1Research projects surrounding notions of “policy instruments” or “government technologies” have multiplied recently in the Francophone world, whereas Anglo-Saxon scholars have debated the subject continuously for over 40 years. L. M. Salamon and O. V. Elliott’s collected volume from 2002 illustrates this fact. [1] Christopher Hood, author of seminal texts in this area [2] cites three phenomena to explain the recent renewal of interest in studying public policy instruments, a sub-field of public policy analysis. First, the emergence of new contexts for public policies (the environment, urban life) or the re-orientation of older policy areas (health, security) have given rise to a new series of questions about how public policy is made, and have brought about numerous reforms. Second, and closely linked to these changes, developed countries have been more and more apt to draw from neo-liberal or managerial models inspired by market regulation. Together with the development of new information and communication technologies, [3] these changes have produced favorable conditions for the dissemination of approaches to public policy leadership which are based on measuring performance. Finally, and often in response to this latter dynamic, Hood points to critical approaches denouncing the onset of a “surveillance society” or, at the very least, the potential risks associated with an increase in new means of control and the normalization of social phenomena. Elements under scrutiny include not just functionalist questions or the pursuit of efficient technologies, but also the concrete modalities of exercising power, or the structures of domination in the Weberian sense.

2Before examining in detail the types of analyses pursued over the past 40 years on the question of “instruments”, it may be useful to review what they have contributed to public policy analysis and to political science more broadly.

3Whether in relation to public policy, or beyond to political regimes, the notion of “change” recurs regularly in our field. An approach via policy instruments represents an excellent way to record change, beyond the political will of actors and political rhetoric. In this sense it constitutes a new methodological perspective for studying public policy: by focusing on the observation and analysis of aspects of change which are as discreet as they are meaningful, we can grasp public policy as a material phenomenon. Among the factors which allow us to characterize public policy, instruments constitute a relatively independent variable (although they are also the product of a particular history and particular mobilizations) which are key to our understanding of it. The use of instruments as an analytic tool is an opportunity for revisiting – even re-interpreting – public policy decisions. Instruments make intentions concrete, and often allow us to recognize more clearly what constitutes a true innovation, and what is a recycled idea or a half measure. More generally speaking, there are two major benefits provided by the study of instruments. First, as far as political sociology is concerned, instruments may constitute a concrete marker of public policy and may as such help to grasp transformations in the relationship between governing and governed. Questions about how individual and collective behavior is influenced, and of the legitimacy of the methods used, are more current than ever. Understanding how instruments are used may help us to grasp the transformations of the state; to envision state practices and the changes they undergo, particularly in the context of permanent tensions between constraint and exhortation. In addition, the use of instruments may help to develop reflections on political regimes and policy styles across traditional sectoral divisions. Taking an international perspective, the ways in which instruments are transposed provides fertile ground for apprehending not only how ideas and models circulate, but also the dynamics of re-appropriation and re-interpretation. Finally, a new generation of Francophone scholars is now working on interpreting public policy through instruments, and in this article we aim to highlight the work of this new generation, all the while helping to spread knowledge of Anglo-Saxon research on public policy instruments within Francophone political science, where Frédéric Varone’s work occupied a pioneering position. [4]

4This special issue of the Revue française de science politique presents a variety of recent research perspectives which focus on public policy instruments. This introduction will provide an overview of the scholarly literature on the subject, touching not only upon the seminal North American research but also on the broad directions in current research. We will show how this approach, which for many years remained dormant in research on public policy, has now come to the fore – first under the influence of North American policy analysis, and later, from neo-institutional and sociological angles. Having established these milestones, we will introduce the three specific research projects described in this issue, all of which develop different approaches on the basis of original empirical data, thus contributing new lines of enquiry and examples to the debate around the question of the use of policy instruments.

A dormant question

5Instruments of public policy remained a long dormant question in the field of public policy analysis. S. H. Linder and B. G. Peters [5] attribute the origins of the discussion to R. A. Dahl and C. Lindblom. As early as the 1950s, [6] Dahl and Lindblom had stressed the importance of public intervention techniques, issues common to all political regimes. In their eyes, the ideological schism between capitalism and socialism was of secondary importance to the choice between a multitude of “politico-economic techniques” for running government programs. For Dahl and Lindblom, the question was crucial, for while the rate of innovation in the field was increasing, the innovations themselves had scarcely been analyzed. [7]

6It is nevertheless the case that for several decades, questions regarding public policy instruments – What are their properties? How are they chosen? What are their effects and how are they used in combination? – remained largely or entirely unconsidered within the study of government action. These questions were thought of as secondary to approaches which prioritized actors, political plans, institutions, or beliefs. Instruments were first studied from a functionalist or technical point of view, and their place in political decisions was thought to be self-evident. If governing meant legislating, taxing, informing, etc., then it was simply a matter of mobilizing ready-made instruments for whatever task a government set out to accomplish, or for whatever set of “problems” it decided to address. As a “legislative solution”, [8] the instrument was part of the technical rationality of the bureaucratic model. When implemented, the instrument was scrutinized in order to evaluate its normative capacities and, in more recent times, its economic efficiency.

7Public policy instruments – or, at least, those that are judged to be innovative – have been the object of recent fascination. In response to growing criticism of the interventionist “command and control” model, governments multiplied “new instruments” of regulation of which “the contract” and “the agreement” are the most commonly invoked forms. This allowed governments to signal a change in a policy sector, or even a complete break with previous policy, without the impact of the supposed innovation actually being analyzed. A fair number of experts warned against this in their research. But the phenomenon of recycling previous mechanisms and the problems of accumulating policy instruments, which were inherent in the adoption of any new instrument, remained largely undiscussed. [9] However, there is renewed interest in innovative policy instruments as a result of the development of new public management strategies on the one hand, [10] and the emergence of new social issues on the other (the environment, [11] new ICTs [12]). Finally, the promotion of new forms of governance (particularly at the European level with the popularity of “new modes of governance” [13]) has been largely dependent upon combinations of instruments intended to ensure better coordination of networks. [14]

8From a theoretical point of view, instruments are one factor within frameworks for public policy analysis, without being of central importance conceptually. In rational choice theory, instruments are seen as institutions put in place according to short-term calculations in the hopes of reinforcing the power of decision-makers. [15] In the case of the Advocacy Coalition Framework, instruments are seen as accessory, linked to secondary beliefs (as opposed to “deep core beliefs” or “policy core beliefs”). Instruments are also seen as being concerned with technical or procedural dimensions (rules, programs, or budgets). [16] In the institutionalist or neo-institutionalist literature, notably in Peter Hall’s work, instruments are given more visibility. Choosing instruments is not seen as inconsequential, but rather as the product of the accumulated weight of institutional history, and the forces at work within it. Institutions’ capacity for action (that is, their degree of coherence, autonomy, and flexibility) is considered in terms of the extent and scope of the instruments available. [17] For Peter Hall, observing instruments is a means for studying the processes of change and learning that are specific to one particular policy. When instruments change setting and use, it becomes possible to identify a first stage of transformation. The introduction of a new instrument altogether reveals a more profound type of change and learning. [18]

Improving the governmental “tool box” and the quest for the “right” instruments

9From the mid-1970s, Canadian authors occupied a pioneering position in research on the question of instruments. [19] Canada was in fact home to the first community of scholars to pursue a structured group of research projects on the subject. Their perspective is classic policy analysis as defined by H. Lasswell; that is, the development of new means of analysis in the social sciences in order to improve upon state efficiency. The group’s approaches were essentially pragmatic, based on a search for economic rationality (cost/advantage). [20] From 1970 to 1985, significant contributors to the field included G. B. Doern and V. S. Wilson (1974), R. Simeon (1976), M. J. Trebilcock, D. G. Hartle, J. Robert, S. Pritchard and D. N. Dewes (1982). [21] Their work insists upon the need to step away from classic theories about the state and focus on the autonomous development of public policy studies in order to understand the conditions under which policies are devised and put into action. Early work consisted of analyses of sectoral policy-making in which particular attention was paid to the chosen methods of intervention: taxation, nationalization, persuasion, etc. The group’s principal focus concerned the degree of coercion involved in instruments, and in this, the group’s members drew their inspiration from the work of Theodore Lowi. [22] However, they also considered the private or public nature of the instrument as well as its appropriateness for the intended target (individuals, groups, environments). Doern and Wilson’s 1974 volume is the first to formally introduce the notion of an instrument, and to structure a series of questions around the topic. [23] The authors describe a continuum within which instruments are classified from the least coercive (distributive spending) to the most coercive (redistributive programs). The choice of instruments is made, they claim, based on a “market in governing instruments” which varies between fields and according to circumstances across the entire spectrum. This rationalist dynamic is strongly tempered by the constraints of political signposting, which seeks to privilege the instrument’s social visibility. Simeon [24] holds that instruments are one of the three dimensions of public policy analysis, in so far as they concern the means of action (means–how), positioned between, on the one hand, the object and the scope of the policy (what – scope); and on the other, between its costs, effects, and benefits (distributive – who gets what). In the early 1980s, in a governmental context favorable to deregulation, the Canadian Economic Council commissioned a study. This study, conducted by Trebilcock et al., [25] became a benchmark text because of its significant use of rational choice theory in policy analysis, which thus became “public choice”. The authors of the study emphasize the interests of elected officials and short-term electoral considerations as the principal determining factors in instrument choice, thus downplaying the importance of technical efficiency as the most important selection criteria.

10Most of the research on instruments in political science and economics from this first period (1970-1985) deals with specific instruments, taken individually and often qualified by “substantive”, “vertical”, “direct”, or “legal”. We shall cite here, among many examples, state-owned corporations; [26] economic regulation (indirect control mechanisms in certain sectors such as commissions, agencies or price-fixing); [27] capacities to spend and tax; [28] or information, public awareness campaigns, or persuasion. [29]

11This early research on instruments makes an implicit hypothesis about the existence of high levels of coordination in state decision-making processes. The instruments are not often put in the context, for example, of the type of state they are used in, the role of the actors involved, or the prevalent ideas and cultural characteristics. Nonetheless, a number of hypotheses emerge which provide a framework for subsequent work. These suggest, generally speaking, that instruments should be considered as dependent variables, and illustrate the constraints weighing upon instrument choice; the various possible reasons for continuing to use one particular instrument, or the sudden attraction of an instrument; or the popularity that certain instruments may enjoy (conventional instruments, market instruments, participative instruments, etc.). [30]

12It was also during this period that the first more-or-less direct critiques of the new public management system began to be formulated. These critiques target two main problems. First, instruments are diverse, numerous, and spread over many policy sectors, which also demonstrates the diversity, even heterogeneity, of their objectives. It is, in fact, rare to find policies that call for only one instrument. Second, there is no rational reflection on the different modes of simultaneous regulation (rules, economic or incentive measures), and criticism of the problems of coordination and overlapping which result from this were first voiced at this time. [31]

The politics of choice and typological frenzy

13In his review of twenty years’ worth of research, Christopher Hood concludes that there are today three main approaches to studying instruments. [32] The first deals with the politics of tools, that is, questions linked to choice and correspondence between a public policy objective and the means of reaching it. The second approach is of a typological nature and deals with the constant diversification of instruments and the coordination problems that result. The third approach, which treats instruments as sociological institutions, is centered upon the dynamics of permanent construction and appropriation by actors. The following sections of this introduction will present a concise synthesis of the research projects that have used these approaches. The current section will summarize the first two approaches, which follow in the steps of those described earlier in that they pursue partly pragmatic reflections on aids to decision-making; the third approach, more distanced and sometimes critical, will be discussed in the next section.

Typological frenzy

21Classification is often a key step in interpretation and thought. For Howlett and Ramesh, the question of classification is omnipresent in the study of public policy instruments. [54] The majority of authors who have undertaken extended reflection on how contemporary societies are regulated have thus differentiated between techniques of public intervention in order both to clarify them and to arrange them into hierarchies. For his thesis in political science, Sylvain Perret [55] undertook a remarkable classification project, distinguishing more than twenty typologies based on the criterion or criteria used to specify and differentiate instruments. [56] He recalls the criteria initially used by Dahl and Lindblom, [57] which use five continua: public/private, constraining/persuasive, direct or indirect expenditure restraint, voluntary/obligatory organization, dependent/independent government agencies. Although this categorization has been forgotten, it remains pertinent today and can be found to varying degrees in most of the recent typologies. In 1983, Hood developed another landmark typology [58] based on an analysis of state functions that he defines as a “tool-kit” or a collection of instruments. [59] Hood proposes to analyze the complexity of public policy by discarding elements specific to a particular field or issue, retaining only those characteristics that are fundamental to the instruments on which policy is based and which are common to many sectors. As such he pursues a classification of fundamental instruments, distinguishing between detectors (instruments that extract social information) and effectors (instruments that aim to orient different behaviors). A second differentiation, which cuts across the first, is based on the different ways in which the state may implement the instrument. There are four: information gathering (nodality), financial means (treasure), authority, and organization. This “generic” perspective studies instruments in a technical manner without referencing the organizational forms in which they are used. The eight types of basic instruments discussed by Hood may be geared toward certain people, geared toward particular groups, or geared toward everyone.

22In a less sophisticated manner, many typologies stress the different means of framing the behaviors of actors targeted by public programs. These means are grouped into three main categories: constraint, financial interest, and persuasion. The most well-known typology is that of Bemelmans-Videc, Rist and Vedung. These authors make a distinction between “sticks” (constraining regulations), “carrots” (economic incentives), and “sermons” (communication techniques). [60] Ultimately, the quest for efficiency and effectiveness drive this typological demonstration. [61]

23Most recently, interest in instruments has largely coincided with the search for new modes of governance. Lester M. Salamon’s book, The Tools of Government. A Guide to the New Governance, is an exemplary case. Transformations in the role of the state have encouraged the redefinition of paradigms to take into account the resources and constraints arising from recent economic and financial evolutions (the fiscal crisis of the state) and changes in territorial scale (globalization, multi-level action). Classic forms of government (“command and control”) are increasingly losing their legitimacy, while market interactions (particularly privatization) and public/private collaborations (contracts, voluntary initiatives) multiply. One way of approaching these complex phenomena is via policy instruments. The “new governance” perspective attempts to account for vast networks of heterogeneous actors (from the public/private or profit/non-profit sectors) in an attempt to better coordinate them. There is also an attempt to confront multidimensional issues which it is no longer possible to deal with in the context of traditional sectoral regulation. Finally, the “new governance” approach stimulates the use of more sophisticated regulatory tools, such as “non-selfexecutive” tools, which are more arbitrary and require actors’ constant mobilization and engagement; or non-directional or unidirectional instruments, which are open-ended, relying on investments made by actors to create dynamics indirectly. This search for alternative models to the traditional bureaucratic one has also favored the development of procedural rather than substantive instruments, a move which is close to the concept of a state which should concentrate on, or limit itself to, “steering” as opposed to “driving”. [62] These hybrid regulatory methods [63] do not necessarily mean a distancing of the state; [64] however, they raise important questions vis-à-vis the transparence, accountability and legitimacy of decision-making processes.

24In his most recent work, Peter John continues in this same typological vein. [65] He identifies and evaluates the chief public policy instruments in the light of past studies in order to guide policy choices. The goal: invest in the right instruments for the right reasons. He aims to use scientific knowledge to benefit the public, and he explicitly aligns this perspective with that of the founders of policy science. Because instruments are superposed, and because they are used in various contexts, John accepts that it is difficult to distinguish the specific effects of specific instruments; he proposes, nevertheless, a framework for analysis based on two criteria: the facility with which an instrument is introduced, and its efficiency (cf. table below).

Table

Government instruments as compared by Peter John

Resource Facility of introduction Effectiveness 1. Law and regulation High Weak 2. Public spending and taxation Medium Weak 3. Bureaucracy and management Medium Medium 4. Institutions Weak High 5. Information, persuasion, deliberation High Medium 6. Networks and governance Medium Weak

Government instruments as compared by Peter John

25The author gives reasons to explain each instrument’s score (high, medium, weak) according to the two criteria used. He concludes that the government should invest in regulations that allow for the greatest effects with the least resources. Without abandoning traditional instruments, John favors combinations of instruments and the quest for better management via a balance between goals and real efficiency. Although situations vary between countries and from one sector to the next, the author makes five recommendations. 1. Do not place too much emphasis on top-down approaches or traditional instruments, for this undermines public policy legitimacy and the motivations of actors; 2. In public management, incentive actions for civil servants may be beneficial; 3. Governance via networks is unlikely to produce many results; 4. Information, persuasion, and deliberation are the most promising instruments; 5. Public organizations remain promising instruments, even if reforms are long and onerous. A combination of the second, fourth, and fifth recommendations should prevent “over-regulation” of society and the wasting of public resources, and should enable long-term change.

Instrument plasticity and robustness: sociological approaches

26The third approach discussed by Christopher Hood treats instruments as “sociological institutions”. This approach analyzes two dynamics: first, the dynamic of instrument construction and adaptation, which deals with how instruments are fitted into different sectors of public policy, and how such choices are justified; and second, how instruments are appropriated by different actors and the resultant effects on the network of actors, on the matters under regulation, and, finally, the repercussion on the instrument itself. Hood places the research group directed by Lascoumes and Le Galès under this heading. [66] Whereas the first approaches recognize the political dimension of instrument choice (which is mainly a result of interests and power relationships), the sociological approach envisions instruments as institutions in the neo-institutionalist sense, that is, in their less formal, symbolic, and cognitive dimensions. The instrument as an institution is also approached via analyses of power, with the formatting of social facts which this implies; but also through the pedagogical actions, the act of framing, and sometimes the manipulations that instruments give rise to. This work reveals the extent to which instruments may have controlling effects, both cognitively and behaviorally. Focusing on discourse and, in particular, on specific practices, this research attempts to retrace past histories and to identify current impacts. Sociological institutionalism, then, underlines two phenomena. On the one hand, it takes into account cognitive frameworks, both global and specific, which the instrument draws on. The instrument is part of the general relationship of power between governed and governing, which it makes concrete in order to ensure the regulatory operation of particular sectoral groups. For instance, regulatory instruments are tied to a conception of the rule of law in which the social realm is oriented by norms, the infringement of which incurs penalties. Almost all sectors are regulated in this manner (from the oldest issues such as taxation, to the newest, such as the internet and new ICTs). Communication instruments are another example, falling within the notion of “public democracy” which they operationalize in a number of contemporary public policy sectors, such as security, pollution, public health, etc. On the other hand, this perspective aims to characterize the constant re-working and renegotiation of the conventions on which the instrument is based. An analysis of the degrees and forms of this flexibility presupposes not just an analysis of the instrument’s own internal properties – its technical and logical constraints – but also an analysis of the expected and unexpected effects which arise when the instrument is variously appropriated. In this perspective, public policy is a sociopolitical space, constructed as much by its instruments and regulation techniques as by the beliefs and strategies of its actors.

27This approach has been developed on the basis of three main approaches. Scholars have taken techniques of political domination, the industrial management sciences, and the sociology of science as starting-points. The rich literature on government technologies and their evolution constitute initial reference points on the subject: Max Weber, Norbert Elias, and the geographer, Claude Raffestin posited bureaucracy, labeling rules, and cartography as techniques of political domination, each with its own logic. [67] Their technical nature is inextricably linked to the effects of social constraint they produce and the ways in which they legitimize state positions. Michel Foucault, for example, drew attention to governmentality, that is, to the practical modes of orienting individual and collective behaviors, which led him to emphasize the role of instruments. Analysis of instrumentation allowed Foucault to differentiate, in particular, “disciplinary” powers from “bio-powers”. [68] Recent research has confirmed the absence of instruments’ axiological neutrality and has emphasized their close relationship with the imposition of power. [69] K. E. Weick, A. Hatchuel, B. Weil, and J.-C. Moisdon, within the context of industrial sociology, have demonstrated the decisive importance of management tools, [70] showing that such tools play a double role in organizations: first, they extract and synthesize information; and second, they may serve to standardize practices by diffusing norms. [71] In 1992, Hatchuel and Weil distinguished instruments through their combination of three indivisible elements: “a technical substrate”, “a philosophy of management” and “a simplified representation of the organization”. [72] Finally, this third approach may be found in the sociology of science. Georges Simondon was the first to study innovation not as the materialization of an initial idea, but rather as an often chaotic dynamic wherein information converges, constraints are integrated, and mediation takes place between divergent paths of development. [73] Simondon talks of concretization in order to account for the combination of heterogeneous factors whose interactions may or may not produce an innovation. Michel Callon or Bruno Latour’s conception of technology [74] and Alain Desrosière’s conception of statistics [75] both contain the notion of compromise and conciliation via convention. Both represent constant pairings of actors and information, which are regularly subject to re-interpretation. Taking these three factors into account – technical-social interactions, the flexibility of arrangements, and the materialization of power relations – leads to the definition of the public policy instrument as a normative device. This device is both technical and social, is generally applicable, and offers a concrete notion of the governed/governing relationship based on a specific conception of regulation.

28Alain Desrosières showed how statistical production uses a common language and representations which create effects of truth and apparent interpretation of the world; effects which are imposed upon every actor and which naturalize the social situations that statistics deal with. It is possible to see public policy instruments in the same way, stripping them of the illusion of neutrality, specifying their effects and revealing the forms of legitimacy that validate them. [76] Public policy instruments are not inert or freely available for sociopolitical mobilizations – rather, their force of action is autonomous, developing over the course of interactions with the actors who use them. Instruments are not static, and cannot be reduced to pure technicality: rather, their effects are often independent of assigned objectives. It is in this sense that appropriation by actors becomes a determining factor. However, seen as institutions, instruments also have their own degree of robustness, some more than others, over and beyond their degrees of materialization. For example, there is a considerable difference between the relative stability of a demographic classification or the calculation of a tax on the one hand, and the malleability of under-defined instruments such as Conseils de quartier (neighborhood councils) [77] or the Open Method of Coordination (OMC) on the other. [78] Indeed, the conventions on which instruments are based are often discussed, critiqued, or modified, but these dynamics are highly variable according to the intrinsic properties and usages of the instruments. In many areas, beyond the signaling of the new objectives of a public policy, it is in practice the longevity of instruments which is noticeable.

29Based on existing research, three main effects may be distinguished:

30First, instruments create an aggregation effect. An obligatory rite of passage, this is part of what Callon has called “translation activities” – activities that allow heterogeneous actors to come together around questions on which they agree to work together in a network. [79] This learning activity requires of the actors involved that they modify or move away from their initial conceptions. Correlatively, the instrument creates inertia effects which may partly explain resistance to change (tensions between actor-users) and to external pressures (governmental change, pressure from interest groups). In this sense, the instrument may be considered as an “actor-network”, [80] central to the definition of public policy programs and the changes they undergo.

31Next, the instrument produces a specific representation of the issue at hand, and in this sense the instrument produces a direct cognitive effect. Conventional definitions of social facts are imposed via a categorization grid. The instrument leads to a specific problematizing of the issue in question in so far as it arranges variables in a hierarchy and may go so far as to propose an explanatory system. The calculation of means or weighting, and the search for statistical regularities leads to causal systems of interpretation, which consistently present themselves as justified by science. Thus “unemployment”, “insecurity”, and “atmospheric pollution” describe phenomena that are first and foremost the products of artifacts: statistical measures or indexes often correlated to other variables. [81]

32Third, the instrument is never an isolated device: it is inextricably linked to contextualized modes of appropriation. Via the instrument it is possible to observe not only professional mobilizations (for example, the affirmation of new competencies) but also reformulations (serving particular interests and power relations between actors) and, finally, resistance (in order to reduce the impact of the instrument or circumvent it by creating paradoxical alliances). [82] We are wrong to interpret these dynamics in terms of the inefficiency of the instrument, as this leads us to either over-or under-interpret its initial intention. Rather, we should consider the instrument as a phenomenon of “normal sociology” (in the Durkheimian sense of the term) with its own regularities.

33*

34* *

35The articles that we have grouped in this issue complement one another in their analyses of instrument use in the practice of governing. They work to show the importance of the links between knowledge and power, demystifying the apparent neutrality of recourse to technical rationality, as well as its supposed apolitical legitimacy. Two broad approaches are prioritized here. On the one hand, a consideration of how instruments are invented, demonstrating the vicissitudes of their development, and, historically, the importance of negotiations between different actors and the compromises that must be made between norms and multiple values. Elsewhere, and often complementarily, the emphasis is placed on how instruments are used, in order to illustrate the scope of the register of potential instrument appropriation and to underline the transformative effects that different uses may bring. The current issue acts as a progress report for this developing research field, drawing attention to the diversity of approaches possible within it.

36Pauline Ravinet’s analysis of the Bologna Process in European higher education is situated within discourse on continental integration, but also within the dominant discourse on economics and the knowledge-based society. The instrument examined in her article reveals the emergence and institutionalization of a specific arena of governance, where participating countries are coordinated outside the realm of European Union policy. Ravinet analyzes the normative development of a dynamic that was at the outset voluntary and non-binding. She also demonstrates the importance of the divergent interpretations and uses made of the Process.

37Using original data, Ravinet examines the functioning and effects of the instruments resulting from the Bologna Process, as well as the relationship between these instruments and the European Union, and with the “new governance”. What began as ministers simply agreeing upon six objectives, several years later had become a process in which increasingly normative instruments had been put in place outside European Union control. These instruments were intended to allow for the learning and identification of good practice. A loose system of participation became a three-step system of monitored coordination. If the Bologna Process has ended up with mechanisms which resemble others, like the OMC, and which also draw their inspiration from practices in the private sector, these were developed in the shadows of the European hierarchy, and, as such, appear to be more legitimate. The author concludes that the Bologna Process is in fact piloted by instruments that have been reinvented and readapted for the “Bolognese sauce” which illustrates the potential of functions activated and developed by actors without their impact being essentially original or surprising. At the same time, instruments do not come from a vacuum: they must be considered in the context of specific institutions and cognitive realms.

38Using a comparative approach (looking at environmental and urban policies) that is also systematic (using 86 instruments) and longitudinal (spanning the years 1972-2006), Charlotte Halpern and Patrick Le Galès aim to determine if, at the European level, we are witnessing the emergence and development of autonomous public policy. Situating their research within work on the choice of instruments, and postulating that this is a profoundly political choice, the authors draw on a database and a classification of instruments adopted over a period of more than 30 years to show that this approach allows them to shed considerable light on public policy in the European Union, beyond formal objectives and shared competencies. In retracing the instruments’ origins and studying the effects of accumulation, two contrasting cases are observed. In the case of the environment, inspiration is drawn from national political systems and the European policy that results is characterized by disparity and inequality on the ground. In the case of urban affairs, it is regional politics that are referenced and, as a result, policies are marginalized. In both cases, to speak of “new instruments” would be to exaggerate, for in reality it is a question of combining traditional mechanisms. That European institutions appear to be quite limited in their capacity to pilot programs leads to the conclusion not only that decisions have long-term effects, but that “without autonomous instruments, there is no autonomous action”.

39Finally, Pierre-Yves Baudot’s study of the development of computerized administration in France (1960-1970) observes the varied uses of instruments over time. Instruments appear not as vectors of a unique, coherent, or stabilized doctrine, nor as unequivocal materializations of reforming interventions. Rather, instruments give consent for operations that lead actors to take action, particularly local actors who thus acquire a new position and participate in the modernization of the state. Baudot analyzes the archives of the instigators of administrative computerization, as well as those of the Interior Ministry’s Information Technology Commission. In doing so, he illustrates how people worked together without consensus on the instrument: how diverse, even contradictory meanings were given to the modernization process and to the different modes of behavior of the different collectives (in the realms of information technology, the administration, and modernization). The instrument emerges as an element of coordination for mixed groups; in practice, it ensures that a plurality of ideas may be rendered compatible. Do instruments map out ideas? Do the possible forms of understanding they propose come from the object itself, or from the actors who implement them? For the author, the responses to these questions lead us away from the initial theories on which the notion of the instrument was formed. “What is at stake is less a dependency on technical solutions and more the possibility of political action.” The article contibutes to thinking about instruments within the perspective of the sociology of translation, revealing the decisive role played by appropriations in a context where no initial agreement exists on the reforming potential of an instrument. In the end, the notion of a “border object” is employed as a way to understand the process of social construction involved in technical solutions.