Autonomy by Decree: How Administrative Law Shapes Bureaucratic Autonomy in Four Administrative Traditions

Abstract

Administrative law is a not a frequent subject of research in public administration, but it can reveal a great deal about the functioning of the public sector and governance in general. The nature of administrative law is, we argue, closely linked with administrative traditions, and therefore administrative law is an especially apt focus for comparative analysis. This article discusses administrative law in four countries representing different administrative traditions. The perspective is that of the student of public governance, rather than that of the lawyer, with the principal concern here understanding bureaucratic autonomy within the administrative system.

Introduction

The study of bureaucratic politics has grown noticeably over the last decades. One of the more noteworthy developments in the field is the increasing attempt to identify causal evidence of bureaucratic behavior. As Bertelli and Riccucci (2022) argue, this strand of research often fails to connect causal evidence of behaviors to the prevailing politico-institutional order. Similar concerns have been raised more generally about the lack of “real-world” deduction and/or excessive use of particularism in case selection (Peters & Pierre, 2017), or more broadly that political science needs to be brought back to the study of the public administration (Peters et al., 2022). We agree that bureaucratic behavior is best explained when robustly anchored to the prevailing institutional order under study.

Another development in our field has been less frequent attention to administrative law and its dynamic impact on public organizations and bureaucratic behavior (cf. Bersch & Fukuyama, 2023). To put it briefly, administrative law sets the basic rules for interactions between the public bureaucracy, politicians, and citizens (Rose-Ackerman et al., 2017); it delineates the process of public policy (Rosenbloom, 2018); and it is highly differentiated in style and substance across national contexts (cf. Bell, 2019). In response, we investigate how administrative law can have direct and indirect consequences for contemporary bureaucratic politics. In doing so, we advance the argument that administrative traditions (Peters, 2021) manifest themselves in administrative law, as well as in organizational structures and administrator behavior. Law, then, creates very different spaces for maneuver for the individual civil servant or public organization as a whole, depending on the nature of that law. Thus, we believe that the comparative-historical framework of administrative traditions together with a legal analysis of administrative law can tell us a good deal about the important concept of bureaucratic autonomy—a key dimension of bureaucratic politics—in contemporary states.

This article is written from the perspective of public administration and political science rather than from the perspective of law per se. We try to demonstrate the utility of examining an aspect of administration that many public administration scholars may consider arcane. We believe that although administrative law is embedded in broader legal systems and their approaches to law (see Damaska, 1986), political and managerial elements sill help define the way in which administrative law functions and its impact on public service delivery. We are seeking to bring administrative law back into public administration, where it has largely been ignored for some years (but see Lienhard et al., 2022; Rosenbloom et al., 2017).

Moreover, we believe that legal analysis of administrative practice is quite relevant to the wider question of institutionalism in political science. The power of social norms has forced us to question how dominant formal institutions are in shaping behavior. At the same time, without formal law, we cannot speak of a firm institutionalization of the public administration, which is essential for talking about administrative or legal traditions. This article therefore attempts to demonstrate how law affects bureaucratic autonomy just as social norms and political power do.

Specifically, while an obvious link between administrative law and bureaucratic autonomy exists, we are interested in studying how administrative law shapes the level and form of autonomy for individual bureaucrats and for bureaucratic organizations. Administrative actors—individuals and organizations—may gain some autonomy by building their reputations or by engaging in bureaucratic politics, but administrative law establishes the parameters within which they must function. This investigation is constructed on a variable-basis with four different dimensions; it is illustrated with four distinct administrative traditions, the Anglo-American, Germanic, Napoleonic, and Scandinavian (cf. Peters, 2021). Bureaucratic autonomy is but one of any number of aspects of public administration that law heavily influences, and we use it to illustrate the general argument in line with Bersch and Fukuyama (2023) that administrative law is an important approach to comparative public administration.

Although some scholars (e.g., Raadschelders & Vigoda-Gadot, 2015) have criticized the concept of administrative traditions as merely being convenient stereotypes of systems, we believe there is substantial utility in the concept. Administrative tradition does point to the historical roots of contemporary behaviors; there is some path dependency in the manner in which institutions including bureaucracies function, and those roots manifest themselves in law. This concept also contributes to crossnational analysis. Although each country is inherently sui generis, it also is useful to understand the commonalities that both exist across some systems and that may be “legacies” of the past (Rugge, 2012). Here, we will point out that administrative law is at once a reflection of these legacies and an instrument through which they are enacted. Finally, there are clear empirical differences in the ways in which bureaucracies within different traditions behave, which can produce different outcomes (Peters, 2021, p. 37) as well as varying degrees of bureaucratic autonomy.

We do not contend that these groups are totally independent of one another, and we recognize that there has been some diffusion of ideas across administrative systems (Heyen, 1989; Sager et al., 2018). Empirically, however, the clear patterns of similarity in structures, law, and behaviors within traditions (see e.g., Ongaro, 2009) and differences across traditions justify traditions as an approach to classification and analysis.

This article is structured as follows. We first describe the concepts of bureaucratic autonomy, administrative traditions, and administrative law. This is followed by a discussion of the variables constituting the basis of our empirical application. Four empirical case studies of the US, Germany, France, and Sweden are then examined, followed by a concluding discussion. These four countries were chosen because the last three represent the prototypical examples of three dominant European traditions, while the US has the most fully codified rules for administrative practice in the Anglo-American tradition (Freedman, 1978).

Understanding Bureaucratic Autonomy

To derive a plausible framework for analysis, we must begin by defining bureaucratic autonomy. Maggetti (2007) offers a basic definition of it as “the ability to translate one’s own preferences into authoritative actions, without external constraints,” which although specified rather vaguely, sets a reasonable intuition into motion. By contrast, a significant share of related applied studies begins with a notion of autonomy from game theory that conceives bureaucrats essentially as utility-maximizers and antagonists of public power who prefer to “shirk” if moral hazard and knowledge asymmetries are high (e.g., Brehm & Gates, 1997; Calvert et al., 1989; Niskanen, 1975). Such “principal-agent” relationships fail, however, to explain all forms of autonomy and counteract a more value-neutral notion of the concept (Pierre & Peters, 2017).

Although autonomy from political actors is important, so too is autonomy from other actors that may attempt to influence or control an agency. The classic issue of “regulatory capture” is the extreme, but more subtle influences from organizations or individuals with which the organization interacts also can reduce its capacity for independent decisions. At the bottom of organizations, connections between street-level bureaucrats and their clients also can inhibit organizational autonomy during implementation (May & Winter, 2007). It is safe to say that bureaucratic autonomy needs wider attention than merely as a problem of suboptimal political control.

Autonomy is not a dichotomous concept; it varies over time and across policy areas and forms of governance. More related to our thesis, countries have established control mechanisms due to path dependent administrative traditions. For instance, the Napoleonic and Germanic traditions have relatively strong traditions to control bureaucracy by law, while the Anglo-American and Scandinavian models strive for more autonomy and managerialism (Peters, 2021; Pollitt & Bouckaert, 2004). The “environmental-institutional” context thus shapes the basic function of the bureaucracy in formalized and institutionalized practices (Maggetti & Verhoest, 2014, p. 248; Yesilkagit & van Thiel, 2008).

Moreover, informal autonomy is the outcome of imperfect formal control by the political level (Majone, 1997). This form of autonomy is directly associated with several specific formats of control, as will be discussed shortly. Yet, bounded rationality, changing external circumstances, and knowledge asymmetries between ministry and agency can generate autonomy that is not the result of deliberate design (Bach & Ruffing, 2013, p. 716).

As we focus on administrative autonomy we should consider the autonomy of bureaucratic organizations from their nominal political masters, the public, and from the regular courts. These three dimensions of autonomy will not necessarily vary together, and often may change in exactly opposite directions. As administrative agencies are given great latitude to make independent decisions without as much control from politicians, they may find themselves more subject to scrutiny by citizens and the media.

Administrative autonomy is in part based on political decisions, but it is also a function of the legal system and the manner in which the law defines the latitude of action for administrative actors. If we return to Maggetti’s definition of bureaucratic autonomy, administrative law represents the granting of the capacity to make decisions (largely through delegation of powers) as well a set of constraints on the exercise of autonomy. Each of the four traditions has mechanisms for dealing with the question of autonomy, as we point out below.

Lastly, the civil service consists of individuals, and the full staff is both heterogeneous and employed at several different levels of hierarchy and responsibility. In consequence, the degree of autonomy as well as autonomy from whom depends on such conditions. Moreover, and related to the first category, the degree to which legal environments are structured to hold individual civil servants accountable varies among administrative traditions.

Several other forms and sources of autonomy are important for understanding the concept’s full power, but space does not permit a detailed description (cf. Bawn, 1995; Callander, 2008; Gailmard & Patty, 2007; Maggetti & Verhoest, 2014; Verhoest et al., 2004). Most important for this article, however, is that autonomy is related to administrative traditions. The institutionalization of autonomy results in path dependent structured behavior. A repeated and “good” bureaucratic activity creates reputation and therefore institutionalizes autonomy (Carpenter & Krause, 2012), although bureaucrats who work in autonomous organizations can be argued to lack autonomy, as their behavior is largely structured by the specific organization’s institutionalized norms.

Administrative Law as a Basic Mechanism for Governing

One important question in thinking about administrative law in the context of public administration and political science is the extent to which that law, and its execution, is dominated by lawyers or by administrators. That question is related to an even larger question: To what extent does administrative law facilitate or control the autonomy of administration? In almost all instances, the fundamental intention of administrative law in democratic political systems is to control administration and to hold it more accountable.

Nonetheless, there are instances in which the evolution of administrative law has helped to make the bureaucracy more autonomous and has validated the autonomy that it has sought through political means–bureaucratic politics. This growth of autonomy has been seen as the “administrative state” in the United States, and Supreme Court decisions have accepted significant discretion by agencies when they make and apply rules (see Metzger, 2017). Likewise, close linkages between administration and the administrative courts in France have tended to maintain, and perhaps reinforce, administrative autonomy. And Supreme Court decisions in the United States have at times interpreted the Administrative Procedure Act as giving agencies substantial autonomy (see below). In the case of the United States and other countries, different agencies may be given greater autonomy than others, whether through legal foundations or political practices.

The development of administrative law also has in some instances tended to reduce administrative discretion and autonomy. This has been especially true because of increased demands for citizen participation in making rules and because of more opportunities for citizens to contest rulings. For instance, as will be elaborated below, Swedish constitutional law grants extensive rights for citizens to scrutinize administrative decisions and public documents following adoption of the Freedom of Press Act in 1766. Likewise, there is some increased tendency to question controls exercised by political officials over their civil servants and over the ordinary procedures of public organizations.

As we discuss the several administrative traditions below, one question stands out perhaps more than others. If this is administrative law, which is dominant—“law” or “administration”? For example, somewhat paradoxically the French administrative system often is cited as being legalistic, but administrative law is dominated by administrators, trained very much like administrators in other grands corps of the state. Meanwhile, American administrative law, operating in a system that is usually deemed managerial, ¹ actually is heavily influenced by law and by the ever-present possibility of administrative actions being appealed to the regular court system. Similar patterns are found in Sweden, where the public administration enjoys high levels of autonomy and managerial values, but where administrative law largely structures those values. The strong application of Rechtstaat suggests a similar analogy in Germany.

We propose and apply a framework for analysis that enables the study of bureaucratic autonomy and administrative traditions in a comparative perspective. This framework is concerned with five central variables or categories that are present in most contemporary polities, albeit to varying degrees and forms. These categories by no means capture the full spectrum of legal instruments for bureaucratic control, and the categories themselves include extensive variation in their sub-components. Nonetheless, we aim to provide a general framework that can guide further data collection and comparative analysis.

The first category refers to the basic legal system (Head, 2011). The basic distinction between and evolution of two legal traditions (Common and Civil) helps us understand the fundamental composition and modes of the administrative state and its involved actors, bearing fundamental consequences for administrative autonomy (Daly, 2021; Duve, 2018).

The second point of comparison regards rulemaking processes. Rulemaking power refers to the capacity of public authorities to amend, repeal, or create administrative regulation. A common explanation for why the political government would grant public authorities with extensive rulemaking powers is because of the latter’s relative expertise on technically or socially complex issues. The rulemaking power is either formal or informal, where the former implies formal judicial hearing for public consultation over proposed rules, while the latter lacks such requirements but are de facto making binding rules (Custos, 2006). However it is made, secondary legislation represents a significant component of the rule-making activity of the State (Raadschelders, 2017), and therefore the powers and constraints contained in administrative law are significant for the governance capacity of the system.

Thirdly, we consider the central concept of judicial review, which refers to the process of judicial scrutiny of administrative action. Judicial courts with powers to invalidate administrative rules or actions is an essential component for checks and balances in the separation of powers. That being said, this fundamental feature is manifested differently across different administrative traditions. Arguably, the most notable aspect of judicial review affecting administrative autonomy is the dichotomy between ex ante versus ex post review. The former signifies judicial review by a reviewing agency that is independent from the agency making a specific regulation, where the review process precedes the adoption of the new regulation. Ex post review, on the other hand, implies review after a regulation is adopted (Asimow et al., 2020).

Liability is another important aspect of judicial review. The exercise of public power by the public bureaucracy and its officials must be correct, professional and lawful. It therefore is essential for the administrative law to describe the ways and the extent to which liability can affect the organization or the individual bureaucrat. Wrongful conduct by individuals can inter alia be punished by disciplinary or financial liability (Chaba, 2020), which we believe will shape bureaucratic autonomy in distinct ways.

The fifth and final dimension considered here is the nature of administrative courts. In addition to general courts, many states have specialized administrative courts to impose judicial review over public authorities. These can be specialized in issues such as taxation or environmental licenses (Bruff, 1991). To what extent are these separate courts able to compel administrative action, and how autonomous are they in relation to general courts? In addition, are the servants in these administrative courts trained separately from other civil servants?

Administrative Law and Autonomy in Four Administrative Traditions

We should emphasize two points about the four administrative traditions prior to examination. The first is that there can be substantial internal variation, especially in the Anglo-American case. There is no absolute uniformity, but still there are clear similarities with a historical foundation. The second point is that these traditions continue to evolve. There is a historical foundation, but these are living ways of coping with changes in governance and the environment of the State.

France

In this analysis France is used rather naturally as an example of the Napoleonic tradition of administration. Although many aspects of administrative practices in France have their roots in the Ancien regime (Dreyfus, 2000), they became codified during the Napoleonic period and have served as the foundations for this and other administrative systems (Ongaro, 2009). The French state has been a developmental state for at least part of its history, and as such it has tended to use the bureaucracy as a central, and to some extent autonomous, mechanism for governance (Silberman, 1993).

Perhaps the most important actor in French administrative law is the Conseil d’Etat. There is an extensive system of administrative courts in France, with the Conseil d’Etat as the apex (and at the floor). The members of the Conseil d’Etat at some point in history had been lawyers first, but after the creation of the École National d’Administration (ENA) most of the members are graduates of that school. ² They have more legal background than would most civil servants in Anglo-Saxon countries, but they are trained to be administrators. The logic of this selection is that who can better control bureaucracy than an expert bureaucrat?

The Conseil d’Etat is the final court of appeal for administrative cases, but it also is involved in reviewing legislation, including secondary legislation, before it is enacted. Under common law, a controversy is required before the courts become involved in addressing the legality or constitutionality of a government action. In the case of the Conseil d’Etat, and analogous bodies in other civil law countries, the design of review is more to prevent any cases arising, especially from poor drafting or conflict with other laws. Although operating within a codified legal system, the Conseil d’Etat and other administrative courts act in some ways more as if they were working with common law, using precedent and interpretation in effect to make new law. As Garner (1924, pp. 538–539) noted some years ago:

Even those who like Dicey have criticized the French system of administrative law as fundamentally wrong have expressed their admiration for the skill and ingenuity which the council of state, in particular, has shown in building up from year to year a vast system of jurisprudence and in devising new remedies for the protection of private individuals against the arbitrary and illegal conduct of administrative authorities.

The public bureaucracy remained a very powerful, and largely centralized, actor shaping policy as well as an actor that applies a codified set of laws. As already noted, however, the legalism that infuses the administrative system somewhat paradoxically means that law and administration are not really separated, and administrative law may be dominated by administrators. That lack of separation accentuates the autonomy of the bureaucracy as an institution.

Although in many ways dominated by administration and administrators, the administrative courts have developed a series of principles that can be used to constrain administration. As mentioned earlier, administrative actions can be reviewed in advance to determine if they are based on law and conform to accepted principles of administration.The principle of “recourse for the excess of powers” is analogous to a writ of ultra vires in the common law. Yet even more clearly than in the United States after Chevron, there is an assumption that if an administrative act falls broadly within the competence of the organization performing the act, it is legal.

Following from the above, if an individual citizen is indeed harmed by an action of the State or its agent, the State is responsible for providing the remedy. In Anglo-Saxon law in most instances, if an agent causes harm, then he or she may face a tort action, or even a criminal action, in the regular courts. ³ In French administrative law, the state itself is responsible, and the Conseil d’Etat has established principles by which individuals may be compensated because of torts committed by its agents, thus limiting autonomy by upholding the rights of citizens.

The autonomy of bureaucracy is further enhanced because of the connections between the bureaucracy and the rest of the economy and society. The civil service is the source of many elites in politics and in the private sector; it tends to be more controlling of, than controlled by, these linkages. Unlike many administrative systems there are relatively few legal constraints on the “revolving door” between the public and the private sectors, so this circulation of elites can link various segments of society (Rouban, 2010).

The United States

The Anglo-American tradition of public administration is less internally consistent than the other traditions. This distinctiveness is manifested in several ways, most notably in the existence of a statute that systematizes the rules for administrative procedures. While many Americans have the image of a runaway bureaucracy, especially after the diffusion of the idea of the “deep state,” constraints on administrative action are more clearly defined in the United States than in other Anglo-American systems (see Baldwin, 1996). US agencies may have autonomy, but it is exercised within legal limits.

That said, common principles of administrative law apply to the United Kingdom and the United States, as well as to other administrative systems within this tradition (Craig, 2001). They both come from the common law tradition and assume that law besides being written in statutes, emerges from interpretation and from common principles. The most general of these principles for public administration is that there should be some legal constraints on administrative action. Following from that, there also are general notions about rights of citizens to participate in the process, although the balancing principle weighs far more heavily against citizens in the United Kingdom (Thomas, 2000). As Mashaw (2012, p. 8) has written, “the task of the administrative constitution is to legitimate institutional designs that appropriately balance the simultaneous demands of political responsiveness, efficient administration, and respect for legal rights.”

The Administrative Procedure Act of 1946 (APA) controls rule-making and rule-adjudication in federal agencies as well as some other aspects of administrative process. Unlike law in most other Anglo-American administrative systems, the procedures for administrative actions are codified in a single statute, albeit one with numerous amendments. ⁴ The passage of this act at the end of World War II reflected the growth of governing during the New Deal and the war, and the jumble of procedures and rules that existed before the APA was adopted (Freedman, 1978).

The APA at once empowers and constrains administrative agencies. On the one hand, the Act clearly gives agencies the capacity to utilize delegated powers to make rules and to enforce them. It also gives agencies the capacity to adjudicate claims made against them, especially from clients who are denied benefits or who do not receive the level of benefits they expected to receive. Under the APA administrative agencies become the major source of rules and the major locus of adjudication in the federal government, far surpassing the activity of Congress or the regular federal courts (see Kerwin & Furlong, 2018).

Those powers to make rules can only be exercised within certain constraints, however. One of the important principles in the APA, especially for rulemaking, is that citizens and affected interests should know of the intentions of an agency to regulate and should be able to try to influence the contents of the regulation. Both major forms of rulemaking—informal and formal—provide access to citizens and interests. Informal rulemaking allows “notice and comment” in which the agency must publicize its intent to regulate and later the draft regulation; informal rulemaking permits comments from the public that become part of the record.

Formal rulemaking, as the name implies, involves an almost judicial process in which there are witnesses, a formal record is kept, and an administrative law judge is the presiding officer. The enabling legislation for individual agencies or individual programs specifis if formal rulemaking is required. Some legislation requires so-called hybrid rulemaking, with a more complete record being kept of the process and more detailed justifications of decision being supplied. The Occupational Safety and Health Administration and the Federal Trade Commission, for example, are required to use hybrid rulemaking.

The rules an agency makes through either process can be appealed in regular court system if there is a substantial constitutional or statutory issue involved, primarily that the agency is acting ultra vires. That said, under the so-called Chevron rule, ⁵ the courts tended to defer to the interpretation of the laws made by the agency unless there is a clear violation of Congressional intent. In 2022 the Supreme Court did, however, constrain that autonomy by emphasizing the “major questions” doctrine: Congress does not intend to delegate major policy decisions to the agencies. ⁶ In general, however, unless the plaintiff can demonstrate that the agency was acting in an “arbitrary and capricious” manner or is violating other aspects of due process, the assumption has been that the agency can write rules as it deems appropriate, so long as they do not breach some invisible threshold of a “major question.” In terms of autonomy, it appears that agencies with better reputations receive more deference than others.

The [Chevron] rule points to another important aspect of administrative law in the United States: the review of administrative actions is ex post. Agencies are given the latitude to act and then must confront any legal challenges. This does not mean that the rule must be put into effect, and courts may issue injunctions to prevent a rule that potentially is not within an agency’s appropriate bounds of action from going into effect (Friedman & Schwartz, 2020).

Although judicial review of agency actions may be ex post, political and economic reviews of those actions are not. In addition to the role of federal courts in regulating the actions of administrative agencies, the executive branch itself attempts to regulate administrative activity. In particular, the Office of Information and Regulatory Affairs (OIRA) within the Executive Office of the President reviews all rules issued by the federal bureaucracy not only for legality but also for conformity with presidential priorities (Sunstein, 2012). Especially under Republican presidents, OIRA has been vigorous in applying cost-benefit standards to regulations and attempting to modify those that are most costly.

Finally, while we have detailed a number of codified aspects of rulemaking and adjudication in the federal bureaucracy, a number of recognized informal processes also are utilized regularly. For example, the Federal Trade Commission may issue preclearance statements to firms seeking a merger but concerned about possible anti-trust violations. These informal mechanisms are not examined with the same scrutiny as other administrative acts, and they enable the agencies and their “clients” to make decisions and reach agreements more easily.

In addition, agencies may attempt to gain greater autonomy by using contractors or other agents to perform some of their tasks. These organizations are not necessarily under the same legal constraints from the APA as are the agencies themselves, and hence are able to exercise discretion in making decisions more easily (see Rosenbloom & Piotrowski, 2005). This can be an effective strategy for gaining additional autonomy, assuming that the contractors can be managed so as to follow the desires of their principal—the agency.

Germany

The German state administration emerged in several stages from the Middle Ages and onwards. Especially after the Thirty Years War (1618–1648), a new class of non-hereditary civil servants emerged and became part of the image of the state. The old idea of the monarch as the state (which Louis XIV referred to as “l’état c’est moi”) transformed into the idea of the state as a servant for the citizens. Shortly after the Thirty Years War, the Policeyrech was established and laid the foundation for an unrestricted lordly administrative authority (ius eminens). The introduction of exclusive legal protection against territorial lords provided the basic separation of public and private law in German legal culture (Bogdandy & Huber, 2017).

As in many other European states at the time, Germany was partitioned into several sovereign states at the beginning of the 19th century, but bound together in a larger legal and political framework, the German Confederation (Deutscher Bund) in 1816. Many notable states of the Confederation (e.g., Prussia, and Bavaria) followed the Napoleonic bureaucratic structure of centralization from the monarch to specialized ministers in specific departments. Provincial governments followed the French system of prefects as well as specialized agencies for welfare activities such as education and health. Moreover, in contrast to the French system, a decentralized self-administration, which had been institutionalized since the Middle Ages, remained in Germany. Thus, the German administrative-legal arrangement is founded on Napoleonic centralism with important elements of local self-government.

The system of various sovereign states operating within the Deutscher Bund was one of several factors that shaped the idea of Germany as a unified state. While the Imperial Constitution (Paulskirchenverfassung) of 1849 laid the foundation for several administrative courts in the various states, creation of the German Empire followed intensive war-making against Denmark, Prussia and Austria, and France. Germany thus evolved as a unitary actor and the most important legislator in the area. This development coincided with increased federalism, creating an intriguing mix of centralism and federalism (Ziblatt, 2006)

Perhaps the most famous German legal concept is Rechtsstaat. It refers to a set of fundamental legal principles of a more administrative rather than constitutional nature that regulate action in the bureaucracy and protect against executive power. Such principles developed between 1850 and 1900, including, for example, the so-called “lawfulness of the administration” (Gesetzmäßigkeit der Verwaltung) that underlines the supremacy of the law as well as legal restrictions for administrative violations of individual freedom and administrative control mechanisms. The concept of administrative law (Verwaltungsrecht) emerged in this context, as there was a gradual transformation of recognizing the German population as subjects to citizens. Decisions taken by the government were now written in statutory form, and administrative law would ensure its impartial implementation. This law-centered relationship between the administration and the citizens greatly influenced Max Weber’s concept of rational legitimation.

The German Administrative Act (Vervaltungsakt) developed in correspondence to the French act administratif. The results were that administrative invention did not need to be preceded by judicial decision and all doctrinal meaning is contained within the normative legal text itself. Yet, as Becker (2017) notes, the German administrative model differs from its French counterpart in that it underscores forms of administrative action. Substantial judicial review is applied to the legally binding individual administrative acts, which steer the rights and duties of citizens towards the state.

Today, public administration is divided into two separate functions; a hierarchically structured direct administration (unmittelbare Staatsverwaltung) and an indirect administration (mittelbare Staatsverwaltung). The latter refers to the delegation of powers to other (regional or local) entities, while the former refers to the formal power of responsible authorities inter alia to maintain public infrastructure or more broadly provide social welfare services (Sommermann, 2021, p. 24). Related to these features, the bureaucracy has lost much rule-making authority compared to many other EU member states during the 20th century, and there exists today comprehensive ex post judicial control over administrative action and a clear principle of proportionality of administrative legal action, which is always subject to judicial review. This principle is not explicitly stated in the constitution, but a consensus remains that it is a fundamental attribute of the rule of law. The level of proportionality is determined by three aspects: suitability (Geeignetheit), necessity (Erforderlichkeit), and balance (Angemessenheit).

Judicial review, which was inspired by the US constitution, is a key feature flowing from the development of the rule of law (Hoffmann-Riem, 2004). Judges were long part of the wider public administration and appointed by the same recruitment system. Administrative courts were, in consequence, and in contrast to criminal and civil courts, not part of the judicial system but rather of the administration. Such a weak stance of administrative courts persisted until 1960 when a statute (Verwaltungsgerichtsordnung) was enacted to make the administrative court system independent from the administration. Introducing the administrative court system into the wider judicial branch of public power has blurred the lines between civil law and public law.

Sweden

Sweden belongs to the codified, civil law tradition but with some distinct Scandinavian features. The single most important legal foundation is the Constitution (Grundlag), in which the Instrument of Government (Regeringsformen) has steered the organization of the public bureaucracy since 1974, but with earlier versions from 1634 and 1809. Fundamentally, it stresses the rule and equality of law, and it holds that no public officials shall make decisions without regulated authority. In addition, it underlines values of objectivity, impartiality, independence, and human rights (Kumlien, 2019; Marcusson, 2018).

The system of judicial review over administration action followed the procedure of the general courts, and no formal codification concerning procedure for administrative cases existed before 1971. Procedural rules for administrative regulations were heterogeneous across substantive areas, and citizens could complain about administrative decisions through internal appeals in the public bureaucracy. After 1971, however, a homogeneous procedural code (Förvaltningsprocesslagen) and a general law for public agencies (Förvaltningslagen) were enacted.

Since then, the three-tiered system of general administrative courts (Länsrätter, Kammarriätter, and Regeringsrätten) developed, and collectively these have received relatively more power vis-à-vis the government than earlier. The administrative courts have wide jurisdiction to review both the legality and the suitability of administrative actions, which is rather unusual from a European comparative vantage point, perhaps reflecting that the administrative courts were traditionally regarded as part of the administration rather than the court system.

Regarding personal liability for civil servants, there exists no sharp legal distinction between civil servants and other employees in the Swedish system. All are responsible for their actions according to penal law even if this responsibility is restricted to cases involving the exercise of public power, and there are also special disciplinary punishments for state employees.

The Administrative Procedures Act (Förvaltningslagen), enacted in 1971, and amended in 1986, protects individuals and guarantees them swift, correct, and efficient handling of administrative matters. It consists of only 34 sections and must be applied by all public agencies, with certain exemptions, and by the courts when handling administrative matters. Relatedly, the 1766 Freedom of Press Act (Tryckfrihetsordningen), providing citizens extensive rights to scrutinize the public bureaucracy, and the Parliamentary Ombudsman (Justitieombudsmannen), an independent agency under the Parliament responsible for controlling administrative decisions, together create a rather controlled public bureaucracy. However, Sweden is often described as having exceptionally high degrees of bureaucratic autonomy because of the prohibition of ministerial rule. According to the Constitution (Chapter 11, section 7), a public agency’s exercise of public power is made independently according to laws and ordinances. This coalescing of, on the one hand, bureaucratic control, and on the other hand, bureaucratic autonomy, constitutes a distinctive arrangement and often creates controversy over responsibility (Hall, 2015).

Most appeals are made to one of the twelve lower administrative courts (Förvaltningsrätter). The review concerns strictly legal aspects and aspects of the suitability of administrative decisions. The specific administrative court may not only rescind or uphold the decision but also issue a new decision on the matter. In this way, the right to administrative appeal at court or to an authority gives an extensive opportunity to question administrative decisions in detail and in broad aspects. The same opportunities for review do not exist for the local or regional levels, however, protecting local self-government.

Conclusion

Understanding administrative behavior, broadly defined, cannot be underaken in isolation from the legal structures under which the individual is nested. We have proposed a comparative framework for analysis of how several aspects of administrative law can influence autonomy. This function is indeed a basic purpose of administrative law: to define the appropriate sphere of action for public organizations and in the process to define where those organizations have latitude for autonomous action. As demonstrated, the core features of administrative law (the legal system, rulemaking authority, liability, courts, and judicial review) vary considerably across administrative traditions. Thus, we see great potential for future studies to extend this research endeavor in alternative contexts and research designs. For example, when studying bureaucratic autonomy, the legal tradition may conflict with contemporary political pressures, with different traditions finding varying mechanisms for resolving the conflict. A number of factors shape these traditions in addition to law, but law helps to define each tradition as well as being defined by it.

We have discussed the four main administrative traditions within Western Europe and its diaspora. The four countries chosen represent important illustrations of those traditions but by no means do they represent the entire tradition. This is especially true for the United States within the Anglo-American tradition. That said, these examples do demonstrate how administrative law can be related to bureaucratic autonomy, and how the legal system can be used to both limit and expand spheres of action for administrative agencies, for individual administrators, and for the bureaucracy as an institution.

This article presents something of a research agenda. We have posited to importance of administrative law for bureaucratic autonomy, and given some examples of how that relationship functions in four settings. A great deal still remains to be done, however, to explicate the linkages among these variables and to detail the characteristics of national legal systems that produce particular types of bureaucratic patterns. Further, this may open a broader research agenda that will examine how law, politics and social factors interact to produce outcomes in the public sector.

We also have brought administrative law back directly into the study of comparative public administration and have shown that this is not some arcane study for lawyers but instead a central component of the functioning of the public sector. Administrative law is one of several important institutional variables that define administrative systems and administrative behavior. It therefore deserves to have a more central place in the contemporary study of public administration.