Bringing the Law Back In: The Law-Government Nexus in an Era of Network Governance

Abstract

Introduction

In this article, we answer to calls to bring the law back in public management theory (Beckett 2010; Christensen, Goerdel, and Nicholson-Crotty 2011; Cooper 2007; Lynn 2009; Moe and Gilmour 1995; Rosenbloom, O’Leary and Chanin 2010; Wright 2011; Zouridis 2011, 2014). We argue that we should not simply bring the law back in public management theory but instead learn from history in order to avoid the pitfalls from the past. Specifically, bringing the law back in should not mean returning to the old ways. On the contrary, in this article, we pick up the challenge to redefine law as a foundation of public management by taking into account some historical lessons taught by public management theory.

The key question in studying how to reconnect law and public management is: “How should the link be understood conceptually?” We introduce the concept of the law-government nexus to answer this question. The law-government nexus refers to the assumptions one holds on the link between law and government. In order to answer to the recent call to bring the law back in, we explore how the law-government nexus has changed during the history of public management theory. Based on the assumptions underlying the law-government nexus, we explore the law-government nexus in three paradigms derived from public management theory. We also formulate three strategies to reconceptualize the law-government nexus in the prevailing paradigm.

In this article, a paradigm refers to the conception on what government is. As we will show with regard to the law-government nexus, three paradigms can and have been distinguished in public management theory: a public administration (PA) paradigm, a new public management (NPM) paradigm, and a (new) public governance (PG) paradigm. Each paradigm displays a different set of assumptions on law, government, and the link between law and government. Hence, each paradigm displays a different law-government nexus. These paradigms have been guiding public management as a science, as an art, and as a profession (Lynn 1996). As Pollitt and Bouckaert (2017) show these paradigms have inspired public management reform and even austerity programs. Next to the NPM paradigm, they distinguish the Neo-Weberian State (that resembles our PA paradigm) and the New Public Governance paradigm (our PG paradigm).

Along with a theoretical exploration of the assumptions on the law-government nexus in each of the paradigms, we use public management practices in a typical government agency such as the Dutch tax agency to illustrate the effects of these assumptions. The Dutch Tax and Customs Administration (DTCA) is illustrative for our article, as it has over the years reorganized and implemented structures and procedures reflecting the three paradigms as well as the law-government nexus that goes along with these paradigms. As we will show, other tax agencies also changed their organizational structures and routines. We will illustrate the advantages and flaws of each of the paradigms that the DTCA and other tax agencies experienced.

Section 2 introduces the theoretical framework on the law-government nexus and its foundations in three paradigms in public management theory. Section 3 presents the DTCA case and some commonalities with tax agencies as described by public management scholars. In section 4, we explore the dynamics between the law-government nexus in these paradigms, identify strengths, weaknesses, and pitfalls, and formulate lessons that should be taken into account in bringing the law back in. In section 5, we build on these lessons to explore three strategies towards a revised law-government nexus. In the final part of this article, we discuss the limitations and implications of the reconceptualization, including some future research directions.

The Law-Government Nexus and its Embeddedness in Public Management Theory

While recent calls to bring the law back in public management theory suggest that law left public management, we argue the opposite: our key argument is that law never left public management theory. We use the concept of public management theory in a broad sense as the overarching field that focuses on leading and controlling public institutions. As an overarching field, public management theory may build on theories of law, economics, organization studies, and psychology.

While law never left public management theory, the approach to the link between law and government did change. We refer to this link as the law-government nexus. In this nexus, we use the concept of law in a formal sense in contrast to natural law approaches (see among others Freeman 2001 for the exact differences). In our conceptualization, law refers to the rules, rights, principles, procedures, checks, and structures enshrined in legislation, case law, and the constitution. The concept of government in the law-government nexus refers to the ensemble of public agencies that govern societies. As these definitions show, law and government are strongly connected. Usually governments produce laws and, in turn, abide these laws.

The definitions also show that analytically law and government can and should be separated. This is not merely a philosophical question. As legal scholars have shown both in the history of law and more recently in the realm of transnational law, law can be separated from government if both the production and enforcement of rules are taken over by private non-state actors. Many examples of law without government have been described in the realm of transnational law and in legal history (e.g., Basedow 2008; Muller et al. 2011; Saiman 2008).

So, while law without government is not something imaginary, neither is government without law. During the Cultural Revolution in China, law was largely replaced by policy and the rule of law only re-appeared during the 1980s (see e.g., Peerenboom 2002). This may sound like a matter of semantics, but it is not. While policies merely express political will, laws also bind political will. If the categories of law and government can be analytically separated, several models spring up with regard to the link between them (e.g., Zamboni 2008).

Any version of the law-government nexus connects with a particular paradigm as developed in public management theory (Lynn 1996; Pollitt and Bouckaert 2017). The history of public management theory shows at least three paradigms. Each paradigm also reveals a particular set of assumptions on the law-government nexus. First, the PA paradigm, born out of a rejection of the then prevailing exclusive constitutional-legal focus on government, considers government as administration, and law as the set of rules governing administrative decision-making. Its law-government nexus assumes a close connection in which law precedes PA. The lawfulness of the administrative state is both a key focus and a key value of the PA paradigm. Research driven by this paradigm focuses on understanding how the administrative apparatus implements and executes the legal orders spelled out in the constitution. This research aims at explaining the discrepancies between the law and everyday administrative practices. Research on bureaucracy, implementation, regulatory justice, street-level bureaucrats, and large-scale administrative organizations illustrates the law-government nexus in the PA paradigm (e.g., Kagan 1978; Lipsky 1980; Merton 1940; Selznick 1949). As Lynn (2001) has argued, this paradigm can be easily misunderstood. The caricature image of bureaucracy and red tape associated with this paradigm distorts a sound understanding of the PA paradigm as an expression of legal-rational governance as understood and portrayed by Weber (1922).

The NPM paradigm assumes that government is not about administrating the law but about managing public services. The shift from the PA to the NPM paradigm was at least partially triggered by the pathologies of the PA paradigm. As Weber (1922), Kaufman (1977), and Bardach and Kagan (1982) among others already argued, a law-based administrative apparatus produces bureaucracy, red tape, and legalism. While the PA paradigm positively values these phenomena since they express lawful government, the NPM paradigm redefines them as pathologies. The NPM paradigm also adopts a different approach to the law-government nexus. While the PA paradigm looks at law as constitutive for the administration and its routines, the NPM paradigm considers law as merely a regulatory tool available for public institutions in order to enhance the efficiency and quality of public services. Law is just a tool that serves the purpose of achieving policy goals (see Baldwin, Cave and Lodge 2011 for an illustration of this approach even though they consider a legislative mandate as a measure of regulatory quality). Legal scholars have fundamentally contested this assumption of law as a regulatory tool referring to the adverse effects of legal instrumentalism on the legitimacy of both law and government (Tamanaha 2006). Whereas lawfulness provided the normative foundation for the PA paradigm, efficient government according to economic standards supplants lawfulness as a normative foundation in the NPM paradigm.

During the past decades, a new paradigm gradually complemented the NPM paradigm. The still increasingly popular PG paradigm builds on the concept of network governance. Again, the phenomena considered as normal in one paradigm become the pathologies of the other paradigms. Efficiency, performance, and contracting out may have been the cornerstones of the NPM paradigm, but the PG paradigm shows how artificial markets dramatically affected the quality of public services and how perverse incentives triggered practices close to administrative fraud (e.g., House of Commons 2014). The PG paradigm approaches government in a broader sense as a set of structures and processes that transcend the realm traditionally referred to as government. Governance is broader than government as it encompasses the structures and processes with the inclusion of private actors such as businesses and not-for-profit organizations. The paradigm shift from NPM to PG also triggered a new law-government nexus. As Wise already argued in 1990, privatization is not simply moving public services to private organizations. It alters the interorganizational field in which the responsibilities of public organizations do not disappear (Wise 1990). The PG paradigm builds on this understanding of public interest and the responsibilities of public organizations. The PG paradigm also challenges the public-private distinction in administrative law (Freeman 2000). Private actors may play public roles, but traditional administrative law and administrative procedure build on the PA paradigm that only recognizes public organizations (Rose-Ackerman and Lindseth 2010). The PG paradigm also reconceptualizes law as a networked phenomenon. Instead of the lawfulness of the PA paradigm and the efficiency of the NPM paradigm, the PG paradigm aims toward an effective ensemble of networks to address complex societal problems and challenges. Academic research focuses on the dynamics of these networks, on the management of these networks, on the effectiveness of networks, and on the possibilities of these networks for governance purposes (e.g., Bovaird and Löffler 2003; Kenis and Provan 2009; Scott and Ulibari 2019; Wachhaus 2012; Xiong et al. 2019). Table 1 summarizes the law-government nexus in the three paradigms.

All three paradigms have figured and interacted in the history of public management theory. While some observe a succession of paradigms, others have emphasized the tensions between the values that underlie these paradigms (e.g., Pollitt and Bouckaert 2017). Kaufmann (1956) analyzes the historical shifts from the values of representativeness to neutral competence and leadership. Hood (1991) uses three competing sets of values. The Sigma-type values of lean and efficient organization prevail in the NPM paradigm, the Theta-type values of fair and honest decision making constitute the PA paradigm, and the Lambda-type values such as resilience match with the PG paradigm.

While our presentation of these paradigms may suggest neat demarcations and a succession of paradigms, reality is more complex. Nevertheless, the paradigms do provide a useful conceptual framework for a better understanding of how law and public management are connected. The law-government nexus in each of the paradigms points at how ways of doing law interact with ways of doing administration and public management. In the PA paradigm, a rule-based concept of law goes along with an administrative and bureaucratic reading of public management. The NPM paradigm triggers an economic perspective on both public management and law, which conceives law as a regulatory tool of government. Because the PG paradigm transcends the traditional boundaries of government, it also transcends traditional ideas on law as a phenomenon inherently connected with public authority. It would be too hyperbolic to argue that law and government become liquid as described by Bauman (2000), but to a certain extent, the PG paradigm does trigger a liquid conception of both law and government. The law-government nexus in each of the paradigms does not only show how ways of doing law interact with ways of doing administration. The nexus also offers tools on how to reconnect law and public management in a prospective way that builds on the lessons taught by history. In the next section, we identify these lessons by examining the strengths and weaknesses of each paradigm and the dynamics between the paradigms using the DTCA case.

The Law-Government Nexus in Public Management Practice: An Illustration

Before we formulate the strategies to redefine the contemporary prevailing law-government nexus of the PG paradigm, we first assess the historical lessons that should be taken into account. In this section, we illustrate the effects of the law-government nexus in the three paradigms with public management practices in the case of the DTCA. We use the case of the DTCA, as over the years it has continuously tried to improve its performance and organization by reorganizing and implementing structures and procedures reflecting the three paradigms. As we show, by following the paradigms the DTCA has gained a reputation for some of its innovative procedures both nationally and internationally, while some of the pathologies of the paradigms led the DTCA into difficulties. However, the DTCA is not unique in this. For example, many tax agencies have been inspired by the NPM paradigm, and changed their bureaucratic organization into a more customer friendly organization (Aberbach and Christensen 2007). Furthermore, more recently, the PG paradigm triggered tax agencies over the world to develop cooperative compliance approaches (Walsh 2012). We will use the DTCA as an exemplary case and draw multiple commonalities and parallels with other tax agencies.

DTCA in PA

The assumptions underlying the law-government nexus in the PA paradigm can be easily detected in tax agencies before the 1980s. For example, tax agencies in the Netherlands, the United Kingdom, and Norway were mainly focused on rules and norms, and the operation and implementation of tax law (Aberbach and Christensen 2007, Tuck, Lamb, and Hoskin 2011). In the Netherlands, both the organization and the processes of the DTCA were guided by tax law (Gerritsen and van Breukelen 1992). For example, the structure of the administration followed the types of tax as defined in law. This means that the DTCA had separate inspections for wage tax and income tax, corporation tax, turnover tax, et cetera. In addition, for the three processes of assessment, control, and collection, separate sections were responsible (Alink and van Kommer 2012). In practice, this segmentation involved numerous local units performing the same or similar tasks. To illustrate, in the Netherlands there were 260 local tax units, in the UK 760 (NAO 1996), and in Norway 435 (Aberbach and Christensen 2007).

Furthermore, regarding the processes of the tax agencies in the Netherlands, United Kingdom and Norway, tax legislation was the main criterion influencing both thinking and acting. This meant that in general tax agencies managed tax returns on an individual basis (Aberbach and Christensen 2007; Alink and van Kommer 2012). For example, in the Netherlands, the philosophy of the administration was to levy taxes in a proper way, implying that levying taxes had to meet three criteria: justice, legal certainty, and legal equality (Voigt 1994). Following this, the DTCA formulated the 100% philosophy, which entails that each tax return was treated in the same way, with the same attention, at the same level, and within the same period, irrespective of whether the problems involved or the tax risks were different (Alink and van Kommer 2012).

Having tax law as both the cornerstone and the foundation comes with advantages and drawbacks. In the case of the DTCA, in the 1970s the administration was perceived as a reliable tax agency with clear responsibilities and occupied by professional bureaucrats (Alink and van Kommer 2012). However, the flip side of the coin was that the strict use of tax law comes with the perception that tax agencies are rigid, introvert and not user-friendly, as was the case in Norway (Aberbach and Christensen 2007). Unfortunately, the DTCA could not uphold its reputation, when in the early 1980s it became apparent that the organization and operation of the DTCA were not sustainable. Tax legislation became increasingly difficult and complex, which exposed the lack of client service, as the DTCA was unable to respond to the questions and demands for assistance provoked by the complex legislation (Alink and van Kommer 2012; van Kommer 1998).

Next to the inefficiency, both the Dutch and Norwegian tax agencies had problems with their large-scale bureaucracy, which created disparate and unclear procedures (Aberbach and Christensen 2007). To illustrate, in the Netherlands, because of the structure of the DTCA, taxpayers had to deal with different (local) offices, each with own approaches, forms, and procedures. Taxpayers who wished to do so were able to play different parts of the DTCA off against each other (Alink and van Kommer 2012). Furthermore, combating misuse of tax returns ran across many units and offices. No unit or inspection had a complete overview, facilitating tax fraud and tax avoidance (de Kam and van den Berg 1992; Voigt 1994).

Finally, because each unit and inspection was responsible for only a part of the process, there was a high degree of inflexibility and legalism. Accordingly, complaints and petitions related to the slow processing of tax returns increased.

DTCA in NPM

In the late 1980s and early 1990s, multiple tax agencies changed their organizational structure and practices (Aberbach and Christensen 2007; Currie, Tuck, and Morell 2015). Generally, they changed from a bureaucratic led organization towards a customer-oriented service provider (Aberbach and Christensen 2007; Alink and van Kommer 2012; Tuck 2013). These changes did not come out of nowhere. Tax agencies felt a great need for change, as they faced heavy arrears (NAO 1996), and great risks for tax revenue and public fiscal ethics (Alink and van Kommer 2012). Contrary to previous concerns, in the Netherlands, revenue became an important element of interest. Hence, the return on investment in the DTCA even became a cornerstone of management philosophy. The DTCA redefined its paradigm shift as a transformation of a strictly hierarchical, bureaucratic, risk-avoiding, and inward-looking organization toward a collegial, outward-looking, service and output-oriented one (Alink and van Kommer 2012; de Kam and van den Berg 1992).

Generally, we observe three organizational reform trends in tax agencies: reorganization of the organizational structure, focus on efficiency, and customer friendliness. In the Netherlands, these reforms involved a clear change in the DTCA’s foundation: from law-oriented to business-oriented. Tax legislation no longer guided the organization of the DTCA (Voigt 1994), but rather the DTCA adopted an entrepreneurial philosophy for its structure, management, and processes (Koster and Muizelaar 1992). Concepts such as decentralization, contract management, planning and control, service orientation, and customer friendly played a pivotal role (Alink and van Kommer 2012; Koster and Muizelaar 1992).

Furthermore, tax agencies restructured their organization and procedures according to the taxpayer’s characteristics. In the United Kingdom, Norway, Australia, the Netherlands, and the United States, tax agencies structured their agency based on target groups or taxpayers segments, rather than the type of tax law or means of taxation. (Aberbach and Christensen 2007; Alink and van Kommer 2012; Tuck 2013).

The restructuring, subsequently, involved a reorganization of the organization. In the Netherlands, for example, middle management slimmed down and reorganized similarly to the classification of the target groups. Responsibilities regarding staff, organizational structures, and the use of financial budgets were decentralized to local units as much as possible (Gerritsen and van Breukelen 1992; van den Berg and Janssen 1994). As a result, the number of units reduced from 260 to 89 (Gerritsen and van Breukelen 1992). In Norway, the 435 local offices merged into 99 units, resulting also in a more horizontal organization with only two levels: regional and central (Aberbach and Christensen 2007). The tax agency in the United Kingdom as well reduced their tax offices from 760 to 380, distinguishing between tax service offices, tax district offices and tax enquiry centers (NAO 1996).

Furthermore, there appears to be a value shift. While before the 1980s the DTCA relied on values such as justice, legal certainty, and equality, after the reorganization values such as economy and efficiency prevailed (Voigt 1994). Consequently, the DTCA integrated the different types of tax and working processes within one local office. As a result, taxpayers could turn to one office for all their matters instead of having to deal with different local offices (Alink and van Kommer 2012; Koster and Muizelaar 1992). In the United Kingdom, the mission of the tax agency was reformulated to “to provide an efficient, effective, and fair tax service to the country and government.” This involved an emphasis on four values “The Four C”s’: Customer service, Compliance, Cost efficiency, and Caring for staff. The first value, Customer service, aimed at a more efficient and modern service. Secondly, with the value of Compliance, the tax agency tried to increase the effectiveness and efficiency of the compliance activities of taxpayers. Thirdly, the value of Cost efficiency emphasized the aspiration of the tax agency to reduce costs. They tried to do this through for example more efficient staffing. Lastly, the tax agency emphasized the value Caring for staff, meaning that the agency aimed at providing opportunities for staff, improving the work environment, and improving communication (NAO 1996).

Finally, as for the processes, tax agencies increased the standardization of the treatment of taxpayers. To illustrate, the DTCA abandoned the philosophy of total control of its taxpayers. The new risk approach was based on risk selection and differentiated tax returns. Simple cases were separated from the complicated, complex, high-risk cases. The latter category got a thorough treatment, while the former could be taken care of administratively or even fully automated with the use of ICTs (Voigt 1994; Zouridis, Van Eck and Bovens 2019). One of the ideas behind the new approach was that improvement of the primary processes could overcome backlog and, therefore, decrease the tendency of taxpayers to misuse or avoid taxes (Gerritsen and van Breukelen 1992). Similarly, in Norway, the rationale of the standardization was to increase the fairness and improve the efficiency (Aberbach and Christensen 2007). The United Kingdom adopted a tax compliance model, the regulatory pyramid (Braithwaite 2007), which distinguishes between customers who comply and those who do not. As customers go up in the pyramid, the enforcement strategies increase (Tuck 2013).

However, while the reforms succeeded in decreasing backlog, creating a smaller tax agency, with more service-oriented and customer-friendly approaches, the tax agencies also became less personal, distant, primarily concerned with products, cost allocation, and performance management, and excessive administrative attention for performance measurement and reports (Aberbach and Christensen 2007; Stevens 2007). In the Netherlands specifically, the reorganization created an introvert organization, which was mainly concerned with its own interests and processes (Gribnau 2005).

Due to the focus on mass processes, the administration started thinking and acting in terms of quantity instead of quality (Gribnau 2005). Individual attention required a disproportional amount of time and money. Therefore, taxpayers became numbers rather than individuals. An accurate and respectful treatment of taxpayers contradicted the businesslike approach. Consequently, the craftsmanship, knowledge, skill, and service of the civil servant were compromised. In fact, civil servants became client managers to be held accountable via production statistics.

Furthermore, following Stevens (1999), the approach of the DTCA provoked citizens to behave like entrepreneurs, resulting in calculating citizens. Indeed, taxpayers were referred to as clients instead of taxpayers (Stevens 1999).

DTCA in PG

The aim of governments to deliver their services more effectively while at the same time lowering their costs is still an important point on the agenda (OECD 2013b). Additionally, tax agencies encounter growing compliance risks with increasing globalization (OECD 2013a, 2019). Motivated by these developments, tax agencies over the world reorganized their management, organizational structure, and approach (Alink and van Kommer 2012; OECD 2006).

In the Netherlands, for example, the DTCA reorganized its management structured in the early 2000s into a more flat management structure. A network structure replaced the old top-down focus. Key values in this network structure became self-steering, mutual adjustment, working together, sharing and exchanging information, and learning (Alink and van Kommer 2012).

For the organizational structure, the OECD (2006) reports that most tax agencies (e.g., The Netherlands, Finland, Canada, Germany et cetera) strategically kept the target group orientation; however, tax agencies abandoned the target group differentiation in their organization structure: the target group divisions were eliminated. Instead, tax agencies based their organizational structures on the functional model, meaning that tax agencies organize their organization in groupings of their core functions, such as for example, registration, accounting, information processing, audit, collection, and appeals (Department of Finance 2003; Hauptman, Horvat, and Korez-Vide 2014; OECD 2006). By doing so, tax agencies tried to counteract the compartmentalization in national policy development, to reduce the distance between policymaking and implementation, to achieve a more balanced law enforcement between the different target groups, to increase the exchange of knowledge, to improve operational efficiency, and to enable greater standardization of work processes across taxes (Department of Finance 2003; OECD 2006).

Thirdly, while previously management has been organized more hierarchical, tax agencies in the PG paradigm, introduced new ways of working to improve productivity by increasing flexibility, autonomy, and cooperation. In the United Kingdom, for example, the tax agency adopted the “Smart Working” approach. With this approach, staff members receive greater autonomy and flexibility, which empowers staff to make their own decisions over how, when and where to work. Additionally, collaboration between teams and departments is improved (El-Gamry and Heselwood 2018). In the Netherlands, the DTCA emphasized self-regulating management teams. The integral manager was no longer solely responsible, but rather all team members became responsible for achieving the goals of the organization. Every team had to set goals, identify tasks, perform, deploy resources, et cetera. Within the teams, proper agreements and arrangements had to be made regarding who was responsible for what, who would pay attention to which specific area, files, items, et cetera (Alink and van Kommer 2012).

Furthermore, in addressing the increased compliance risks, tax agencies introduced cooperative approaches to increase their compliance efforts (OECD 2013a). These strategies differ from one tax agency to another. In general, the strategies involve a change from the traditional control approach, which based enforcement on increasing the chances of being caught, and strong criminal penalties, to a cooperative compliance strategy that takes factors that influence taxpayers’ behavior into account (Walsh 2012), relies on the responsibilities of taxpayers (van Lint 2016) and maximizes voluntary compliance (OECD 2013a). For the most part, these approaches are aimed at large businesses (OECD 2013b).

The compliance strategies differ from one tax agency to another. For example, in Sweden, the “Right from the Start” model is adopted, in which the tax agency creates a comprehensive view of the taxpayer by gaining knowledge regarding, for example, education and active engagement of the taxpayer (Walsh 2012). In Australia, the tax agency adopted the attitude that taxpayers are willing to do the right thing. The tax agency makes a risk categorization of large businesses being examined enabling to differentiate between cases who need more attention (OECD 2013b).

In the Netherlands, the compliance strategy is called horizontal monitoring (Horizontaal Toezicht). This type of supervision is based on mutual trust between taxpayer and DTCA. The DTCA aims both to redress the trust relationship with the taxpayer and to gain efficiency. In order to check whether businesses can be trusted, the DTCA uses internal and external audit mechanisms such as the accountants and certification schemes.

With horizontal monitoring, the DTCA builds on the appropriateness of the tax return by using the monitoring systems of large companies and tax advisors. The goal of horizontal monitoring is to conclude voluntary enforcement agreements. In these agreements, mutual requirements regarding the past and the future are defined. In addition, arrangements can be made on yet unforeseen acts or activities with fiscal consequences. The assessment of tax returns addresses primarily the acceptability rather than the correctness (Raaphorst 2018). Trust rather than distrust is the basis. Horizontal monitoring has become increasingly popular. Over the years, the concept of horizontal monitoring has expanded. Firstly, only large companies were involved in the agreements with the DTCA. Later on, agreements were also made with tax intermediaries, financial advisors, and sector organizations (Stevens 2012). The DTCA increasingly considers these networks of businesses, tax intermediaries, software developers, and tax consultants as opportunities to enhance tax compliance.

While these compliance strategies increase the flexibility, transparency, and mutual trust between taxpayers and tax agencies, there are also points of attention and drawbacks. Firstly, most cooperative compliance strategies are not embedded in law or regulations. The majority of countries formalized the cooperative compliance in agreements or declaration of intention (OECD 2013b). While these strategies aim at creating more certainty, the lack of formalization leads to ambiguity, for example, regarding accession criteria or the balance between fair play and quality of arms (Björklund Larsen 2016; Stevens 2012). Moreover, most taxpayers cannot receive a binding response on issues taking up in the cooperative compliance strategy. This was mainly a problem in Denmark and Finland. The responses taxpayers get from the tax agency are informal notifications. This is at odds with the promise to receive certainty in exchange for transparency (Björklund Larsen et al. 2018). Crucially, problems with equality and legality, especially in Sweden, were a major reason to put the compliance strategy on hold. The strategy implied that taxpayers would not be treated in the same way, which is against the Swedish constitution. In addition, it was not clear who could make what type of legal decision. Without a change of law, the strategy was not tenable (Björklund Larsen et al. 2018).

In the Netherlands, the horizontal monitoring evaluation committee concluded that the standard agreements in the contract jeopardize custom-made services. Although taxpayers have transparency and predictability advantages, they are very limited when it comes to taking ownership in the agreements. Finally, horizontal monitoring increases the chances of corruption because of the close non-business relationships that occur between the DTCA and taxpayers.

Paradigm Dynamics and Lessons

The example of the DTCA and other tax agencies clearly illustrates the strengths, weaknesses, and pitfalls of all three paradigms. While the PA paradigm is strong on lawful government, checks and balances, and its embeddedness in democracy, it is quite weak on efficiency. Its legalism and government-centeredness prevent a customer orientation let alone the effective governance of complex problems such as tax avoidance. The NPM paradigm may be strong on efficiency, the performance of public institutions, and the quality of public services, but it triggers perverse effects and transaction costs and its legal instrumentalism is detrimental to the rule of law. It also leads to administrative practices that deviate from the exact legal rules and even the legal system. Finally, the PG paradigm is quite strong on effective governance for complex problems and mobilizing actors for the public interest, but network governance hampers transparency, checks and balances, and lawfulness. These strengths and the counterproductive effects and pathologies of the PG paradigm have been identified in the literature, and the DTCA illustrates these effects. Moreover, as shown in table 2, the strengths of the paradigms largely mirror the weaknesses and flaws of the other paradigms. The vulnerabilities of the PG paradigm, for example, are its weak legal and constitutional foundations, which are the strengths of the PA paradigm. Calls to bring the law back in should therefore not come as a surprise ever since the PG paradigm became more popular.

Both the overview and the dynamics of these paradigms and their approaches to the law-government nexus provide two historical lessons on how to bring the law back in public management theory. A first lesson is that bringing the law back in should not mean reversing history and going back to the PA paradigm. Learning from the PA paradigm may restore the under-legalized conception of public management, but it comes at the cost of bureaucracy, red tape, legalism, and government-centeredness. In turn, it will trigger an over-legalized conception of public management.

A second lesson is that moving forward while bringing the law back in should build on the benefits of both the NPM and PG paradigms. Bringing the law back in should not ignore the performance, efficiency, and quality of public services that came along with the NPM paradigm. Neither should bring the law back in ignore the benefits of network governance, as emphasized in the PG paradigm, such as involving non-state actors in governance networks. So, bringing the law back in means that the NPM and PG paradigms should not be rejected. Instead, in order to avoid the flaws and weaknesses of the PG paradigm we propose to constitutionalize the PG paradigm. In the next section, we present three strategies to constitutionalize the PG paradigm.

Three Strategies

Since law was never pushed out of public management theory, bringing the law back in implies redefining the law-government nexus in the age of the PG paradigm. Redefining the law-government nexus starts with addressing the weaknesses and vulnerabilities of the PG paradigm. At least four weaknesses should be addressed. Firstly, as argued by Amsler (2016), theories on collaborative governance greatly ignore the legal framework in which the collaborative interaction takes place. Without an understanding of constitutional law and administrative procedure, the processes and the outcomes of collaborative governance may be vulnerable to judicial review, and they may also deviate from legal standards and principles.

Secondly, the focus on networks and flexibility rather than lawfulness results in diffused responsibilities, as networks almost inherently produce diffuse responsibilities and rules. The network acts, develops laws and regulations, and enforces public or public-private standards. As Wise (1990) argued, the public organization configuration alters in such a way that public responsibilities are blurred. In network governance contexts, it is not clear who is responsible for what let alone who is responsible for the network interactions and interdependencies. In effective and successive networks, this question will not be urgent. Yet, for failing networks, the question is rather important. Who, for example, will claim responsibility and account for the governance failure?

A third pathology follows from the diffuse responsibilities and rules produced by the PG paradigm. Co-creation may be a no-regret strategy, but what happens if it eventually leads to failure or inaction when action is required? For academics, there might be a perfect match between the complex and contested problems to which the PG paradigm claims to be a solution, but in real life, these problems have real consequences. Migration, organized crime, tax avoidance, corruption, cybersecurity, and climate change may be interesting phenomena for academics because of the complexity, the transnational nature, and the “wickedness” of these issues, but from the perspective of citizens, the effects are very real, local, and operational. If the networks that should address these problems produce failure or inaction and no single actor can be held accountable, public dissatisfaction rises.

Lastly, we consider public dissatisfaction in a more generic sense to be a real risk of the PG paradigm on the law-government nexus. Two decades ago, Pierre and Guy Peters (2000) argued that governance networks require strong states and politics governing the networks. Without strong states, the networks will not be able to address complex social problems. A radical turn to PG and network governance without strong PA and NPM, therefore produces governance that fails to meet the demands of citizens. As Rosanvallon (2008) suggested, the very concept of governance may already de-politicize the networks and become a key source of public dissatisfaction and populism.

Bringing the law back in thus means that these weaknesses should be addressed instead of reversing history and going back to the PA paradigm. How can the messy responsibility structures in governance networks, the lack of legal certainty because of the rule jungles developed in these governance networks, and the closed-door processes in governance regimes be countered by law? Going back to the PA paradigm would address these issues but at the high cost of re-introducing bureaucracy, red tape, and legalism. Instead, we propose to reconcile the benefits of the PG paradigm with the strengths of the PA paradigm. Such a reconciliation requires a redefinition of the law-government nexus in the PG paradigm in at least three ways that touch upon both the structural and the cultural aspects of network governance.

On the structural aspects, the PG paradigm goes along with blurred and diffused responsibilities of actors that, in turn, cause responsibility gaps. These responsibility gaps may cause governance failure and, in turn, trigger blame games and accountability gaps that foster public dissatisfaction (e.g., Hood 2010; Resodihardjo 2019). The capacity to deal with complex governance problems, to a certain extent, requires diffuse responsibilities, emerging networks, and bottom-up coordination. In the PA paradigm, the law clearly demarcates legal responsibilities and attributes legal tasks to actors. Redefining the law-government nexus thus means a stronger focus on the clarity and legal basis of the responsibility structures in governance networks. The traditional PA perspective on the law-government nexus can be used to identify the legal basis of responsibilities in governance networks as well as responsibility gaps in these networks. The adaptiveness and the flexibility of network structures require vague network roles and emerging responsibility structures, but a revised law-government nexus also reveals the need for lawful and clear-cut responsibilities. Including a focus on responsibility and accountability structures also better warrants legal certainty for all actors involved in the governance network.

Governance networks are claimed to be strong on effectively dealing with complex governance problems such as climate change, transnational organized crime, tax avoidance, and urban development. Due to the nature of network governance, these networks are also prone to failure because of group think and other collective cognitive biases, such as the inertia bias (Janis 1972; Meyer and Kunreuther 2019). Again, the PG paradigm may benefit from a redefined law-government nexus. A redefined law-government nexus should not return to meticulous implementation of legislation but instead build on rule of law-related values, checks, and balances in governance networks. Checks and balances can and should better control and restrain governance networks. Governance networks may be effective, but they are also powerful. Checks, such as openness on information, balanced power structures that include countervailing powers within the governance network, and, for example, a stronger focus on the position of courts in governance networks may be used to control and restrain the powers of these networks. Restraining governance power is not an end in itself, but it is a prerequisite to guarantee lawful governance and fundamental rights and freedoms.

A key strength of the PA paradigm is its natural focus on lawfulness. Since lawfulness of the administration is the prime focus of this paradigm, research usually addresses the legal basis of routines and procedures in administrative agencies and the mechanisms that affect the discrepancies between law and everyday administrative practices. Bureaucratic culture and a bureaucratic ethos may have many disadvantages, but lawfulness is its key value and main benefit. Since governance networks are preoccupied with effectiveness going back to the traditional PA version of lawfulness would endanger the effectiveness of these networks. A third strategy to bring the law back in the PG paradigm would thus be to bring in rule of law-related values such as lawfulness, legal equality, and legal certainty as well as infuse governance networks with constitutional competence (Cooper 2007). This requires replacing the rule-driven concept of law with a value and principle-driven constitutional or rule of law ethos.

The original law-government nexus in the PG paradigm considered law in a neutral sense as the outcome and negotiated context of network governance. Bringing the law back in adjusts this nexus by adding an emphasis on clear and determinate public responsibilities, lawful network action, checks and balances both within governance networks and as checks on governance networks, and, above all, constitutional values and competencies of both public and private actors.

Conclusions: Possibilities and Limitations

In this article, we argued that that law and public management theory should be reconnected now in order to avoid a foreseeable crisis of the current PG paradigm. Law never left public management theory, yet the approach to the link between law and government did change. We refer to this link as the law-government nexus. In this nexus, law refers to the rules, rights, principles, procedures, checks, and institutions as laid down in legal documents, while government refers to the ensemble of public agencies that govern societies. Competing, connected, but also successive paradigms in public management theory have guided different versions of the law-government nexus. Bringing the law back in public management theory neither means that a new paradigm has to be invented nor that we should reverse history and go back to the PA paradigm with its over-legalized conception of public management. We need to address the weaknesses of the PG paradigm with regard to its law-government nexus in order to restore the under-legalized conception of public management.

We explored three strategies to reconceptualize the law-government nexus in order to constitutionalize the current PG paradigm. These strategies offer a renewed law-governance nexus within the PG paradigm. Bringing the law back in adjusts this nexus by adding an emphasis on clear and determinate public responsibilities, lawful network action, checks and balances in and around governance networks, and, above all, rule of law related values, constitutional competences and a constitutional ethos of both public and private actors. Adopting the redefined law-government nexus in the PG paradigm opens up new research directions of which three can be mentioned. Firstly, as Wise (1990) has already argued, network governance reconfigures public and private organizations. While administrative law and administrative procedure still predominantly focus on the organizations of the state, comparative research provides both conceptual and legal tools to regulate network governance (see also Rose-Ackerman and Lindseth 2010). Comparative research on the regulation of private organizations with public tasks will also show what works and why certain legal solutions do or do not work.

Secondly, research and a comparative analysis may provide clues for a constitutional and legal framework on network governance. Even if both the actions of public institutions and private organizations with public tasks are regulated, the structure of the network, the interactions in the network, and the outcomes of these interactions largely take place below the legal and constitutional radar. While some degree of flexibility and adaptability is required to achieve effective network governance, legal and constitutional checks may provide legal protection of actors both within the network and outside the network. Since the traditional constitutional checks and balances do not suffice, research may develop and test new types of checks and balances for network governance.

Finally, much research has been published on customer orientation and public sector motivation in both public and private organizations (e.g., Perry, Hondeghem, and Wise 2010; Pollitt 1990; Walker et al. 2011). Research on constitutional competence and constitutional ethos in governance networks will add to a better understanding of the law-government nexus adopted by actors in governance networks. The necessity of this research direction should not be underestimated. Research may even show that actors who engage with network governance already have a strong focus on lawfulness, legal certainty, and legal equality, which are the key values of constitutional government and the rule of law. Whatever the outcome, this research direction emphasizes that redefining the law-government nexus is not only about restructuring governance networks. In the end, all public management theory is about public managers. Bringing the law back in by redefining the law-government nexus thus entails revising the values and the behavior of public managers.

In conclusion, bringing the law back in means that we should not live in the past. Rather, we should move beyond nostalgia for what was and selectively pick the best components of past paradigms in order to secure the future of public management theory.

References