly burdensome. It is not connected to more general ideas of transparency and political legitimacy. The doctrines of the “substitution of reasons” and “overabundant reasons” add to the problem. Although they may help speed up the administrative process, those doctrines create further incentives to limit the transparency of executive policymaking.

Putting together the limits of substantive review and the promise of procedural review of policymaking, and considering the goal of balancing the three aspects of state legitimacy through the courts, France appears to have the worst combination of our four cases. It has very aggressive review of substance and weak review of process. Of course, this does not mean that the French approach to judicial review is problematic overall. However, it privileges the protection of rights to the detriment of other goals, such as administrative efficiency, technical competence, and political accountability. After France, comes Italy. Although its courts also engage in aggressive review of substance, it offers a greater review of process, at least in the context of independent agencies. Canada has found a good balance for review of substance but lacks review of rulemaking procedures, except for the cases of formal regulation. The United States has a rather inconsistent record on the review of substance, but its relatively deferential judicial practice places it just behind Canada. On judicial review of the rulemaking process it dominates the other cases in its explicit concern for the democratic legitimacy of delegated policymaking. However, this favorable view is conditioned by the time consuming nature of the process, which delays the implementation of important rules. The decision to adopt procedural requirements that further state legitimacy is just a first step. A further discussion concerns the actual procedures required. Who should be given the opportunity to participate? If participation is restricted to interested parties, how should they be defined? How can we ensure that some (better organized, better financed) groups will not dominate the consultation process? How can participation be designed to avoid excessive cost and delay? The answer to these and other questions will help to balance democratic responsiveness with the other conflicting goals of administrative law.

  • Jordao, Eduardo Ferreira and Rose-Ackerman, Susan, Judicial Review of Executive Policymaking in Advanced Democracies: Beyond Rights Review (April 3, 2014). Administrative Law Review, Vol. 66, 2014; Yale Law School, Public Law Research Paper No. 499; Yale Law & Economics Research Paper No. 497. Available at SSRN:

(Eduardo Jordão and Susan Rose-Ackerman)

{F6EE6C1D-B3D6-4986-A44C-519047584308}_jordaoEduardo Ferreira Jordao, FGV Direito Rio
Doutor em Direito Público Pela Université Paris I Panthéon-sorbonne e Pela Sapienza – Università Di Roma, em Co-tutela. Ll.m Pela London School Of Economics And Political Science. Mestre em Direito Econômico Pela Usp. Bacharel em Direito Pela Ufba. Pesquisador Visitante na Yale Law School, nos Estados Unidos, e Pesquisador Bolsista nos Institutos Max-planck de Heidelberg e de Hamburgo, na Alemanha. Pesquisador do Centro de Justiça e Sociedade (cjus) e Coordenador de Relações Institucionais da Fgv Direito Rio.
rose-ackerman_susanSusan Rose-Ackerman, Yale Law School
Susan Rose-Ackerman is the Henry R. Luce Professor of Jurisprudence (Law and Political Science) with joint appointments between Yale Law School and the Yale Department of Political Science. She has taught and written widely on corruption, law and development, administrative law, law and regulatory policy, the nonprofit sector, and federalism.