The Standardization of “Good Governance” in the Age of Reﬂexive Modernity

Standards, indicators, and benchmarks are well-studied instruments of global governance. What has received less attention so far is the trend toward the standardization of governance itself. Catalogs of “good governance” are drawn up for universal use and promise to optimize the processes of governance. International organizations, development agencies, and academics compile such catalogs to make the quality of governance measurable, comparable, and “rankable.” Good governance standards have attracted sustained criticism. They project normative concepts from the West, where they are usually formulated, to other parts of the world. As societies and cultures are not standardized, it is problematic to expect standardized tools to have uniform and beneﬁcial effects regardless of local context. Good governance as a ready-made toolbox also is at odds with the very ideal of democracy as self-determination. On the other hand, good governance standards can have empowering effects when they enable societal actors to challenge oppression, corruption, or violations of human rights. How can we make sense of this ambivalence? In this article, we argue that the contested standardization of governance, with all its ambivalences, represents one of the typical reﬂexive loops that characterize the radicalization of the modernization process that sociologist Ulrich Beck described. Standardization, as a practice of modernity, has become highly self-reﬂexive and the debates over good governance catalogs show how, in modern functionally complex societies, different ideals of modernity can be at odds with each other. Les normes, indicateurs et points de repère sont des instruments bien étudiés de la gouvernance mondiale. Mais la tendance à la normalisation de la gouvernance a jusqu’ici bénéﬁcié de moins d’attention. Des catalogues de « bonne gouvernance » sont élaborés pour une utilisation universelle et promettent d’optimiser les processus de gouvernance. Des organisations internationales, des agences de développement et des universitaires compilent de tels catalogues pour

In this article, we study the global spread of good governance norms through the analytical lens of reflexive standardization. Standards are key instruments of modern governance. Firmly anchored in the logic of optimization and rationalization that marks industrial modernity, they are used to improve the quality of products and services, and to guarantee interoperability. Over the last decades, standardization increasingly turned to governance itself, as evidenced by the rise of good governance norms that pertain to issues such as public participation, accountability, transparency, policy evaluation, and anti-corruption. International organizations (IOs), development agencies, nongovernmental organizations (NGOs) and academics present contending definitions of what good governance is, and how the quality of governance can be made measurable, comparable, and "rankable." Even the world's most eminent standard-setter, the International Organization for Standardization (ISO), is currently in the business of developing "guidance for the governance of organizations" (ISO 37000).
Such attempts at standardizing the quality of governance globally are controversial, however. Critics argue that normative concepts from the West, where standards are usually formulated, are projected to other parts of the world, with little regard for local particularities and traditions. Others contend that technocratic conceptions of good governance are undermining the traditional ideals and practices of democracy, understood as self-governance. On the other hand, good governance standards have the potential to facilitate critical activism, when they enable societal actors to challenge oppression, corruption, or violations of human rights. There is good evidence that they can help holding powerful elites to account.
How can we make sense of this ambivalence? In this article, we argue that the standardization of governance represents one of the reflexive loops that characterize the radicalization of the modernization project. In the age of "reflexive modernity," as sociologist Ulrich Beck called it, the inner contradictions of industrial modernity and its ideational foundations come to the fore (Beck 1992). Expectations of linear progress are increasingly questioned, along with the means by which the ideals of the enlightenment have been pursued. In our discussion of Beck's approach, we distinguish a normative and a functional dimension of reflexivity and argue that this distinction can help us to better understand that there are two logics of reflexivity at play that, under certain conditions, can be at loggerheads with each other.
This article is organized as follows. In the next section, we bring together the extensive literatures on standardization and good governance. We interpret standard-setting as a typically high-modern, rationalizing practice, and the standardization of governance (most prominently in the form of standards of "good governance") as a particular variety of it. In the third part of the article, we examine the standards of "good governance" that are currently promoted. Part four discusses the arguments of the critics and carves out the deep ambivalence of standardizing governance. In the last section, we show how the theory of reflexive modernity can help us understand this ambivalence in the light of the inner conflictions of the project of modernity.

Standardization, Rationalization, and Governance
Standardization plays a pivotal role in the governance of modern societies (Brunsson and Jacobsson 2000). The ISO has defined standardization as the "activity of establishing, with regard to actual or potential problems, provisions for common and repeated use, aimed at the achievement of the optimum degree of order in a given context." 1 Standards are used to steer societies in a rather indirect way, giving rise to new practices of "governing at a distance" (Higgins and Tamm Hallström 2007;Broome and Quirk 2015a). They are regarded as technocratic policy instruments because experts tend to dominate in their formulation and application (Hülsse and Kerwer 2007, 629-31). As standards are issued mostly by independent agencies or private bodies, standardmaking routinely bypasses the traditional law-making procedures of the state (Abbott andSnidal 2001, 2009).
Any historiography of standards is likely to touch upon the character of industrial modernity in some way or another. Analytical concepts deployed in the literature to capture this intimate connection are, inter alia, "objectification," "formalization," "quantification," "regulation," and "rationalization" (Timmermans and Epstein 2010, 74). Historians and social scientists thus often "conceive the expansion of standardisation as a product of a dominant instrumental rationality" (Peña 2015, 56). The literature traces the origins of standardization as we know it today, and the factors that facilitated its global expansion, back to the second half of the nineteenth century (Yates and Murphy 2019). That period was marked by not only intense industrialization, but also economic, technological, and communicative globalization. Ever more goods and people crossed boundaries, thus further incentivizing transnational standardization of technologies and services. One should, however, steer clear of too simplistic, functional accounts of this development that cite efficiency gains and the reduction of transaction costs only. The creation of standards was often not the rational and efficient solution to an "objective" problem but resulted from protracted struggles among economic, political, and social forces (e.g., McGuire 1998, 2005).
The creation of standards was driven by industry as well as by governments and occurred with varying speed in different economic sectors. The military, in particular, played a great role in the advance of technical and organizational standardization (Busch 2011, 93-95). The apparatus of the modern army needed efficient ways of organizing itself, which prompted the search for ever new "best solutions" for internal procedures, governance of the organization, and monitoring and "optimization" of "results." Standardization also had a role in the constitution of the modern state itself as it enabled the compilation of statistics and thus increasing control over the territory and what happened there. In his careful historical analysis, Alain Desrosières has shown how statistics, being at the same time "outils de prevue" (tools of prognosis) and "outils de gouvernement" (tools of government), have been involved in the birth of a science de l'État and hence in the constitution of the modern state (Desrosières 1998(Desrosières , 2008(Desrosières , 2014also Porter 1995). This points to standardization's intertwined epistemic and political dimensions. The rise of modern sciences and their experimental techniques helped to standardize the societal production of knowledge and education (Busch 2011, 106-7).
Today, standard-setting is increasingly recognized as a "form of governance in its own right" (Hülsse and Kerwer 2007, 638). The majority of standards are nonbinding recommendations whose epistemic quality needs to make up for a lack of legal obligation (Brunsson, Rasche, and Seidl 2012, 615-16). The appeal of standards thus rests on claims about effective coordination, resource efficiency, and problem-solving. Nevertheless, voluntary standards can always be embedded into law and become legally binding if governments deem it necessary (Büthe and Mattli 2010, 441). Even standards that formally remain recommendations can become binding when public bodies explicitly endorse or reference them. For instance, the World Trade Organization (WTO) works on the assumption that trade restrictions citing consumer health or environmental concerns are admissible as long as they are "in accordance with relevant international standards." 2 To become effective, standards also need operationalization and measurement. As Busch (2011, 42-52) points out, measuring can function in different ways. Measures can work with thresholds, that is, minimum requirements for inclusion into a category, a concept, a social group, an institution, etc. They can also establish ordinal scales (rankings) or nominal classification into unranked categories (e.g., into certain subtypes). Especially rankings offer insights into the real-world impact of standards, including their unintended consequences (Espeland and Sauder 2007). In his analysis of performance measurements and indicators as policy instruments, Le Galès interprets their growing use as a "new wave of rationalization organized by the state in the classic Weberian sense" and as "one of the symbols of the transformation of governance" in many countries (Le Galès 2016, 508-9). He offers a nuanced view, however. While the spread of such indicators may reveal a diminished capacity of states to exercise their authority, governance activities may also spread into new spheres. Instead of signaling a retreat of the state, governing with indicators can be seen as a reconfiguration of the state that is trying to overcome policy failure and to find new ways of penetrating society and regaining capacity to govern (Le Galès 2016, 514-16).
Conventionally, standards were understood as instruments of governance, i.e., means by which public and private actors engage in the regulation and steering of society (Lascoumes and Le Galès 2004; Le Galès 2016). The dynamics of standardization, however, eventually engulfed the process of governance itself in a self-reflexive manner, most prominently in the form of standards of good governance. One of the earliest attempts at standardizing good governance transnationally were the "standards of civilization" promoted in the age of European imperialism (Gong 1984;Stroikos 2014). There never was a generally accepted codification of these standards but a clear sense that they described a modern way of governing that had developed in Europe and its settler colonies. This way of governing implied a disciplining of the passions and violence, rational forms of societal organization, and a will to control the forces of nature (Towns 2009;Linklater 2016;Yao 2019). Standards of civilization were an export product of the West and were imposed all over the world with the promise that they would lift colonized peoples out of their "backwardness." Yet, their more immediate, practical purpose was that they allowed for a classification of states into "civilized" and "non-civilized" in international law and diplomacy (Hobson 2006). These political uses of standards of civilization can be illustrated with a quote from a treatise on international law, published by the Scottish jurist James Lorimer in 1883: "As a political phenomenon, humanity, in its present condition, divides itself into three concentric zones or spheres-that of civilised humanity, that of barbarous humanity, and that of savage humanity. To these, whether arising from peculiarities of race or from various stages of development in the same race, belong, of right, at the hands of the civilised nations three stages of recognition-plenary political recognition, partial political recognition, and natural or mere human recognition" (Lorimer 1883, 101; emphasis in the original).
Lorimer reserved full recognition of "civilized humanity" to European states (except Turkey) and their ex-colonies in the Americas, which were ruled by white elites of European descent. Partial recognition was awarded to the "barbarians" of Turkey and other Asian states, such as Persia, Siam, and Japan, which had escaped colonization. The population of the rest of the world was declared savages. International law, according to Lorimer's doctrine, was applicable only in relations among civilized states of the first category. In relations with other territories it remained at the discretion of the Europeans how they wanted to treat them (Lorimer 1883, 102-3). Progress on the scale of civilization would lead to promotion to the first group of nations, which Lorimer expected in particular for Japan.
Standards of governance thus carry a considerable baggage of colonialism. However, and as Jack Donnelly pointed out, standards of civilization were not only vehicles of domination. The colonized could turn such universal expectations of civilized conduct also against their colonizers, denouncing selective application of their standards. "Even as the classic standard of civilization was being imposed on resistant African and Asian peoples, however, it appealed to (allegedly) universal moral values. And those values were more than a cover for the pursuit of self-interest" (Donnelly 1998, 6).
"Standards of civilization" are not just a phenomenon of the past but continue to be used as a concept until the present day. Donnelly argues that human rights now function as global normative benchmark (Donnelly 1998). Christopher Hobson sees the ideal of democracy as a "new standard of civilization" (Hobson 2008). Its dominant version, i.e., liberal democracy with free markets, defines the limits of international society and "rightful membership" in it. It structures relations between democracies and nondemocracies, establishes norms of "rightful conduct," and defines what counts as "legitimate statehood" (see also Bowden and Seabrooke 2006). The ultimate aim of such standardization is the "creation of an international society of reasonably uniform states based on a largely Western idealtype of liberal-democracy coupled to a market economy" (Bowden 2014, 616).
The projection of Western norms to other areas of the world thus did not end with the age of colonialism, even if the concept of a civilization that can be exported is treated with suspicion today. Ian Manners famously credited the European Union with possessing a certain type of "normative power," defined as the "ability to shape conceptions of what is 'normal'", while trying to defend his argument against charges that this was simply a restatement of the old colonial mission civilatrice (Manners 2002, 239). As Barry Buzan put it, Western governance norms "are no longer referred to using the term 'standard of civilisation', which along with 'barbarian' and 'savage' have largely dropped out of polite conversation. They generally come wrapped in more anodyne and bureaucratic terms such as 'conditionality', 'good governance', and 'development'" (Buzan 2014, 577).

The Catalogs of Good Governance
Today's standards of good governance are drawn up by IOs, national development agencies, NGOs, and not least by academics. It is impossible to map the entire field here. To illustrate the spread of governance standards, we will discuss some prominent examples. Among the public IOs that standardize good governance, the World Bank is a veteran. As a multilateral lender, the Bank was always interested in a sound base for development project and policies, founded in statistical evidence and amenable to cross-country comparison. Its focus on governance goes back to the development debates of the 1980s and 1990s, when liberal economists increasingly discussed the quality of government and public administration as factors conditioning economic growth and development. To measure the quality of public governance, the World Bank in 1996 formulated a set of Worldwide Governance Indicators. These indicators capture six dimensions of public governance (called Voice and Accountability, Political Stability and Lack of Violence, Government Effectiveness, Regulatory Quality, Rule of Law, and Control of Corruption; see also critical discussion in Langbein and Knack 2010). Measuring and comparing these qualities in almost all countries of the world, the World Bank's dataset quickly became widely used by practitioners and academics alike (Arndt and Oman 2006, 28).
Next to the World Bank, the Organization for Economic Cooperation and Development (OECD) was also remarkably influential in promoting good governance ideas. The OECD is a forum of sufficiently industrialized market democracies dedicated to economic growth and efficiency in government (Woodward 2004;Mahon and McBride 2009). The OECD's work on public governance began in the 1980s, but only in 1990 was a committee on public management formally established. Known by its then acronym PUMA, this committee became an advocate of New Public Management (NPM) techniques. While the enthusiasm for NPM has faded, it seems still fair to say that the OECD adopts a distinctly managerial perspective on public governance, concerned as it is with budgetary practice, human resource management, and outsourcing of services (Pal 2012).
In 1995, the OECD published a landmark report on public sector reform. That report, "Governance in Transition" (OECD 1995), and its sequel, "Modernising Government" (OECD 2005), were widely discussed among scholars and practitioners alike. Oriented toward measuring governance outputs, efficiency, and regulatory quality, the approach still invites analogies with private sector governance, in which the OECD is also quite active. In 1999, the OECD also tabled principles of "good corporate governance" that were revised twice since and endorsed by G 20 leaders in 2015. These principles aim at more than just the self-governing of businesses. They chiefly addressed policymakers and independent agencies that create the regulatory environment in which companies operate.
Subsequently, the OECD also suggested a catalog of public governance standards. In 2017, the OECD Council issued a recommendation on "open government" that calls for "a culture of governance that promotes the principles of transparency, integrity, accountability and stakeholder participation." 3 Open government thus conceived is advertised as a "catalyst for good governance." It is interesting to note that the talk of open government is displacing the traditional legitimating narrative of democracy that founded public power in the will of the people. The OECD replaces this with a more executive perspective on public sector management that imagines citizens as consumers of government services and as stakeholders of public governance who need to be duly informed and consulted. Such a managerial perspective "is likely to emphasize technocratic elements-output, expertise-and combine them with liberal constraints-the rule of law and individual rights-as well as certain forms of broader input through mechanisms of participation and accountability" (Krisch 2016, 679).
Over the last decade also, ISO has ventured into standardizing governance. Unlike the OECD and the World Bank, ISO is not an intergovernmental body but a hybrid, composed of national standardization organizations that are in part private and in part public. Next to standardizing products and services, ISO has turned to standardizing processes, such as quality management (ISO 9000), environmental 3 OECD, Recommendation of the Council on Open Government, OECD/LEGAL/0438, December 14, 2017, Article I. management (ISO 14000), compliance management (ISO 19600), and risk management (ISO 31000). In 2016, ISO set up a technical committee with a rather broad remit to work on "standardization in the field of governance" (ISO/TC 309). This committee has the task to "develop an international understanding of what good governance of organizations should look like." 4 Distinctive of ISO's approach to codifying good governance is the presumption that the resulting standards of governance are applicable to all types of organizations, no matter if public or private. The first standard of governance finalized by the technical committee in 2017 was ISO 37001 on anti-bribery management systems. This standard contains requirements for compliance and enables an organization to obtain external certification of its anti-bribery measures. Currently, under discussion is a more general standard ISO 37000 entitled "Guidance for the Governance of Organizations," due to be published in the autumn of 2021. It responds to the perceived "need for a market-driven International Standard based on knowledge and information on which there is global consensus that gives relevant guidance on governance for all types of organizations." 5 Here, ISO clearly adopts the OECD notion of "governance as management." National institutions also promote conceptions of good governance with transnational reach. For instance, the German federal ministry for economic cooperation and development (BMZ) suggests a conception of good governance that requires "respect for, protection and fulfilment of all human rights"; democratic and rule of law principles, such as "popular participation, a responsible parliament, lawfulness of state action and an adequate division of powers"; "a transparent, efficient and citizen-orientated administration"; fight against corruption; and, interestingly, "a cooperative stance within the international community" (BMZ 2009, 6). NGOs are also in the standard-setting business, normally concentrating on one field of activity. Perhaps most prominently, Transparency International has been instrumental in codifying norms against corruption and become a point of references in its measurement through the Corruption Perceptions Index.
Not least, academics have made suggestions of what good governance is. Particularly influential was Bo Rothstein's approach to measuring the "quality of government," constructed around the core concept of impartiality (Rothstein and Teorell 2008;Rothstein 2011). The normative goal here is governance that is "trustworthy, reliable, impartial, uncorrupted and competent." 6 With his collaborators, he developed an index to measure the quality of government empirically and founded an institute dedicated to such studies. Law scholars have also been active in this field. Daniel Esty developed a "toolbox" for good governance that contains "(1) controls on self-dealing, corruption, and special interest influence; (2) systematic and sound rulemaking; (3) transparency and public participation; and (4) powersharing" (Esty 2006(Esty , 1524.
Good governance, it seems, is becoming standardized in many places, resembling the piecemeal and decentered way in which many standardization processes unfold. There is not one global center of standardization but rather a range of different actors drawing up competing codifications.
Nevertheless, existing catalogs of good governance show a remarkable overlap of criteria and recurring components (Gisselquist 2012(Gisselquist , 2014. Van Doeveren (2011) found that conceptions of good governance are converging around the five principles of accountability, efficiency and effectiveness, openness and transparency, participation, and rule of law. While there still is notable diversity among conceptions of good governance, there is something like a core that seems to have stabilized over time. The more critical analysis of standardization and good governance, to which we turn in the next section, also tends to identify and problematize certain core features that link the prima facie diverse aspects of good governance.

The Critique of Standardization and Good Governance
Critical approaches to standardization and good governance have taken their clues from different theoretical foundations and disciplines. For one, science and technology studies (STS) alert us to issues of (in)visibility, exclusion, and domination. Just as technical objects and physical infrastructure reflect entrenched values, norms, and cultural representations that are prone to prejudice (Bowker and Star 1999, 37-38;, it is plausible to say that standards also "have politics" (Winner 1986, 19-46). In many ways related are Foucauldian governmentality approaches that highlight the role of standards in rendering certain issues "governable" (Erkkilä and Piironen 2009;Hansen 2012). One of the main points of Michel Foucault's work on biopolitics was to show how, through the use of statistics, both the people as a metaphorical body and peoples' physical bodies became subject to political control (Foucault 2009). In a similar vein, anthropologist James C. Scott focused on the nexus of power and knowledge in the modern state. The increasing capacity of the state to shape societies with great public schemes was premised on practices of measurement, categorization, and simplification (Scott 1998).
As explained above, governance with standards requires classifications and quantifications in the form of indicators, benchmarks, and rankings (see Espeland and Stevens 1998;Bowker and Star 1999;Hansen and Mühlen-Schulte 2012;Hansen and Porter 2012). No matter if standards provide precise targets or minimum thresholds, compliance with them needs to be measurable and that gives rise to a "politics of numbers" (Broome and Quirk 2015b), an angle from which much criticism has been leveled at standards of good governance. Building on Foucault's work, Hansen (2012), for example, draws attention to the way in which practices of quantification and comparison constitute "spheres of governance" and "governable subjects." Issues, actors, and relationships are made "governable" when measurement renders them visible.
Quantification in governance is often criticized for the expansion of state power to new areas of social life. Moreover, it is a partial and selective way of defining both the issues that are to be governed and the most appropriate way of doing so, strongly influenced by underlying assumptions and dominant paradigms co-produced by science and politics (cf. Allan (2017) on how climate change was constituted as an object of global governance). Far from being neutral (and more or less faithful) representations of the social world, indicators have a performative dimension in which they act upon the phenomena they seek to describe (Hansen and Porter 2012).
In response, critical approaches often use strategies of making visible the "event" or political decision that was initially behind a standard (Foucault 2007(Foucault [1978, 59). Stan-dards, in that view, are not merely bureaucratic or technical devices resulting from a neutral application of expertise. Rather, they are, intertwined with questions of power and can thereby reinforce structures of domination, discrimination, and exclusion (Porter 1995;Busch 2011, 42). In reaction to being evaluated, observed, or measured, individuals subjected to governance by indicators frequently devise strategies of "gaming the numbers" (Espeland and Sauder 2007, 6). This "reactivity" can have perverse effects when actors change their behavior in response to a standard or indicator in a way that contradicts the original intentions behind the instrument.
Standardization, in its pervasiveness, can be described as extending the "infrastructural" power of the modern state, that is, "its capacity, for good or for ill, to penetrate its territories and coordinate social life" (Timmermans and Epstein 2010, 82). Given the central role of (technical) expertise in the creation of standards, we might frame standards as part of the infrastructure of modernity, or more precisely as part of knowledge infrastructures and infrastructures of governance. As such, standardization can also create enormous political, material, and epistemic lock-in effects. The vibrant debate on "moving beyond GDP" as the key indicator of economic and societal well-being shows how a widely used indicator can shape the structure of societies and economies, as well as academic and policymaking discourse (Schmelzer 2016). Even in the case of a palpable conceptual shift, it takes a considerable amount of time for a change in discourse to translate into new indicators. From a praxeological point of view, it may also be that policymakers are prone to stick to the existing tools since at least they still provide some clear practical usability. 7 Political dynamics can further complicate matters as another prominent example that brings us back to standards of civilization shows, namely that of the distinction between developed and developing countries. For one, as Farias (2019, 669) cogently puts it, despite being analytically increasingly outdated (never mind its normative questionability), this concept "remains firmly embedded in a myriad of real-world operational purposes (e.g., access to preferential lending arrangement and trading schemes)." It also lives on due to its use as a "self-identifier [that] is in fact shored up by most emerging powers' interest in reinforcing their developing country identity, based on self-interest" (Farias 2019, 669). Dislodging a deeply culturally, politically, and technically entrenched standard or indicator can thus be a herculean effort-the prospects of questioning modern society's adherence to indicators tout court in governing public affairs seem to be even more limited.
Critical approaches tend to link standardized measurement and quantification in global governance to a "neoliberal" political rationality (Higgins and Tamm Hallström 2007;Fougner 2008;Gibbon and Henriksen 2012). On a technical plane, indicators used for measurements of good governance are often criticized for a strong bias to business perspectives (Kurtz and Schrank 2007). Via the concept of neoliberalism, further connections are drawn to phenomena such as audit procedures in "New Public Management" (Power 1997). While these affinities are obvious, the historically informed literature warns us not to associate such governance instruments exclusively with neoliberal ideology. "Scientific" approaches to governance that rely heavily on statistics and quantitative evidence date back way before the era identified with the "rise of neoliberalism" in the 1970s. They were, for example, also a key feature of Keynesian policymaking (Le Galès 2016).
The narrative of one single ideological rationale that drives modern standardization also turns out to be unduly simplistic when looking at the actors involved. For example, De Francesco's analysis of governance performance benchmarks of the World Bank and the OECD shows how different actors involved in the creation and promotion of such instruments follow different rationales, especially when they are competing in the promotion and legitimization of their indices (De Francesco 2016). IOs and NGOs are competing for attention and resources and use the creation of measurement instruments as a "strategic tool for producing authoritative expertise-or at least the public appearance of expertise" (Broome and Quirk 2015a, 824).
When we look at it from a global perspective, the standardization of governance can be characterized as one way in which normative concepts are being projected from the core to the periphery of the modern world (Abrahamsen 2012). The idealized institutions and practices of governance have emerged in some of the most advanced industrialization countries, and in their entirety are present in very few of them-"Sweden or Denmark on a good day, perhaps" (Andrews 2008, 380). Critics of such idealizations question the "one size fits all approach" of governance indicators and find a lack of sensitivity to local context and empirical realities of politics (Harrison 2005;Grindle 2007).
According to Löwenheim, governance rankings also tend to shift responsibility for "good governance" to the national level and distract from the responsibilities of powerful international actors for socioeconomic and political problems in poor and instable countries. Good governance indicators function as standards of normalcy in a discursive field where legitimacy and responsibilities are assigned (Löwenheim 2008). Even though most examinations of governance performance are not obligatory for the countries involved, and even if a country might object to a rating undertaken, for example, by the World Bank, they can hardly escape their effects when powerful members of the international community rely on such instruments, for example, when deciding over the granting of aid.
Scholars from the Global South in particular criticize that "the 'good governance' discourse does not admit of such relationships of power. Rather it presents itself as a moral paradigm, distinguishing between the good, the bad and the evil" (Shivji 2003). This criticism points to the conceptual peculiarity that the governance discourse takes the entitlement to rule as given and unproblematic. The question is no longer who exerts power and why this is legitimate, but how that power is to be exercised. In the context of corporate governance, this may even be a plausible conception. When transferred to the public realm, that construction clashes with the ideal of popular sovereignty. Universally best solutions of governance are not grounded in the will of a populace but justified by pointing to the competence of experts who devised them. Erkkiläa andPiironen (2009, 2014) pinpoint these depoliticizing effects of the good governance discourse.
On the other hand, codified standards of governance can have empowering effects. Social movements, too, have developed practices and policy instruments involving quantification, rankings, and performance measurements (Le Galès 2016, 517; see also Broome and Quirk 2015a). Such movements are part of the broader phenomenon of "statactivism" that comprises cases where different types of actors use existing standards to demand accountability, denounce specific indicators, rally against an entire set of quantifying practices, or even come up with own metrics and statistical representations to render certain issues visible in the first place (Bruno, Didier, and Vitale 2014). Such movements have a considerable history, dating back to the nineteenth century's social reformers, union militants, and scholars critical of capitalism, and hence played their role in the "extension of social rights and of systems of social protection" (Desrosières 2014, 349). This is a crucial point since it leads beyond simplistic images of domination by the state and helps us consider the possibility of critical reappropriation and subversion of governance instruments, thus pointing to a more complex interplay of dynamics of depoliticization and repoliticization (Wood and Flinders 2014). In fact, standards have also been critically used by activist organizations to denounce corruption, collusion, nepotism, and violations of human rights. 8 One of the most prominent examples would be how Transparency International advanced the fight against corruption by setting one of the de facto world indicators on this issue and can plausibly said to have strongly influenced existing standards drafted by other actors (Wang and Rosenau 2001;Larmour 2006). The Corruption Perception Index "constitutes a reality, based on a widely accepted intuition, and by doing so it sets the epistemic ground on which the normative and policy debates about corruption take place" (Urueña 2018, 384). In a similar vein, the World Justice Project, founded in 2006, promotes the rule of law and created the Rule of Law Index to measure the impartiality of justice, access to legal remedies, the openness of government, and accountability of public institutions. Not content with the measurement of corruption and other governance defects, the organization Global Integrity decided to focus instead on measuring integrity of governance as the "opposite of corruption." Standards of good governance seem not only to constrain, but also to critically enable societal activism. As discussed above, once standardization of good governance has taken hold, it does not seem easy to turn back the clock. Under conditions of modernity, it seems to take a standard to defeat a standard or, by analogy, a new governance indicator to defeat an old one.

The Reflexive Standardization of Governance
So far, we have interpreted the rise of good governance and its indicators as part of a trend toward standardization that is linked not only to the material realities of industrial society, but also to an intellectual "project of modernity" aiming at rationalization and universal progress. We discussed proposals for good governance standards as submitted by proponents, but also the arguments of their critics. What emerged from the discussion was a deep ambivalence: standards of good governance and their indicators can be a technocratic instrument of steering societies from the top, while they can also be used for emancipatory purposes and civil society action. To make sense of this ambivalence, we focus in this section on the reflexivity that is built into the standardization process.
To understand this reflexivity better and shed light on the particular role that standardization plays in modernity, we turn to sociological theories of "reflexive modernity." Proponents of this theory, such as Ulrich Beck, Anthony Giddens, and Scott Lash, identified increasing reflexivity as a typical feature of late modernity, when modern societies began to critically scrutinize their own foundations. They claim that this happened when the unintended consequences of the "first" industrial modernity became increasingly felt (Beck, Giddens, and Lash 1994). For Beck in particular, it was awareness of the mounting ecological crisis that ushered in this reflexivity and forced industrial modernity to re-examine its own goals, instruments, and procedures (Beck 1992).
Reflexive modernization theory rejects the idea that with the crisis of industrial modernity in the twentieth century, we have entered a new, postmodern era. "What 'reflexive modernization' refers to is a distinct second phase: the modernization of modern society. When modernization reaches a certain stage it radicalizes itself" (Beck, Bonss, and Lau 2003, 1). This radicalization can produce seemingly paradoxical effects when the so far "unquestioned basis of modernization is itself examined in terms of its rationality" (Beck, Bonss, and Lau 2003, 16). Yet, theorists of reflexive modernity see continuity in the process of modernization rather than a radical break or rupture. They argue that the very potential for (self-)critique had been inscribed in the Enlightenment project of modernity from the very beginning (Eisenstadt 2000). What disappears is the unquestioned belief that modernization will almost automatically lead to universal progress and welfare. This disenchantment with the project of the Enlightenment, rather than its rejection tout court, marks the transition from classic (or "first") to reflexive (or "second") modernity.
Reflexive modernity is often assessed either as a commendable self-critical examination of modern society or as revelation of the normative hypocrisy of the project of modernity. Both approaches help little to understand the origins and real-world implications of reflexivity. To develop a sharper lens with which to see the conflicted reflexivity in the standardization of governance, we propose distinguishing between a normative and a functional dimension of reflexivity, which are nevertheless intricately linked. The normative dimension is visible in the universalization of enlightenment values. Beck shows that certain modern ideals, which were previously only selectively applied in society, become applied more broadly in the age of reflexive modernity. This leads to the paradoxical effect of undermining what was often thought to be the pillars of modernity. Beck gives the example of how the move toward gender equality meant that women claimed many of the privileges previously granted only to men. In denouncing the incomplete realization of equality, critical activists uncovered one of the many ambivalences of a "first" modernity, which did not fully live up to its own ideals. By realizing the ideal of equality more thoroughly in the gender dimension, reflexive modernization undermined one of the pillars on which the "first" industrial modernity and many of its institutions rested in practice: a strongly gendered labor-market privileging the male breadwinner (Beck 1992, 104-10).
Next to the normative dynamics, there is a functional dimension of reflexivity, which is intimately related to the functional differentiation of modern societies. Functional differentiation renders modern societies increasingly complex and fragments them into a myriad of subsystems. Beck here takes up Niklas Luhmann's version of functional differentiation theory, according to which all subsystems of modern societies, like jurisprudence, the economy, or science have their own inherent logic, which distinguishes them from other subsystems (Luhmann 1995). Functional reflexivity firstly means that a social subsystem applies the logic that it uses in dealing with the outside world (which Luhmann calls "the environment") to its own practices. Beck's prime example for functional reflexivity is modern science with its procedures of critical scrutiny and falsification, which scientists used to study nature and human action. When traditional epistemology morphed into a sociology of science, it began to critically examine and ultimately deconstruct the foundations on which it was built, that is, objectivity and impartiality (Beck 1992, 165-67).
But secondly, functional reflexivity may also occur when the logic of one subsystem begins to pervade society as a whole and is then confronted with itself in different forms. If we follow Beck, this happened with the scientification of society that gave rise to what is called a "knowledge society." This trend had some paradoxical effects. Although alternative ways of producing knowledge outside of science (such as "local knowledge" or "traditional knowledge") are increasingly acknowledged, even critics of modern science often resort to scientific argument (although mostly pseudoscience or selectively applied) in their attacks on modern science (Beck 1992, 161).
As a technique of optimization and rationalization, standardization of products and services is a phenomenon intimately tied to Beck's "first modernity." The standardization of procedures (including governance), in contrast, can be classified as a reflexive phenomenon in a very basic functional understanding because it applies the technique it used in dealing with its subject matter (standardization) increasingly to itself. The turn to standardizing procedure that has accelerated, roughly, since the 1980s is visible in many places. It also has a normative dimension when it brings values, such as equality, into play. One of the best examples to illustrate the two faces of the increasingly reflexive standardization is the "essential requirements" for processes of standardization that the American National Standards Institute (ANSI) publishes. "Due process means that any person (organization, company, government agency, individual, etc.) with a direct and material interest has a right to participate [in a standard-setting process] by: a) expressing a position and its basis, b) having that position considered, and c) having the right to appeal. Due process allows for equity and fair play" (American National Standards Institute 2016, para 1.0). As with many procedural standards, the aim here is to define thresholds ("minimum acceptable due process requirements") rather than detailed specifications of what needs to be done. ANSI's proposals for the reflexive standardization of standardization not only epitomize the broader trend to critically reflect on procedures of governance in order to optimize them. They also transcend the traditional aims of "first modernity" in standardization, that is, maximum efficacy of function and efficiency of resource use. The norms of equity, fairness, and due process are also typically modern values that are extended here to the sphere of standardization procedures.
In fact, the term "reflexive standardization" has been used in the standards literature as shorthand for inclusive and participatory processes of standard setting that have numerous feedback loops built in (van Loon 2015). 9 Van Loon 9 Reflexive standardization thus conceived has affinities with the idea of metagovernance, or the "governance of governance" (Sørensen and Torfing 2007), because it addresses "the organization of the conditions for governance in its broadest sense" (Jessop 2002, 48;Meulemann 2008). According to Kooiman and Jentoft, meta-governance aims "to guide the institutional and problem-solving choices, ex ante in underpinning future-oriented activities, and ex post in appraising governance activities by serving as criteria for evaluation" (Kooiman and Jentoft 2009, 823, original emphasis). Historically, the concept of metagovernance was introduced to describe transformations of the state and its way of governance when faced with the forces of globalization and the rise of new powerful nonstate actors (Jessop 1997, 574-75). It responded to the pluralization of substantial concerns and critical audiences that questioned established ways of "doing politics" in the second half of the twentieth century. emphasizes what we called the normative dimension of reflexive modernity, the realization and balancing of values (van Loon 2015, 20-21). Others drew more on Beck's functional line of argument, when they employed the term "reflexive standardization" to mean that standardization processes are struggling with their own unintended side-effects (Hanseth et al. 2006). Reflexivity here is arising from the complexity of standardization in which attempts to establish a certain order via standards can unintentionally create disorder elsewhere or interfere with other orders.
Reflexive standardization in the field of standards of governance often implies that those who govern (critically) examine their own institutions and procedures and seek to improve them. The impetus for standardizing governance mostly comes from the governing institutions themselves and not from citizens (as in the phenomena that Beck describes). Even many of the prominent civil society organizations active in the field of good governance (such as TI and the Global Justice Project) are closely related to governing institutions, which in turn use their indices and rankings. These organizations are formed by societal elites rather than being grassroots movements, in the sense of self-organization of the subaltern. They are vulnerable to charges of codeveloping and pursuing policy programs that these elites favor.
Hence, this (self-)reflexivity of the executive stands in a tension with another modern ideal of governance, which is democratic participation. What critics of standardized governance often assert is that the process in which norms of good governance are developed and deployed excludes the people who will have to live under the government in question. It is not a standard of governance that they have chosen. To the extent that standards of good governance comprise standards for the possibilities of self-government, imposing them top-down is also at odds with the very democratic ideals they purport to insure.
Reflexive standardization is thus characterized by the same ambivalence that science faces in late-modern societies. While pervading all realms of society, science finds itself caught between the technical-professional exclusiveness that its own standards of rigor and replicability demand and the normative ideal of democratically inclusive forms of self-governing, legitimation, and accountability. In short, the normative ambition of reflexive standardization (transparency and inclusiveness) is at odds with its functional form of application that continues to rest on exclusion.
Note that reflexive standardization is exclusive not just in terms of its organizational form, being still dominated by experts, but also in terms of its content. As our review of empirical examples in the previous sections shows, the discourse on good governance tends to focus on the executive dimension of governance, and to a lesser extent, on the judicial dimension. As Fukuyama (2013, 347) said about the state of research on good governance, "everyone is interested in studying political institutions that limit or check power-democratic accountability and rule of law-but very few people pay attention to the institution that accumulates and uses power, the state." While there is a rich body of measures of the quality of democracy (like the Freedom House or Polity indexes), such indicators focus on how power is limited or checked, and far less on how it is constituted and legitimated. This is equally true for measures that claim to aim at more "substantive" dimensions of democracy, such as the effective democracy index (Alexander, Inglehart, and Welzel (2012) for an overview of the debate around it). Corruption indices might be a useful proxy to examine the quality not just of the executive but also of the legislative process as well, at least when they have a broader under-standing of political corruption (Ceva and Ferretti 2021, 34-36). However, the point is that discourses around good governance take the power exerted in governance as given and do not look at how it is legitimized in the first place. From the perspective of democratic theory, however, it is almost imperative to ask such critical questions when good governance is at issue.
To summarize, what we observe in the discourses over good governance is a sequence of critical questions posed to good governance or democracy indices that are likely to lead to ever new refinements and revisions. There is a spiral of critical (self-)interrogation at work here, a debate that characterizes reflexive modernity and is unlikely to lead to anything like closure or universal agreement any time soon. As long as we believe that governance can objectively be improved, and not just shaped in one way or the other, we will not be able to exit the critical debate over indicators and standards.

Conclusion
In this article, we studied the trend from using standards as an instrument of governance toward a reflexive standardization of governance itself. We interpreted competing conceptions and indicators of good governance as a key example of this trend, thus bringing together the standardization and good governance literatures. We then scrutinized the arguments of advocates of good governance catalogs, as well as the criticism leveled against them. It emerged that the standardization of good governance is riddled with ambivalence. First of all, it appears as a rather old phenomenon when we consider "standards of civilization" from the colonial era as a set of good governance norms pretending to be globally applicable. Unlike those moralizing and chauvinist "standards of civilization," today's standards of good governance come in a technical language of efficiency, suggesting that they are evidence based and have been put to rigorous empirical test, even if this is rarely true.
We agreed with the critics that today's good governance agenda has affinities with neoliberal policy programs when citizens are framed as consumers of governance services, when power relations or distributive questions are widely disregarded, and when the smooth functioning of market economies becomes the primary objective of good governance. Yet, we argued that it would be simplistic to attribute the rise of good governance as a concept and ideal solely to the impact of neoliberalism, or an interest in governmentality. Notions of good governance not only smoothen the exercise of power, but can also complicate it. They enable civil society to challenge an oppressive state or even the very IOs that promote good governance catalogs. The World Bank, for one, was confronted with the principles it has been preaching elsewhere (Woods 2000). Thus, good governance catalogs can have emancipatory effects and prompt critical discourses about the very governing institutions that issue them.
Rather than trying to close the debate by joining one side or the other, we tried to better understand this ambivalence. Our suggestion in this article was to see the contested standardization of governance as a typical phenomenon of "reflexive modernity." In Beck's conception of reflexive modernity, industrial modernity critically turns to its own intellectual foundations. It calls into question expectations of linear progress but is unwilling to give up the original enlightenment ideal of rationalization. Standardization, as a typically "first-modern" practice has become highly self-reflexive and debates over good governance catalogs show how modern ways of governing societies are critically scrutinizing their own principles and procedures. The tensions between ideals of self-determination and the promotion of "best solutions" that are known in advance and (allegedly) universally applicable are unlikely to disappear. As long as we believe that governance can objectively be improved, we will hardly be able to exit the critical debate over the appropriate standards, indicators, and measurement.